Collab11 PDF
Collab11 PDF
Collab11 PDF
x
Solution Reference Network Designs
(SRND)
THE SPECIFICATIONS AND INFORMATION REGARDING THE PRODUCTS IN THIS MANUAL ARE SUBJECT TO CHANGE WITHOUT NOTICE. ALL
STATEMENTS, INFORMATION, AND RECOMMENDATIONS IN THIS MANUAL ARE BELIEVED TO BE ACCURATE BUT ARE PRESENTED WITHOUT
WARRANTY OF ANY KIND, EXPRESS OR IMPLIED. USERS MUST TAKE FULL RESPONSIBILITY FOR THEIR APPLICATION OF ANY PRODUCTS.
THE SOFTWARE LICENSE AND LIMITED WARRANTY FOR THE ACCOMPANYING PRODUCT ARE SET FORTH IN THE INFORMATION PACKET THAT
SHIPPED WITH THE PRODUCT AND ARE INCORPORATED HEREIN BY THIS REFERENCE. IF YOU ARE UNABLE TO LOCATE THE SOFTWARE LICENSE
OR LIMITED WARRANTY, CONTACT YOUR CISCO REPRESENTATIVE FOR A COPY.
The Cisco implementation of TCP header compression is an adaptation of a program developed by the University of California, Berkeley (UCB) as part of UCBs public
domain version of the UNIX operating system. All rights reserved. Copyright 1981, Regents of the University of California.
NOTWITHSTANDING ANY OTHER WARRANTY HEREIN, ALL DOCUMENT FILES AND SOFTWARE OF THESE SUPPLIERS ARE PROVIDED AS IS WITH
ALL FAULTS. CISCO AND THE ABOVE-NAMED SUPPLIERS DISCLAIM ALL WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING, WITHOUT
LIMITATION, THOSE OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OR ARISING FROM A COURSE OF
DEALING, USAGE, OR TRADE PRACTICE.
IN NO EVENT SHALL CISCO OR ITS SUPPLIERS BE LIABLE FOR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, OR INCIDENTAL DAMAGES, INCLUDING,
WITHOUT LIMITATION, LOST PROFITS OR LOSS OR DAMAGE TO DATA ARISING OUT OF THE USE OR INABILITY TO USE THIS MANUAL, EVEN IF CISCO
OR ITS SUPPLIERS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
Cisco and the Cisco logo are trademarks or registered trademarks of Cisco and/or its affiliates in the U.S. and other countries. To view a list of Cisco trademarks, go to this
URL: www.cisco.com/go/trademarks. Third-party trademarks mentioned are the property of their respective owners. The use of the word partner does not imply a partnership
relationship between Cisco and any other company. (1110R)
Any Internet Protocol (IP) addresses and phone numbers used in this document are not intended to be actual addresses and phone numbers. Any examples, command display
output, network topology diagrams, and other figures included in the document are shown for illustrative purposes only. Any use of actual IP addresses or phone numbers in
illustrative content is unintentional and coincidental.
Cisco Collaboration System 11.x SRND
2012-2015 Cisco Systems, Inc. All rights reserved.
CONTENTS
Preface
xxxvii
xxxviii
xxxviii
CHAPTER
Introduction
xxxviii
xxxix
xxxix
1-1
PART
CHAPTER
2-3
High Availability
Capacity Planning
CHAPTER
2-1
2-3
2-4
Network Infrastructure
3-1
3-3
iii
Contents
3-19
3-61
iv
Contents
CHAPTER
4-1
4-1
4-16
Contents
4-17
4-25
4-26
4-30
4-31
4-32
4-34
4-36
CHAPTER
Gateways
4-41
5-1
5-1
5-2
vi
Contents
5-3
5-14
IP Gateways 5-15
Cisco Unified Border Element 5-15
Cisco Expressway 5-16
Expressway-C and Expressway-E Deployment for Business-to-Business Communications
IP-Based Dialing for Business-to-Business Calls 5-20
High Availability for Expressway-C and Expressway-E 5-21
Security for Expressway-C and Expressway-E 5-23
Scaling the Expressway Solution 5-24
Considerations for Outbound Calls 5-28
5-17
CHAPTER
5-34
6-1
6-2
6-3
6-4
6-6
6-9
Cisco Collaboration System 11.x SRND
vii
Contents
6-14
viii
Contents
6-51
Media Resources
6-57
6-58
6-56
6-57
CHAPTER
6-48
6-60
7-1
7-2
7-7
ix
Contents
MTP Resources
Trusted Relay Point
Annunciator
7-15
7-16
7-16
7-18
7-30
7-35
Contents
CHAPTER
Collaboration Endpoints
7-43
8-1
8-2
8-4
8-6
8-11
8-16
xi
Contents
8-30
xii
Contents
8-37
8-40
8-43
8-44
8-44
CHAPTER
Call Processing
8-43
8-45
9-1
9-2
9-8
xiii
Contents
9-20
9-22
9-25
9-29
9-31
CHAPTER
10
10-1
10-1
10-1
10-3
10-6
xiv
Contents
10-7
10-23
10-39
10-40
10-44
xv
Contents
10-52
CHAPTER
11
11-1
11-3
11-3
11-4
xvi
Contents
11-66
11-69
xvii
Contents
PART
CHAPTER
12
12-2
High Availability
Capacity Planning
CHAPTER
13
12-3
12-3
Bandwidth Management
13-1
12-1
13-1
13-2
13-11
xviii
Contents
Utilizing the Operating System for QoS Trust, Classification, and Marking 13-29
Endpoint Identification and Classification Considerations and Recommendations 13-32
WAN Queuing and Scheduling 13-32
Dual Video Queue Approach 13-33
Single Video Queue Approach 13-34
Provisioning and Admission Control 13-37
Enhanced Locations Call Admission Control 13-39
Call Admission Control Architecture 13-40
Unified CM Enhanced Location Call Admission Control 13-40
Network Modeling with Locations, Links, and Weights 13-41
Location Bandwidth Manager 13-48
Enhanced Location CAC Design and Deployment Recommendations and Considerations 13-50
Intercluster Enhanced Location CAC 13-51
LBM Hub Replication Network 13-52
Common Locations (Shared Locations) and Links 13-54
Shadow Location 13-56
Location and Link Management Cluster 13-57
Intercluster Enhanced Location CAC Design and Deployment Recommendations and
Considerations 13-59
Enhanced Location CAC for TelePresence Immersive Video 13-60
Video Call Traffic Class 13-60
Endpoint Classification 13-61
SIP Trunk Classification 13-61
Examples of Various Call Flows and Location and Link Bandwidth Pool Deductions 13-63
Video Bandwidth Utilization and Admission Control 13-67
Upgrade and Migration from Location CAC to Enhanced Location CAC 13-72
Extension Mobility Cross Cluster with Enhanced Location CAC 13-74
Design Considerations for Call Admission Control 13-74
Dual Data Center Design 13-75
MPLS Clouds 13-76
Call Admission Control Design Recommendations for Video Deployments 13-79
Enhanced Location CAC Design Considerations and Recommendations 13-81
Design Recommendations 13-81
Design Considerations 13-82
Design Recommendations for Unified CM Session Management Edition Deployments with Enhanced
Location CAC 13-83
Recommendations and Design Considerations 13-83
Design Recommendations for Cisco Expressway Deployments with Enhanced Location CAC 13-86
Recommendations and Design Considerations 13-86
xix
Contents
Design and Deployment Best Practices for Cisco Expressway VPN-less Access with Enhanced
Location CAC 13-90
Bandwidth Management Design Examples 13-91
Example Enterprise #1 13-91
Identification and Classification 13-92
WAN Queuing and Scheduling 13-100
Provisioning and Admission Control 13-101
Example Enterprise #2 13-108
Identification and Classification 13-109
WAN Queuing and Scheduling 13-116
Provisioning and Admission Control 13-117
CHAPTER
14
Dial Plan
14-1
14-2
14-8
xx
Contents
14-53
xxi
Contents
CHAPTER
15
Emergency Services
15-1
15-2
15-4
15-7
15-9
15-9
xxii
Contents
CHAPTER
16
15-22
15-23
15-24
16-1
16-2
16-3
16-4
xxiii
Contents
16-31
16-32
PART
CHAPTER
17
17-2
High Availability
Capacity Planning
CHAPTER
18
17-1
17-3
17-4
18-1
18-2
18-18
18-19
xxiv
Contents
CHAPTER
19
18-49
19-1
19-2
19-2
19-4
xxv
Contents
Single-Site Messaging
Centralized Messaging
Distributed Messaging
19-4
19-5
19-5
19-28
19-47
xxvi
Contents
CHAPTER
20
20-1
20-2
Presence 20-2
On-Premises Cisco IM and Presence Service Components 20-3
On-Premises Cisco IM and Presence Service User 20-3
Enhanced IM Addressing and IM Address Schemes 20-4
Single Sign-On (SSO) Solutions 20-5
IM and Presence Collaboration Clients 20-5
Jabber Desktop Clients Modes 20-6
SAML Single Sign On 20-7
Cisco Unified CM User Data Service (UDS) 20-8
LDAP Directory 20-9
AD Groups and Enterprise Groups 20-9
AD Group Considerations for Groups and User Filters 20-9
WebEx Directory Integration 20-10
Common Deployment Models for Jabber Clients 20-10
On-Premises Deployment Model 20-11
Cloud-Based Deployment Model 20-12
Hybrid Cloud-Based and On-Premises Deployment Model 20-13
Client-Specific Design Considerations 20-14
Phone-Specific Presence and Busy Lamp Field
Unified CM Presence with SIP 20-14
Unified CM Presence with SCCP 20-16
Unified CM Speed Dial Presence 20-16
Unified CM Call History Presence 20-17
Unified CM Presence Policy 20-17
Unified CM Presence Guidelines 20-18
20-14
xxvii
Contents
20-41
20-52
CHAPTER
21
Mobile Collaboration
21-1
21-3
xxviii
Contents
21-10
21-28
xxix
Contents
CHAPTER
22
22-1
22-2
22-10
22-12
22-19
22-20
xxx
Contents
CHAPTER
23
22-21
23-1
23-1
PART
CHAPTER
24
Capacity Planning
25
24-1
24-2
High Availability
CHAPTER
23-17
24-3
24-3
25-1
25-2
25-8
25-9
xxxi
Contents
25-10
25-12
25-13
xxxii
Contents
25-47
CHAPTER
26
26-2
26-1
26-2
26-3
26-5
26-6
26-6
26-10
Using Cisco Prime Collaboration Deployment for Migration from Physical Servers to Virtual
Machines 26-10
Cisco Prime Collaboration Deployment Migration Types 26-11
Cisco Prime Collaboration Deployment Migration Prerequisites 26-11
Simple Migration 26-11
Network Migration 26-12
Migrating Video Endpoints from Cisco VCS to Unified CM
26-12
26-13
26-14
26-15
26-15
Cisco Collaboration System 11.x SRND
xxxiii
Contents
CHAPTER
27
Network Management
27-1
27-2
27-3
27-6
27-4
27-10
27-13
Provisioning 27-13
Provisioning Concepts 27-15
Best Practices 27-16
Prime Collaboration Design Considerations
Redundancy and Failover 27-18
Provisioning Ports and Protocol 27-18
27-17
xxxiv
Contents
27-25
27-27
GLOSSARY
INDEX
xxxv
Contents
xxxvi
Preface
Revised: July 30, 2015
This document provides design considerations and guidelines for deploying Cisco Collaboration
solutions, including Cisco Unified Communications Manager 11.x, Cisco TelePresence System, and
other components of Cisco Collaboration System Release 11.x.
This document has evolved from a long line of Solution Reference Network Design (SRND) guides
produced by Cisco over the past decade. As Ciscos voice, video, and data communications technologies
have developed and grown over time, the SRND has been revised and updated to document those
technology advancements. This latest version of the SRND includes Ciscos full spectrum of
collaboration technologies such as TelePresence, WebEx, and support for a wide range of end-user
devices. As Cisco continues to develop and enhance collaboration technologies, this SRND will continue
to evolve and be updated to provide the latest guidelines, recommendations, and best practices for
designing collaboration solutions.
This document should be used in conjunction with other documentation available at the following
locations:
xxxvii
Unless stated otherwise, the information in this document applies to all Cisco Collaboration System 11.x
releases.
Within each chapter of this guide, new and revised information is listed in a section titled Whats New
in This Chapter.
Although much of the content in this document is similar to previous releases of the Cisco Collaboration
SRND, it has been reorganized and updated extensively to reflect more accurately the architecture of the
current Cisco Collaboration System Release. Cisco recommends that you review this entire document,
starting with the Introduction, page 1-1, to become familiar with the technology and the system
architecture.
Revision History
This document may be updated at any time without notice. You can obtain the latest version of this
document online at:
http://www.cisco.com/go/ucsrnd
Visit the above website periodically and check for documentation updates by comparing the revision date
of your copy with the revision date of the online document.
The following table lists the revision history for this document.
Revision Date
Comments
xxxviii
Conventions
This document uses the following conventions:
Convention
Indication
bold font
italic font
Document titles, new or emphasized terms, and arguments for which you supply
values are in italic font.
[ ]
{x | y | z }
[x|y|z]
string
A nonquoted set of characters. Do not use quotation marks around the string or
the string will include the quotation marks.
courier
font
Terminal sessions and information the system displays appear in courier font.
< >
[ ]
!, #
An exclamation point (!) or a pound sign (#) at the beginning of a line of code
indicates a comment line.
Note
Means reader take note. Notes contain helpful suggestions or references to material not covered in the
manual.
Tip
Means the following information will help you solve a problem. The tips information might not be
troubleshooting or even an action, but could be useful information, similar to a Timesaver.
Caution
Means reader be careful. In this situation, you might perform an action that could result in equipment
damage or loss of data.
xxxix
Timesaver
Warning
Means the described action saves time. You can save time by performing the action described in
the paragraph.
Warning
Statements using this symbol are provided for additional information and to comply with regulatory
and customer requirements.
xl
CH A P T E R
Introduction
Revised: July 30, 2015
Collaboration means working together to achieve a common goal. Not very long ago, the best way for
people to collaborate was for them to be in the same location at the same time so that they were in direct
contact with each other. In todays globalized economy with decentralized business resources,
outsourced services, and increasing costs for office facilities and travel, bringing people together in the
same physical location is not the most efficient or effective way to collaborate. But with Cisco
Collaboration Solutions, workers can now collaborate with each other anytime, anywhere, with a
substantial savings in time and expenses.
Cisco Collaboration Solutions support the full range of voice, video, and data communications,
including the latest advances in mobile communications and social media. Cisco Collaboration Solutions
also provide an extensive set of applications and services that can be deployed either on premises or in
the cloud.
1-1
Chapter 1
Introduction
Web Conference
Customer Care
TelePresence Systems
Fundamental
Technologies
Deployment
Models
Management
On Premises
Collaboration Services
Call Control
IM and Presence
Directory
Social Media
Content
Management
Conferencing
Scheduling
Edge Services
Messaging
Media Services
Security
QoS
Hosted or
Managed
Standards
Cloud
Virtualization
Network
Compute
Storage
Medianet
348983
Figure 1-1
Collaboration Infrastructure
Cisco has long been recognized as the world leader in routing and switching technology. This technology
forms the core of the network infrastructure for Cisco Collaboration Solutions. The Quality of Service
(QoS) mechanisms available on Cisco switches and routers ensure that the voice, video, and data
communications will be of the highest quality throughout the network. In addition, Cisco gateways
provide a number of methods for connecting your enterprises internal network to an external wide area
network (WAN) as well as to the public switched telephone network (PSTN) and to legacy systems such
as a PBX. And the Cisco Hosted Collaboration Solution (HCS) enables Cisco partners to offer customers
cloud-based, hosted collaboration services that are secure, flexible, low-cost, scalable, and always
current with the latest technology.
Cisco Collaboration Systems Release 11.x is deployed using virtualization with the VMware vSphere
ESXi Hypervisor. The Cisco Collaboration application nodes are deployed as virtual machines that can
run as single or multiple application nodes on a server. These virtualized applications can provide
collaboration services for small and medium businesses, and they can scale up to handle large global
enterprises such as Cisco.
In most cases you will want your collaboration sessions to be secure. That is why Cisco has developed
a number of security mechanism to protect each level of the collaboration path, from the network core
to the end-user devices.
Once your collaboration solution is implemented, you will want to monitor and manage it. Cisco has
developed a wide variety of tools, applications, and products to assist system administrators in
provisioning, operating, monitoring and maintaining their collaboration solutions. With these tools the
system administrator can monitor the operational status of network components, gather and analyze
statistics about the system, and generate custom reports.
1-2
Chapter 1
Introduction
Cisco End-to-End Collaboration Solutions
Instant messaging (IM) and presence The Cisco IM and Presence Service enables Cisco Jabber,
Cisco Unified Communications Manager applications, and third-party applications to increase user
productivity by determining the most effective form of communication to help connect collaborating
partners more efficiently.
Collaborative rich media conferencing Cisco WebEx incorporates audio, high-definition (HD)
video, and real-time content sharing in a platform that provides easy setup and administration of
meetings, interactive participation in meetings, and the ability to join meetings from any type of
device such as an IP phone, a tablet device, or a desktop computer. For on-premises conferencing,
Cisco TelePresence Server in combination with Cisco TelePresence Conductor enables ad hoc,
scheduled, and permanent audio and video conferencing along with content sharing for
TelePresence video endpoints, video-enabled desk phones, and software-based mobile and desktop
clients.
Telepresence Cisco TelePresence technology brings people together in real-time without the
expense and delay of travel. The Cisco TelePresence portfolio of products includes an array of
high-definition (HD) video endpoints ranging from individual desktop units to large multi-screen
immersive video systems for conference rooms. And Cisco TelePresence products are designed to
interoperate with other Cisco collaboration products such as Cisco WebEx and Cisco Unified IP
Phones with video capability.
Voice messaging Cisco products provide several voice messaging options for large and small
collaboration systems, as well as the ability to integrate with third-party voicemail systems using
standard protocols.
Customer contact Cisco Unified Contact Center products provide intelligent contact routing,
call treatment, and multichannel contact management for customer contact centers. Cisco Unified
Customer Voice Portal can be installed as a standalone interactive voice recognition (IVR) system,
or it can integrate with the contact center to deliver personalized self-service for customers. In
addition, Cisco SocialMiner is a powerful tool for engaging with customers through the social
media.
Call recording and monitoring Cisco Collaboration Solutions can employ a variety of
technologies to record and monitor audio and/or video conferences as well as customer
conversations with contact center personnel. The call recording and monitoring technologies include
solutions based on Cisco Unified Communications Manager, Cisco MediaSense, Cisco Agent
Desktop, Cisco TelePresence Content Server, and Switched Port Analyzer (SPAN) technology.
1-3
Chapter 1
Introduction
Wide variety of collaboration endpoints Cisco produces a compete line of endpoint devices
ranging from basic voice-only phones, to phones with video and Internet capability, and to
high-resolution telepresence and immersive video devices. Cisco Collaboration Technology also
provides the ability to integrated third-party endpoint devices into the collaboration solution.
Cisco BYOD Smart Solution With the Cisco Bring Your Own Device (BYOD) Smart Solution,
users can work from their favorite personal device, be it a smartphone, tablet, or PC. In addition to
enhancing the work experience, the Cisco BYOD Smart Solution ensures greater network security
and simplifies network management by providing a single policy for wired and Wi-Fi access across
your organization.
Mobile collaboration Cisco mobile collaboration solutions provide mobile workers with
persistent reachability and improved productivity as they move between, and work at, a variety of
locations. Cisco mobility solutions include features and capabilities such as: Extension Mobility to
enable users to log onto any phone in the system and have that phone assume the users default phone
settings; Cisco Jabber to provide core collaboration capabilities for voice, video, and instant
messaging to users of third-party mobile devices such as smartphones and tablets; and Single
Number Reach to provide a single enterprise phone number that rings simultaneously on an
individual users desk phone and mobile phone.
Applications and service As mentioned previously, Cisco has developed many advanced
applications and services to enrich the collaboration experience for end users (see Collaboration
Applications and Services, page 1-3). Whenever possible, Cisco strives to adhere to widely accepted
industry standards in developing its collaboration technology so that you can easily integrate
third-party applications and services into your collaboration solutions. In addition, the application
programming interfaces available with many Cisco collaboration products enable you to develop
your own custom applications.
1-4
Chapter 1
Introduction
About this Document
1-5
Chapter 1
Introduction
1-6
PART
Network Infrastructure
Gateways
Media Resources
Collaboration Endpoints
Call Processing
CH A P T E R
Network Infrastructure Ensures a redundant and resilient foundation with QoS enabled for
Unified Communications and Collaboration applications.
Voice Security Ensures a general security policy for Unified Communications and Collaboration
applications, and a hardened and secure networking foundation for them to rely upon.
Deployment Models Provide tested models in which to deploy Unified Communications and
Collaboration call control and applications, as well as best practices and design guidelines to apply
to Unified Communications and Collaboration deployments.
The chapters in this part of the SRND cover the networking subjects mentioned above. Each chapter
provides an introduction to the subject matter, followed by discussions surrounding architecture, high
availability, capacity planning, and design considerations. The chapters focus on design-related aspects
rather than product-specific support and configuration information, which is covered in the related
product documentation.
This part of the SRND includes the following chapters:
2-1
Chapter 2
2-2
Chapter 2
Architecture
The networking architecture lays the foundation upon which all other components of the Unified
Communications and Collaboration System are deployed. Figure 2-1 illustrates, in a generalized way,
the overall architecture of the Cisco Unified Communications and Collaboration System.
Figure 2-1
Remote
Office
Central Site
Monitoring and Scheduling
Applications
PSTN/ISDN
Unified CM
Media Resources
IP WAN
Cisco
E
Expressway-E
Conferencing Resources
Internet
348619
Cisco
Expressway-C
All aspects of the Unified Communications and Collaboration System, including call routing, call
control, applications and services, and operations and serviceability, rely heavily on proper design and
deployment of the system architecture.
High Availability
Proper design of the network infrastructure requires building a robust and redundant network from the
bottom up. By structuring the LAN as a layered model (access, distribution, and core layers) and
developing the LAN infrastructure one step of the model at a time, you can build a highly available, fault
tolerant, and redundant network. Proper WAN infrastructure design is also extremely important for
normal operation on a converged network. Proper infrastructure design requires following basic
configuration and design best-practices for deploying a WAN that is as highly available as possible and
that provides guaranteed throughput. Furthermore, proper WAN infrastructure design requires deploying
end-to-end QoS on all WAN links.
2-3
Chapter 2
Capacity Planning
Wireless LAN infrastructure design becomes important when IP telephony is added to the wireless LAN
(WLAN) portions of a converged network. With the addition of wireless Unified Communications and
Collaboration endpoints, voice and video traffic has moved onto the WLAN and is now converged with
the existing data traffic there. Just as with wired LAN and wired WAN infrastructures, the addition of
voice and video in the WLAN requires following basic configuration and design best-practices for
deploying a highly available network. In addition, proper WLAN infrastructure design requires
understanding and deploying QoS on the wireless network to ensure end-to-end voice and video quality
on the entire network.
After designing and implementing the network infrastructure properly, you can add network and
application services successfully across the network, thus providing a highly available foundation upon
which your Unified Communications and Collaboration services can run.
Capacity Planning
Scaling your network infrastructure to handle the Unified Communications and Collaboration
applications and services that it must support requires providing adequate available bandwidth and the
capability to handle the additional traffic load created by the applications.
For a complete discussion of system sizing, capacity planning, and deployment considerations related to
sizing, refer to the chapter on Collaboration Solution Sizing Guidance, page 25-1.
2-4
CH A P T E R
Network Infrastructure
Revised: June 15, 2015
This chapter describes the requirements of the network infrastructure needed to build a Cisco Unified
Communications System in an enterprise environment. Figure 3-1 illustrates the roles of the various
devices that form the network infrastructure, and Table 3-1 summarizes the features required to support
each of these roles.
Unified Communications places strict requirements on IP packet loss, packet delay, and delay variation
(or jitter). Therefore, it is important to enable most of the Quality of Service (QoS) mechanisms available
on Cisco switches and routers throughout the network. For the same reasons, redundant devices and
network links that provide quick convergence after network failures or topology changes are also
important to ensure a highly available infrastructure
The following sections describe the network infrastructure features as they relate to:
3-1
Chapter 3
Figure 3-1
Network Infrastructure
Central Site
IP
IP
IP
IP
IP
IP
IP
IP
IP
Campus access
layer
Campus distribution
layer
Campus core
layer
WAN aggregation
ISDN backup
Branch
router
PSTN
IP WAN
IP
IP
IP
IP
IP
IP
IP
IP
77290
Branch
switch
Branch offices
3-2
Chapter 3
Network Infrastructure
Whats New in This Chapter
Table 3-1
Infrastructure Role
Required Features
In-Line Power1
Traffic Classification
Traffic Reclassification
Traffic Shaping
Link Efficiency
Traffic Classification
Traffic Reclassification
Branch Router
(Spoke site)
LFI2
Link Efficiency
Traffic Classification
Traffic Reclassification
In-Line Power1
1. Recommended.
2. For link speeds less than 786 kbps.
Described in
Revision Date
3-3
Chapter 3
Network Infrastructure
LAN Infrastructure
LAN Infrastructure
Campus LAN infrastructure design is extremely important for proper Unified Communications
operation on a converged network. Proper LAN infrastructure design requires following basic
configuration and design best practices for deploying a highly available network. Further, proper LAN
infrastructure design requires deploying end-to-end QoS on the network. The following sections discuss
these requirements:
For more information on campus design, refer to the Design Zone for Campus at
http://www.cisco.com/go/designzone
3-4
Chapter 3
Network Infrastructure
LAN Infrastructure
Class C subnets (that is, a 23-bit subnet masked Class C address). For more information on the campus
access layer, refer to the documentation on available at
http://www.cisco.com/en/US/products/hw/switches/index.html.
Note
The recommendation to limit the number of devices in a single Unified Communications VLAN to
approximately 512 is not solely due to the need to control the amount of VLAN broadcast traffic.
Installing Unified CM in a VLAN with an IP subnet containing more than 1024 devices can cause the
Unified CM server ARP cache to fill up quickly, which can seriously affect communications between the
Unified CM server and other Unified Communications endpoints.
Figure 3-2
Distribution
Switches
Access
Switches
IP
VLAN=10
VVID=111
IP
VLAN=11
VVID=310
IP
VLAN=30
VVID=311
VVID=312
IP
IP
VLAN=31
VLAN=32
Stackable Switches
253921
VVID=110
When you deploy voice, Cisco recommends that you enable two VLANs at the access layer: a native
VLAN for data traffic (VLANs 10, 11, 30, 31, and 32 in Figure 3-2) and a voice VLAN under Cisco IOS
or Auxiliary VLAN under CatOS for voice traffic (represented by VVIDs 110, 111, 310, 311, and 312
in Figure 3-2).
Separate voice and data VLANs are recommended for the following reasons:
Address space conservation and voice device protection from external networks
Private addressing of phones on the voice or auxiliary VLAN ensures address conservation and
ensures that phones are not accessible directly through public networks. PCs and servers are
typically addressed with publicly routed subnet addresses; however, voice endpoints may be
addressed using RFC 1918 private subnet addresses.
3-5
Chapter 3
Network Infrastructure
LAN Infrastructure
To provide high-quality voice and to take advantage of the full voice feature set, access layer switches
should provide support for:
802.1Q trunking and 802.1p for proper treatment of Layer 2 CoS packet marking on ports with
phones connected
Multiple egress queues to provide priority queuing of RTP voice packet streams
The ability to classify or reclassify traffic and establish a network trust boundary
Inline power capability (Although inline power capability is not mandatory, it is highly
recommended for the access layer switches.)
Layer 3 awareness and the ability to implement QoS access control lists (These features are
recommended if you are using certain Unified Communications endpoints such as a PC running a
softphone application like Jabber that cannot benefit from an extended trust boundary.)
PortFast
Enable PortFast on all access ports. The phones, PCs, or servers connected to these ports do not
forward bridge protocol data units (BPDUs) that could affect STP operation. PortFast ensures that
the phone or PC, when connected to the port, is able to begin receiving and transmitting traffic
immediately without having to wait for STP to converge.
3-6
Chapter 3
Network Infrastructure
LAN Infrastructure
Note
With the introduction of RSTP 802.1w, features such as PortFast and UplinkFast are not required
because these mechanisms are built in to this standard. If RSTP has been enabled on the Catalyst switch,
these commands are not necessary.
Core
Layer 3
Distribution
HSRP Active
Root Bridge
HSRP
Standby
Layer 2
VLAN 2 Voice
VLAN 102 Data
VLAN 3 Voice
VLAN 103 Data
VLAN n Voice
VLAN 100 + n Data
271569
Access
The purpose of the distribution switch in this design is to provide boundary functions between the
bridged Layer 2 portion of the campus and the routed Layer 3 portion, including support for the default
gateway, Layer 3 policy control, and all the multicast services required.
An alternative configuration to the traditional distribution layer model illustrated in Figure 3-3 is one in
which the access switch acts as a full Layer 3 routing node (providing both Layer 2 and Layer 3
switching) and the access-to-distribution Layer 2 uplink trunks are replaced with Layer 3 point-to-point
routed links. This alternative configuration, in which the Layer 2/3 demarcation is moved from the
distribution switch to the access switch (as shown in Figure 3-4), appears to be a major change to the
design but is actually just an extension of the current best-practice design.
3-7
Chapter 3
Network Infrastructure
LAN Infrastructure
Figure 3-4
Core
Layer 3
Distribution
Access
VLAN 3 Voice
VLAN n Voice
VLAN 103 Data VLAN 100 + n Data
271570
VLAN 2 Voice
VLAN 102 Data
Layer 2
In both the traditional Layer 2 and the Layer 3 routed access designs, each access switch is configured
with unique voice and data VLANs. In the Layer 3 design, the default gateway and root bridge for these
VLANs is simply moved from the distribution switch to the access switch. Addressing for all end
stations and for the default gateway remains the same. VLAN and specific port configurations remain
unchanged on the access switch. Router interface configuration, access lists, "ip helper," and any other
configuration for each VLAN remain identical but are configured on the VLAN Switched Virtual
Interface (SVI) defined on the access switch instead of on the distribution switches.
There are several notable configuration changes associated with the move of the Layer 3 interface down
to the access switch. It is no longer necessary to configure a Hot Standby Router Protocol (HSRP) or
Gateway Load Balancing Protocol (GLBP) virtual gateway address as the "router" interfaces because all
the VLANs are now local. Similarly, with a single multicast router, for each VLAN it is not necessary
to perform any of the traditional multicast tuning such as tuning PIM query intervals or ensuring that the
designated router is synchronized with the active HSRP gateway.
Improved convergence
Of these advantages, perhaps the most significant is the improvement in network convergence times
possible when using a routed access design configured with Enhanced Interior Gateway Routing
Protocol (EIGRP) or Open Shortest Path First (OSPF) as the routing protocol. Comparing the
convergence times for an optimal Layer 2 access design (either with a spanning tree loop or without a
loop) against that of the Layer 3 access design, you can obtain a four-fold improvement in convergence
times, from 800 to 900 msec for the Layer 2 design to less than 200 msec for the Layer 3 access design.
3-8
Chapter 3
Network Infrastructure
LAN Infrastructure
For more information on routed access designs, refer to the document on High Availability Campus
Network Design Routed Access Layer using EIGRP or OSPF, available at
http://www.cisco.com/application/pdf/en/us/guest/netsol/ns432/c649/ccmigration_09186a0080811
468.pdf
Like HSRP and VRRP, Cisco's Gateway Load Balancing Protocol (GLBP) protects data traffic from a
failed router or circuit, while also allowing packet load sharing between a group of redundant routers.
When HSRP or VRRP are used to provide default gateway redundancy, the backup members of the peer
relationship are idle, waiting for a failure event to occur for them to take over and actively forward
traffic.
Before the development of GLBP, methods to utilize uplinks more efficiently were difficult to implement
and manage. In one technique, the HSRP and STP/RSTP root alternated between distribution node peers,
with the even VLANs homed on one peer and the odd VLANs homed on the alternate. Another technique
used multiple HSRP groups on a single interface and used DHCP to alternate between the multiple
default gateways. These techniques worked but were not optimal from a configuration, maintenance, or
management perspective.
GLBP is configured and functions like HSRP. For HSRP, a single virtual MAC address is given to the
endpoints when they use Address Resolution Protocol (ARP) to learn the physical MAC address of their
default gateways (see Figure 3-5).
3-9
Chapter 3
Network Infrastructure
LAN Infrastructure
Figure 3-5
vIP
10.88.1.10
HSRP 1 ip 10.88.1.10
vMAC 0000.0000.0001
HSRP 1 ip 10.88.1.10
vMAC 0000.0000.0001
ARP
reply
.1
.2
10.88.1.0/24
.4
A
253919
.5
Two virtual MAC addresses exist with GLBP, one for each GLBP peer (see Figure 3-6). When an
endpoint uses ARP to determine its default gateway, the virtual MAC addresses are checked in a
round-robin basis. Failover and convergence work just like with HSRP. The backup peer assumes the
virtual MAC address of the device that has failed, and begins forwarding traffic for its failed peer.
Figure 3-6
GLBP Uses Two Virtual MAC Addresses, One for Each GLBP Peer
vIP
10.88.1.10
GLBP 1 ip 10.88.1.10
vMAC 0000.0000.0001
GLBP 1 ip 10.88.1.10
vMAC 0000.0000.0002
ARP
reply
.1
.2
10.88.1.0/24
.5
B
253920
.4
The end result is that a more equal utilization of the uplinks is achieved with minimal configuration. As
a side effect, a convergence event on the uplink or on the primary distribution node affects only half as
many hosts, giving a convergence event an average of 50 percent less impact.
For more information on HSRP, VRRP, and GLBP, refer to the Campus Network for High Availability
Design Guide, available at
http://www.cisco.com/application/pdf/en/us/guest/netsol/ns431/c649/ccmigration_09186a008093b
876.pdf
3-10
Chapter 3
Network Infrastructure
LAN Infrastructure
Routing Protocols
Configure Layer 3 routing protocols such as OSPF and EIGRP at the distribution layer to ensure fast
convergence, load balancing, and fault tolerance. Use parameters such as routing protocol timers, path
or link costs, and address summaries to optimize and control convergence times as well as to distribute
traffic across multiple paths and devices. Cisco also recommends using the passive-interface command
to prevent routing neighbor adjacencies via the access layer. These adjacencies are typically
unnecessary, and they create extra CPU overhead and increased memory utilization because the routing
protocol keeps track of them. By using the passive-interface command on all interfaces facing the
access layer, you prevent routing updates from being sent out on these interfaces and, therefore, neighbor
adjacencies are not formed.
Redundant devices
Redundancy here ensures that, in the event of a device failure, another device in the network can
continue performing tasks that the failed device was doing.
The Cisco Catalyst switches with Virtual Switching System (VSS) is a method to ensure redundancy in
all of these areas by pooling together two Catalyst supervisor engines to act as one. For more information
regarding VSS, refer to the product documentation available at
http://www.cisco.com/en/US/products/ps9336/index.html
Routing protocols at the core layer should again be configured and optimized for path redundancy and
fast convergence. There should be no STP in the core because network connectivity should be routed at
Layer 3. Finally, each link between the core and distribution devices should belong to its own VLAN or
subnet and be configured using a 30-bit subnet mask.
Data Center and Server Farm
Typically, Cisco Unified Communications Manager (Unified CM) cluster servers, including media
resource servers, reside in a firewall-secured data center or server farm environment. In addition,
centralized gateways and centralized hardware media resources such as conference bridges, DSP or
transcoder farms, and media termination points may be located in the data center or server farm. The
placement of firewalls in relation to Cisco Unified Communications Manager (Unified CM) cluster
servers and media resources can affect how you design and implement security in your network. For
design guidance on firewall placement in relation to Unified Communications systems and media
resources, see Firewalls, page 4-22.
3-11
Chapter 3
Network Infrastructure
LAN Infrastructure
Because these servers and resources are critical to voice networks, Cisco recommends distributing all
Unified CM cluster servers, centralized voice gateways, and centralized hardware resources between
multiple physical switches and, if possible, multiple physical locations within the campus. This
distribution of resources ensures that, given a hardware failure (such as a switch or switch line card
failure), at least some servers in the cluster will still be available to provide telephony services. In
addition, some gateways and hardware resources will still be available to provide access to the PSTN
and to provide auxiliary services. Besides being physically distributed, these servers, gateways, and
hardware resources should be distributed among separate VLANs or subnets so that, if a broadcast storm
or denial of service attack occurs on a particular VLAN, not all voice connectivity and services will be
disrupted.
Caution
The use of power injectors or power patch panels to deliver PoE can damage some devices because power
is always applied to the Ethernet pairs. PoE switch ports automatically detect the presence of a device
that requires PoE before enabling it on a port-by-port basis.
In addition to Cisco PoE inline power, Cisco now supports the IEEE 802.3af PoE and the IEEE 802.3at
Enhanced PoE standards. For information on which Cisco Unified IP Phones support the 802.3af and
802.3at standards, refer to the product documentation for your particular phone models.
3-12
Chapter 3
Network Infrastructure
LAN Infrastructure
The IP phone can be restarted only when the switch powers it on again. After power is restored, the IP
phones will reboot and undergo a recovery process that includes requesting a new IP address,
downloading a configuration file, applying any new configuration parameters, downloading new
firmware or locales, and registering with Cisco Unified CM.
The EnergyWise schedule is configured and managed on the Cisco Network Infrastructure. It does not
require any configuration on the IP phone or on Cisco Unified CM. However, power consumption on the
phone can also be managed by a device profile configured on Unified CM. The energy saving options
provided by Unified CM include the following:
Note
The Cisco EnergyWise Power Save Plus mode is supported on most Cisco IP Phones and Collaboration
Desk Endpoints. To learn which endpoints support EnergyWise Power Save Plus, refer to the data sheets
for your endpoint models:
http://www.cisco.com/c/en/us/products/collaboration-endpoints/product-listing.html
3-13
Chapter 3
Network Infrastructure
LAN Infrastructure
For information on configuring these modes, refer to the administration guides for the Cisco IP Phones
and Collaboration Desk Endpoints:
http://www.cisco.com/c/en/us/products/collaboration-endpoints/product-listing.html
Core
Si
Si
Typical 4:1
Data Oversubscription
Instantaneous
Interface
Congestion
Distribution
Si
Si
Typical 20:1
Data Oversubscription
Access
IP
IP
IP
IP
IP
IP
IP
IP
IP
IP
IP
114469
IP
Voice
Data
This oversubscription, coupled with individual traffic volumes and the cumulative effects of multiple
independent traffic sources, can result in the egress interface buffers becoming full instantaneously, thus
causing additional packets to drop when they attempt to enter the egress buffer. The fact that campus
switches use hardware-based buffers, which compared to the interface speed are much smaller than those
found on WAN interfaces in routers, merely increases the potential for even short-lived traffic bursts to
cause buffer overflow and dropped packets.
Applications such as file sharing (both peer-to-peer and server-based), remote networked storage,
network-based backup software, and emails with large attachments, can create conditions where network
congestion occurs more frequently and/or for longer durations. Some of the negative effects of recent
worm attacks have been an overwhelming volume of network traffic (both unicast and broadcast-storm
based), increasing network congestion. If no buffer management policy is in place, loss, delay, and jitter
performance of the LAN may be affected for all traffic.
3-14
Chapter 3
Network Infrastructure
LAN Infrastructure
Another situation to consider is the effect of failures of redundant network elements, which cause
topology changes. For example, if a distribution switch fails, all traffic flows will be reestablished
through the remaining distribution switch. Prior to the failure, the load balancing design shared the load
between two switches, but after the failure all flows are concentrated in a single switch, potentially
causing egress buffer conditions that normally would not be present.
For applications such as voice, this packet loss and delay results in severe voice quality degradation.
Therefore, QoS tools are required to manage these buffers and to minimize packet loss, delay, and delay
variation (jitter).
The following types of QoS tools are needed end-to-end on the network to manage traffic and ensure
voice and video quality:
Traffic classification
Classification involves the marking of packets with a specific priority denoting a requirement for
class of service (CoS) from the network. The point at which these packet markings are trusted or not
trusted is considered the trust boundary. Trust is typically extended to voice devices (phones) and
not to data devices (PCs).
Queuing or scheduling
Interface queuing or scheduling involves assigning packets to one of several queues based on
classification for expedited treatment throughout the network.
Bandwidth provisioning
Provisioning involves accurately calculating the required bandwidth for all applications plus
element overhead.
The following sections discuss the use of these QoS mechanisms in a campus environment:
Traffic Classification
It has always been an integral part of the Cisco network design architecture to classify or mark traffic as
close to the edge of the network as possible. Traffic classification is an entrance criterion for access into
the various queuing schemes used within the campus switches and WAN interfaces. Cisco IP Phones
mark voice control signaling and voice RTP streams at the source, and they adhere to the values
presented in Table 3-3. As such, the IP phone can and should classify traffic flows.
Table 3-3 lists the traffic classification requirements for the LAN infrastructure.
3-15
Chapter 3
Network Infrastructure
LAN Infrastructure
Table 3-3
Layer-3 Classification
Layer-2 Classification
Application
Differentiated Services
Per-Hop Behavior (PHB) Code Point (DSCP)
Routing
CS6
48
Voice Real-Time
Transport Protocol
(RTP)
EF
46
Videoconferencing
AF41
34
IP video
AF41
34
Immersive video
CS4
32
Streaming video
AF31
26
Call signaling
CS3
24
Transactional data
AF21
18
Network management
CS2
16
Scavenger
CS1
Best effort
Real-Time Interactive
For more information about traffic classification, refer to the QoS design guides available at
http://www.cisco.com/c/en/us/solutions/enterprise/design-zone-ipv6/design-guide-listing.html
Traffic Classification for Video Telephony
Voice
Voice is classified as CoS 5 (IP Precedence 5, PHB EF, or DSCP 46).
Videoconferencing
Videoconferencing is classified as CoS 4 (IP Precedence 4, PHB AF41, or DSCP 34).
Call signaling
Call signaling for voice and videoconferencing is classified as CoS 3 (IP Precedence 3, PHB CS3,
or DSCP 24).
Cisco highly recommends these classifications as best practices in a Cisco Unified Communications
network.
QoS Marking Differences Between Video Calls and Voice-Only Calls
The voice component of a call can be classified in one of two ways, depending on the type of call in
progress. A voice-only telephone call would have its media classified as CoS 5 (IP Precedence 5 or
PHB EF), while the voice channel of a video conference would have its media classified as CoS 4
(IP Precedence 4 or PHB AF41). All the Cisco IP Video Telephony products adhere to the Cisco
3-16
Chapter 3
Network Infrastructure
LAN Infrastructure
Corporate QoS Baseline standard, which requires that the audio and video channels of a video call both
be marked as CoS 4 (IP Precedence 4 or PHB AF41). The reasons for this recommendation include, but
are not limited to, the following:
Cisco is in the process of changing this requirement for endpoints to mark the audio and video channels
of a video call separately, thus providing the flexibility to mark both the audio and video channels of a
video call with the same DSCP value or different DSCP values, depending on the use cases. For more
information on DSCP marking, see the chapter on Bandwidth Management, page 13-1.
The signaling class is applicable to all voice signaling protocols (such as SCCP, MGCP, and so on) as
well as video signaling protocols (such as SCCP, H.225, RAS, CAST, and so on).
Given the recommended classes, the first step is to decide where the packets will be classified (that is,
which device will be the first to mark the traffic with its QoS classification). There are essentially two
places to mark or classify traffic:
On the originating endpoint the classification is then trusted by the upstream switches and routers
On the switches and/or routers because the endpoint is either not capable of classifying its own
packets or is not trustworthy to classify them correctly
A Trusted Relay Point (TRP) can be used to enforce and/or re-mark the DSCP values of media flows
from endpoints. This feature allows QoS to be enforced for media from endpoints such as softphones,
where the media QoS values might have been modified locally.
A TRP is a media resource based upon the existing Cisco IOS media termination point (MTP) function.
Endpoints can be configured to "Use Trusted Relay Point," which will invoke a TRP for all calls.
For QoS enforcement, the TRP uses the configured QoS values for media in Unified CM's Service
Parameters to re-mark and enforce the QoS values in media streams from the endpoint.
TRP functionality is supported by Cisco IOS MTPs and transcoding resources. (Use Unified CM to
check "Enable TRP" on the MTP or transcoding resource to activate TRP functionality.)
Interface Queuing
After packets have been marked with the appropriate tag at Layer 2 (CoS) and Layer 3 (DSCP or PHB),
it is important to configure the network to schedule or queue traffic based on this classification, so as to
provide each class of traffic with the service it needs from the network. By enabling QoS on campus
switches, you can configure all voice traffic to use separate queues, thus virtually eliminating the
possibility of dropped voice packets when an interface buffer fills instantaneously.
Although network management tools may show that the campus network is not congested, QoS tools are
still required to guarantee voice quality. Network management tools show only the average congestion
over a sample time span. While useful, this average does not show the congestion peaks on a campus
interface.
Transmit interface buffers within a campus tend to congest in small, finite intervals as a result of the
bursty nature of network traffic. When this congestion occurs, any packets destined for that transmit
interface are dropped. The only way to prevent dropped voice traffic is to configure multiple queues on
campus switches. For this reason, Cisco recommends always using a switch that has at least two output
queues on each port and the ability to send packets to these queues based on QoS Layer 2 and/or Layer 3
3-17
Chapter 3
Network Infrastructure
LAN Infrastructure
classification. The majority of Cisco Catalyst Switches support two or more output queues per port. For
more information on Cisco Catalyst Switch interface queuing capabilities, refer to the documentation at
http://www.cisco.com/en/US/products/hw/switches/index.html
Bandwidth Provisioning
In the campus LAN, bandwidth provisioning recommendations can be summarized by the motto, Over
provision and under subscribe. This motto implies careful planning of the LAN infrastructure so that the
available bandwidth is always considerably higher than the load and there is no steady-state congestion
over the LAN links.
The addition of voice traffic onto a converged network does not represent a significant increase in overall
network traffic load; the bandwidth provisioning is still driven by the demands of the data traffic
requirements. The design goal is to avoid extensive data traffic congestion on any link that will be
traversed by telephony signaling or media flows. Contrasting the bandwidth requirements of a single
G.711 voice call (approximately 86 kbps) to the raw bandwidth of a FastEthernet link (100 Mbps)
indicates that voice is not a source of traffic that causes network congestion in the LAN, but rather it is
a traffic flow to be protected from LAN network congestion.
3-18
Chapter 3
Network Infrastructure
LAN Infrastructure
QoS Design Considerations for Virtual Unified Communications with Cisco UCS
Servers
Unified Communications applications such as Cisco Unified Communications Manager (Unified CM)
run as virtual machines on top of the VMware Hypervisor. These Unified Communications virtual
machines are connected to a virtual software switch rather than a hardware-based Ethernet. The
following types of virtual software switches are available:
From the point of view of virtual connectivity, each virtual machine can connect to any one of the above
virtual switches residing on a blade server. When using Cisco UCS B-Series blade servers, the blade
servers physically connect to the rest of the network through a Fabric Extender in the UCS chassis to a
UCS Fabric Interconnect Switch (for example, Cisco UCS 6100 or 6200 Series). The UCS Fabric
Interconnect Switch is where the physical wiring connects to a customer's Ethernet LAN and FC SAN.
From the point of view of traffic flow, traffic from the virtual machines first goes to the software virtual
switch (for example, vSphere Standard Switch, vSphere Distributed Switch, or Cisco Nexus 1000V
Switch). The virtual switch then sends the traffic to the physical UCS Fabric Interconnect Switch
through its blade server's Network Adapter and Fabric Extender. The UCS Fabric Interconnect Switch
carries both the IP and fibre channel SAN traffic via Fibre Channel over Ethernet (FCoE) on a single
wire. The UCS Fabric Interconnect Switch sends IP traffic to an IP switch (for example, Cisco Catalyst
or Nexus Series Switch), and it sends SAN traffic to a Fibre Channel SAN Switch (for example, Cisco
MDS Series Switch).
Congestion Scenario
In a deployment with Cisco UCS B-Series blades servers and with Cisco Collaboration applications only,
network congestion or an oversubscription scenario is unlikely because the UCS Fabric Interconnect
Switch provides a high-capacity switching fabric, and the usable bandwidth per server blade far exceeds
the maximum traffic requirements of a typical Collaboration application.
However, there might be scenarios where congestion could arise. For example, with a large number of
B-Series blade servers and chassis, a large number of applications, and/or third-party applications
requiring high network bandwidth, there is a potential for congestion on the different network elements
of the UCS B-Series system (adapters, IO modules, Fabric Interconnects). In addition, FCoE traffic is
sharing the same network elements as IP traffic, therefore applications performing a high amount of
storage transfer would increase the utilization on the network elements and contribute to this potential
congestion.
To address this potential congestion, QoS should be implemented.
3-19
Chapter 3
Network Infrastructure
LAN Infrastructure
Note
Fibre Channel over Ethernet (FCoE) traffic has a reserved QoS system class that should not be used by
any other type of traffic. By default, this system class has a CoS value of 3, which is the same value
assigned to the system class used by voice and video signaling traffic in the example above. To prevent
voice and video signaling traffic from using the FCoE system class, assign a different CoS value to the
FCoE system class (2 or 4, for instance).
Note
The decision to use the Nexus 1000V will vary on a case-by-case basis, depending on the available
bandwidth for Unified Communications applications within the UCS architecture. If there is a possibility
that a congestion scenario will arise, then the Nexus 1000V switch should be deployed.
If the Nexus 1000V is not deployed, it is still possible to provide some QoS, but it would not be an
optimal solution. For example, you could create multiple virtual switches and assign a different CoS
value for the uplink ports of each of those switches. For example, virtual switch 1 would have uplink
ports configured with a CoS value of 1, virtual switch 2 would have uplink ports configured with a CoS
value of 2, and so forth. Then the application virtual machines would be assigned to a virtual switch,
depending on the desired QoS system class. The downside to this approach is that all traffic types from
a virtual machine will have the same CoS value. For example, with a Unified CM virtual machine,
real-time media traffic such as MoH traffic, signaling traffic, and non-voice traffic (for example,
backups, CDRs, logs, Web traffic, and so forth) would share the same CoS value.
3-20
Chapter 3
Network Infrastructure
LAN Infrastructure
Campus
Immersive
video unit
Immersive
video unit
CS4
Video
IP phone
Video Portal
Client
Streaming
Server
Encoder
AF41
AF31
AF31
Video
IP phone
AF31
CS3
IP
IP phone
Web Server/
Content
Repository
EF
IP phone
Pre-recorded Content
Live Content
292585
IP
Unlike voice, real-time IP video traffic in general is a somewhat bursty, variable bit rate stream.
Therefore video, unlike voice, does not have clear formulas for calculating network overhead because
video packet sizes and rates vary proportionally to the degree of motion within the video image itself.
From a network administrator's point of view, bandwidth is always provisioned at Layer 2, but the
variability in the packet sizes and the variety of Layer 2 media that the packets may traverse from
end-to-end make it difficult to calculate the real bandwidth that should be provisioned at Layer 2.
However, the conservative rule that has been thoroughly tested and widely used is to over-provision
video bandwidth by 20%. This accommodates the 10% burst and the network overhead from Layer 2 to
Layer 4.
3-21
Chapter 3
Network Infrastructure
LAN Infrastructure
Network Services
The deployment of an IP Communications system requires the coordinated design of a well structured,
highly available, and resilient network infrastructure as well as an integrated set of network services
including Domain Name System (DNS), Dynamic Host Configuration Protocol (DHCP), Trivial File
Transfer Protocol (TFTP), and Network Time Protocol (NTP).
Resolve fully qualified domain names to IP addresses for SIP route patterns based on domain name
Resolve service (SRV) records to host names and then to IP addresses for SIP trunk destinations
3-22
Chapter 3
Network Infrastructure
LAN Infrastructure
When DNS is used, Cisco recommends defining each Unified CM cluster as a member of a valid
sub-domain within the larger organizational DNS domain, defining the DNS domain on each Cisco
Unified CM server, and defining the primary and secondary DNS server addresses on each Unified CM
server.
Table 3-4 shows an example of how DNS server could use A records (Hostname-to-IP-address
resolution), Cname records (aliases), and SRV records (service records for redundancy, load balancing,
and service discovery) in a Unified CM environment.
Table 3-4
Host Name
Type
TTL
Data
12 Hours
182.10.10.1
CUCM1.cluster1.cisco.com
Host (A)
Default
182.10.10.1
CUCM2.cluster1.cisco.com
Host (A)
Default
182.10.10.2
CUCM3.cluster1.cisco.com
Host (A)
Default
182.10.10.3
CUCM4.cluster1.cisco.com
Host (A)
Default
182.10.10.4
TFTP-server1.cluster1.cisco.com
Host (A)
12 Hours
182.10.10.11
TFTP-server2.cluster1.cisco.com
Host (A)
12 Hours
182.10.10.12
CUP1.cluster1.cisco.com
Host (A)
Default
182.10.10.15
CUP2.cluster1.cisco.com
Host (A)
Default
182.10.10.16
www.CUCM-Admin.cisco.com
Alias (CNAME)
Default
CUCM-Admin.cluster1.cisco.com
_sip._tcp.cluster1.cisco.com.
Service (SRV)
Default
CUCM1.cluster1.cisco.com
_sip._tcp.cluster1.cisco.com.
Service (SRV)
Default
CUCM2.cluster1.cisco.com
_sip._tcp.cluster1.cisco.com.
Service (SRV)
Default
CUCM3.cluster1.cisco.com
_sip._tcp.cluster1.cisco.com.
Service (SRV)
Default
CUCM4.cluster1.cisco.com
For Jabber clients, refer to the Cisco Jabber DNS Configuration Guide, available at
http://www.cisco.com/web/products/voice/jabber.html
3-23
Chapter 3
Network Infrastructure
LAN Infrastructure
network, all you have to do is add DHCP voice scopes to an existing DHCP server for these new voice
devices. Because IP telephony devices are configured to use and rely on a DHCP server for IP
configuration information, you must deploy DHCP servers in a redundant fashion. At least two DHCP
servers should be deployed within the telephony network such that, if one of the servers fails, the other
can continue to answer DHCP client requests. You should also ensure that DHCP server(s) are
configured with enough IP subnet addresses to handle all DHCP-reliant clients within the network.
Note
If the primary TFTP server is available but is not able to grant the requested file to the phone (for
example, because the requesting phone is not configured on that cluster), the phone will not attempt to
contact the secondary TFTP server.
Cisco highly recommends using a direct IP address (that is, not relying on a DNS service) for Option 150
because doing so eliminates dependencies on DNS service availability during the phone boot-up and
registration process.
Note
Even though IP phones support a maximum of two TFTP servers under Option 150, you could configure
a Unified CM cluster with more than two TFTP servers. For instance, if a Unified CM system is
clustered over a WAN at three separate sites, three TFTP servers could be deployed (one at each site).
Phones within each site could then be granted a DHCP scope containing that site's TFTP server within
Option 150. This configuration would bring the TFTP service closer to the endpoints, thus reducing
latency and ensuring failure isolation between the sites (one site's failure would not affect TFTP service
at another site).
3-24
Chapter 3
Network Infrastructure
LAN Infrastructure
prevent depletion of DHCP-managed subnet addresses. Mobile devices typically use IP addresses for
short increments of time and then might not request a DHCP renewal or new address for a long period
of time. Longer lease times will tie up these IP addresses and prevent them from being reassigned even
when they are no longer being used.
Cisco Unified IP Phones adhere to the conditions of the DHCP lease duration as specified in the DHCP
server's scope configuration. Once half the lease time has expired since the last successful DHCP server
acknowledgment, the IP phone will request a lease renewal. This DHCP client Request, once
acknowledged by the DHCP server, will allow the IP phone to retain use of the IP scope (that is, the IP
address, default gateway, subnet mask, DNS server (optional), and TFTP server (optional)) for another
lease period. If the DHCP server becomes unavailable, an IP phone will not be able to renew its DHCP
lease, and as soon as the lease expires, it will relinquish its IP configuration and will thus become
unregistered from Unified CM until a DHCP server can grant it another valid scope.
In centralized call processing deployments, if a remote site is configured to use a centralized DHCP
server (through the use of a DHCP relay agent such as the IP Helper Address in Cisco IOS) and if
connectivity to the central site is severed, IP phones within the branch will not be able to renew their
DHCP scope leases. In this situation, branch IP phones are at risk of seeing their DHCP lease expire,
thus losing the use of their IP address, which would lead to service interruption. Given the fact that
phones attempt to renew their leases at half the lease time, DHCP lease expiration can occur as soon as
half the lease time since the DHCP server became unreachable. For example, if the lease time of a DHCP
scope is set to 4 days and a WAN failure causes the DHCP server to be unavailable to the phones in a
branch, those phones will be unable to renew their leases at half the lease time (in this case, 2 days). The
IP phones could stop functioning as early as 2 days after the WAN failure, unless the WAN comes back
up and the DHCP server is available before that time. If the WAN connectivity failure persists, all phones
see their DHCP scope expire after a maximum of 4 days from the WAN failure.
This situation can be mitigated by one of the following methods:
Set the DHCP scope lease to a long duration (for example, 8 days or more).
This method would give the system administrator a minimum of half the lease time to remedy any
DHCP reachability problem. Long lease durations also have the effect of reducing the frequency of
network traffic associated with lease renewals.
Configure co-located DHCP server functionality (for example, run a DHCP server function on the
branch's Cisco IOS router).
This approach is immune to WAN connectivity interruption. One effect of such an approach is to
decentralize the management of IP addresses, requiring incremental configuration efforts in each
branch. (See DHCP Network Deployments, page 3-25, for more information.)
Note
The term co-located refers to two or more devices in the same physical location, with no
WAN or MAN connection between them.
3-25
Chapter 3
Network Infrastructure
LAN Infrastructure
deployed at the central site, both servers' IP addresses must be configured as ip helper-address.
Also note that, if branch-side telephony devices rely on a centralized DHCP server and the WAN
link between the two sites fails, devices at the branch site will be unable to send DHCP requests or
receive DHCP responses.
Note
By default, service dhcp is enabled on the Cisco IOS device and does not appear in the
configuration. Do not disable this service on the branch router because doing so will disable
the DHCP relay agent on the device, and the ip helper-address configuration command will
not work.
Centralized DHCP Server and Remote Site Cisco IOS DHCP Server
When configuring DHCP for use in a centralized multisite Unified CM deployment, you can use a
centralized DHCP server to provide DHCP service to centrally located devices. Remote devices
could receive DHCP service from a locally installed server or from the Cisco IOS router at the
remote site. This type of deployment ensures that DHCP services are available to remote telephony
devices even during WAN failures. Example 3-1 lists the basic Cisco IOS DHCP server
configuration commands.
Example 3-1
Note
The term co-resident refers to two or more services or applications running on the same server or virtual
machine.
3-26
Chapter 3
Network Infrastructure
LAN Infrastructure
Ringer files
Softkey files
The TFTP server manages and serves two types of files, those that are not modifiable (for example,
firmware files for phones) and those that can be modified (for example, configuration files).
A typical configuration file contains a prioritized list of Unified CMs for a device (for example, an SCCP
or SIP phone), the TCP ports on which the device connects to those Unified CMs, and an executable load
identifier. Configuration files for selected devices contain locale information and URLs for the
messages, directories, services, and information buttons on the phone.
When a device's configuration changes, the TFTP server rebuilds the configuration files by pulling the
relevant information from the Unified CM database. The new file(s) is then downloaded to the phone
once the phone has been reset. As an example, if a single phone's configuration file is modified (for
example, during Extension Mobility login or logout), only that file is rebuilt and downloaded to the
phone. However, if the configuration details of a device pool are changed (for example, if the primary
Unified CM server is changed), then all devices in that device pool need to have their configuration files
rebuilt and downloaded. For device pools that contain large numbers of devices, this file rebuilding
process can impact server performance.
Note
The TFTP server can perform a local database read from the database on its co-resident subscriber
server. Local database read not only provides benefits such as the preservation of user-facing features
when the publisher in unavailable, but also allows multiple TFTP servers to be distributed by means of
clustering over the WAN. (The same latency rules for clustering over the WAN apply to TFTP servers as
apply to servers with registered phones.) This configuration brings the TFTP service closer to the
endpoints, thus reducing latency and ensuring failure isolation between the sites.
When a device requests a configuration file from the TFTP server, the TFTP server searches for the
configuration file in its internal caches, the disk, and then alternate Cisco file servers (if specified). If
the TFTP server finds the configuration file, it sends it to the device. If the configuration file provides
Unified CM names, the device resolves the name by using DNS and opens a connection to the
Unified CM. If the device does not receive an IP address or name, it uses the TFTP server name or IP
address to attempt a registration connection. If the TFTP server cannot find the configuration file, it
sends a "file not found" message to the device.
3-27
Chapter 3
Network Infrastructure
LAN Infrastructure
A device that requests a configuration file while the TFTP server is rebuilding configuration files or
while it is processing the maximum number of requests, will receive a message from the TFTP server
that causes the device to request the configuration file later. The Maximum Serving Count service
parameter, which can be configured, specifies the maximum number of requests that can be concurrently
handled by the TFTP server. (Default value = 500 requests.) Use the default value if the TFTP service is
run along with other Cisco CallManager services on the same server. For a dedicated TFTP server, use
the following suggested values for the Maximum Serving Count: 1500 for a single-processor system or
3000 for a dual-processor system.
The Cisco Unified IP Phones 8900 Series and 9900 Series request their TFTP configuration files over
the HTTP protocol (port 6970), which is much faster than TFTP.
3-28
Chapter 3
Network Infrastructure
LAN Infrastructure
Figure 3-9
Data Packet
Acknowledgement
Time = 20ms
Acknowledgement
Time = 30ms
Data Packet
191938
Data Packet
If a response is not received in the timeout period (4 seconds by default), the sender will resend the data
packet or acknowledgment. When a packet has been sent five times without a response, the TFTP session
fails. Because the timeout period is always the same and not adaptive like a TCP timeout, packet loss
can significantly increase the amount of time a transfer session takes to complete.
Because the delay between each data packet is, at a minimum, equal to the network round-trip time,
network latency also is a factor in the maximum throughput that a TFTP session can achieve.
In Figure 3-10, the round-trip time has been increased to 40 ms and one packet has been lost in transit.
While the error rate is high at 12%, it is easy to see the effect of latency and packet loss on TFTP because
the time to complete the session increased from 30 ms (in Figure 3-9) to 4160 ms (in Figure 3-10).
Figure 3-10
Data Packet
Acknowledgement
Time = 80ms
Data Packet
Acknowledgement
Acknowledgement
Time = 4 sec + 120ms = 4120ms
Data Packet
191939
Use the following formula to calculate how long a TFTP file transfer will take to complete:
FileTransferTime = FileSize [(RTT + ERR Timeout) / 512000]
Where:
FileTransferTime is in seconds.
FileSize is in bytes.
RTT is the round-trip time in milliseconds.
ERR is the error rate, or percentage of packets that are lost.
Timeout is in milliseconds.
3-29
Chapter 3
Network Infrastructure
LAN Infrastructure
Note
The exact process each phone goes through on startup and the size of the files downloaded will depend
on the phone model, the signaling type configured for the phone (SCCP, MGCP, or SIP) and the previous
state of the phone. While there are differences in which files are requested, the general process each
phone follows is the same, and in all cases the TFTP server is used to request and deliver the appropriate
files. The general recommendations for TFTP server deployment do not change based on the protocol
and/or phone models deployed.
Under normal operations, a phone in subnet 10.1.1.0/24 will request TFTP services from
TFTP1_Primary, while a phone in subnet 10.1.2.0/24 will request TFTP services from
TFTP1_Secondary. If TFTP1_Primary fails, then phones from both subnets will request TFTP services
from TFTP1_Secondary.
Load balancing avoids having a single TFTP server hot-spot, where all phones from multiple clusters
rely on the same server for service. TFTP load balancing is especially important when phone software
loads are transferred, such as during a Unified CM upgrade, because more files of larger size are being
transferred, thus imposing a bigger load on the TFTP server.
3-30
Chapter 3
Network Infrastructure
LAN Infrastructure
C2_TFTP_Primary
C3_TFTP_Primary
Proxy requests to
alternate TFTP hosts
M
M
C1_TFTP_Primary
IP
C1_TFTP_Secondary
153371
3-31
Chapter 3
Network Infrastructure
LAN Infrastructure
Note
Cisco does not recommend enabling auto-registration on the centralized TFTP node cluster. If
auto-registration is enabled on the centralized TFTP cluster and any of the alternate cluster TFTP nodes
are down, phones provisioned on those alternate TFTP clusters will get auto-registered to the centralized
TFTP cluster rather than registering to their home cluster.
3-32
Chapter 3
Network Infrastructure
WAN Infrastructure
WAN Infrastructure
Proper WAN infrastructure design is also extremely important for normal Unified Communications
operation on a converged network. Proper infrastructure design requires following basic configuration
and design best practices for deploying a WAN that is as highly available as possible and that provides
guaranteed throughput. Furthermore, proper WAN infrastructure design requires deploying end-to-end
QoS on all WAN links. The following sections discuss these requirements:
For more information on bandwidth management, see the chapter on Bandwidth Management,
page 13-1.
Deployment Considerations
WAN deployments for voice and video networks may use a hub-and-spoke, fully meshed, or partially
meshed topology. A hub-and-spoke topology consists of a central hub site and multiple remote spoke
sites connected into the central hub site. In this scenario, each remote or spoke site is one WAN-link hop
away from the central or hub site and two WAN-link hops away from all other spoke sites. A meshed
topology may contain multiple WAN links and any number of hops between the sites. In this scenario
there may be many different paths to the same site or there may be different links used for
communication with some sites compared to other sites. The simplest example is three sites, each with
a WAN link to the other two sites, forming a triangle. In that case there are two potential paths between
each site to each other site.
For more information about centralized and distributed multisite deployment models as well as
Multiprotocol Label Switching (MPLS) implications for these deployment models, see the chapter on
Collaboration Deployment Models, page 10-1.
WAN links should, when possible, be made redundant to provide higher levels of fault tolerance.
Redundant WAN links provided by different service providers or located in different physical
ingress/egress points within the network can ensure backup bandwidth and connectivity in the event that
a single link fails. In non-failure scenarios, these redundant links may be used to provide additional
bandwidth and offer load balancing of traffic on a per-flow basis over multiple paths and equipment
within the WAN.
3-33
Chapter 3
Network Infrastructure
WAN Infrastructure
Voice, video, and data should remain converged at the WAN, just as they are converged at the LAN. QoS
provisioning and queuing mechanisms are typically available in a WAN environment to ensure that
voice, video, and data can interoperate on the same WAN links. Attempts to separate and forward voice,
video, and data over different links can be problematic in many instances because the failure of one link
typically forces all traffic over a single link, thus diminishing throughput for each type of traffic and in
most cases reducing the quality of voice. Furthermore, maintaining separate network links or devices
makes troubleshooting and management difficult at best.
Because of the potential for WAN links to fail or to become oversubscribed, Cisco recommends
deploying non-centralized resources as appropriate at sites on the other side of the WAN. Specifically,
media resources, DHCP servers, voice gateways, and call processing applications such as Survivable
Remote Site Telephony (SRST) and Cisco Unified Communications Manager Express (Unified CME)
should be deployed at non-central sites when and if appropriate, depending on the site size and how
critical these functions are to that site. Keep in mind that de-centralizing voice applications and devices
can increase the complexity of network deployments, the complexity of managing these resources
throughout the enterprise, and the overall cost of a the network solution; however, these factors can be
mitigated by the fact that the resources will be available during a WAN link failure.
When deploying voice in a WAN environment, it is possible to reduce bandwidth consumption by using
the lower-bandwidth G.729 codec for any voice calls that will traverse WAN links because this practice
will provide bandwidth savings on these lower-speed links. Furthermore, media resources such as MoH
can also be configured to use multicast transport mechanism when possible because this practice will
provide additional bandwidth savings.
Delay in IP Voice Networks
Recommendation G.114 of the International Telecommunication Union (ITU) states that the one-way
delay in a voice network should be less than or equal to 150 milliseconds. It is important to keep this in
mind when implementing low-speed WAN links within a network. Topologies, technologies, and
physical distance should be considered for WAN links so that one-way delay is kept at or below this
150-millisecond recommendation. Implementing a VoIP network where the one-way delay exceeds
150 milliseconds introduces issues not only with the quality of the voice call but also with call setup and
media cut-through times because several call signaling messages need to be exchanged between each
device and the call processing application in order to establish the call.
Guaranteed Bandwidth
Because voice is typically deemed a critical network application, it is imperative that bearer and
signaling voice traffic always reaches its destination. For this reason, it is important to choose a WAN
topology and link type that can provide guaranteed dedicated bandwidth. The following WAN link
technologies can provide guaranteed dedicated bandwidth:
Leased Lines
Frame Relay
These link technologies, when deployed in a dedicated fashion or when deployed in a private network,
can provide guaranteed traffic throughput. All of these WAN link technologies can be provisioned at
specific speeds or bandwidth sizes. In addition, these link technologies have built-in mechanisms that
help guarantee throughput of network traffic even at low link speeds. Features such as traffic shaping,
3-34
Chapter 3
Network Infrastructure
WAN Infrastructure
fragmentation and packet interleaving, and committed information rates (CIR) can help ensure that
packets are not dropped in the WAN, that all packets are given access at regular intervals to the WAN
link, and that enough bandwidth is available for all network traffic attempting to traverse these links.
Best-Effort Bandwidth
There are some WAN topologies that are unable to provide guaranteed dedicated bandwidth to ensure
that network traffic will reach its destination, even when that traffic is critical. These topologies are
extremely problematic for voice traffic, not only because they provide no mechanisms to provision
guaranteed network throughput, but also because they provide no traffic shaping, packet fragmentation
and interleaving, queuing mechanisms, or end-to-end QoS to ensure that critical traffic such as voice will
be given preferential treatment.
The following WAN network topologies and link types are examples of this kind of best-effort
bandwidth technology:
The Internet
DSL
Cable
Satellite
Wireless
In most cases, none of these link types can provide the guaranteed network connectivity and bandwidth
required for critical voice and voice applications. However, these technologies might be suitable for
personal or telecommuter-type network deployments. At times, these topologies can provide highly
available network connectivity and adequate network throughput; but at other times, these topologies can
become unavailable for extended periods of time, can be throttled to speeds that render network
throughput unacceptable for real-time applications such as voice, or can cause extensive packet losses
and require repeated retransmissions. In other words, these links and topologies are unable to provide
guaranteed bandwidth, and when traffic is sent on these links, it is sent best-effort with no guarantee that
it will reach its destination. For this reason, Cisco recommends that you do not use best-effort WAN
topologies for voice-enabled networks that require enterprise-class voice services and quality.
3-35
Chapter 3
Network Infrastructure
WAN Infrastructure
Note
There are some new QoS mechanisms for DSL and cable technologies that can provide guaranteed
bandwidth; however, these mechanisms are not typically deployed by many service providers. For any
service that offers QoS guarantees over networks that are typically based on best-effort, it is important
to review and understand the bandwidth and QoS guarantees offered in the service provider's service
level agreement (SLA).
Note
Upstream and downstream QoS mechanisms are now supported for wireless networks. For more
information on QoS for Voice over Wireless LANs, refer to the Voice over Wireless LAN Design Guide,
available at
http://www.cisco.com/en/US/solutions/ns340/ns414/ns742/ns818/landing_wireless_uc.html.
Enable queuing policies at every node where the potential for congestion exists, which generally
equates to attaching a comprehensive queuing policy to every WAN/VPN edge.
Protect the control plane and data plane by enabling control plane policing (on platforms supporting
this feature) as well as data plane policing (scavenger class QoS) to mitigate and constrain network
attacks.
To this end, this design section provides best-practice recommendations for enabling QoS over the wide
area network. However, it is important to note that the recommendations in this section are not
autonomous, but rather, they depend on the campus QoS design recommendations presented in the
section on LAN Quality of Service (QoS), page 3-14, having already been implemented. Traffic
traversing the WAN can thus be assumed to be correctly classified and marked with Layer 3 DSCP (as
well as policed at the access-edge, as necessary).
Furthermore, this design section covers fundamental considerations relating to wide area networks.
Before strategic QoS designs for the WAN can be derived, a few WAN-specific considerations need to
be taken into account, as are discussed below. Further information on bandwidth management in a
Collaboration solution can be found in the chapter on Bandwidth Management, page 13-1.
3-36
Chapter 3
Network Infrastructure
WAN Infrastructure
Each of these WAN QoS design considerations is discussed in the following sections.
Serialization (fixed)
Propagation (fixed)
Queuing (variable)
Serialization refers to the time it takes to convert a Layer 2 frame into Layer 1 electrical or optical pulses
onto the transmission media. Therefore, serialization delay is fixed and is a function of the line rate (that
is, the clock speed of the link). For example, a (1.544 Mbps) T1 circuit would require about 8 ms to
serialize a 1,500 byte Ethernet frame onto the wire, whereas a (9.953 Gbps) OC-192/STM-64 circuit
would require just 1.2 microseconds to serialize the same frame.
3-37
Chapter 3
Network Infrastructure
WAN Infrastructure
Usually, the most significant network factor in meeting the latency targets for over the WAN is
propagation delay, which can account for over 95% of the network latency time budget. Propagation
delay is also a fixed component and is a function of the physical distance that the signals have to travel
between the originating endpoint and the receiving endpoint. The gating factor for propagation delay is
the speed of light, which is 300,000 km/s or 186,000 miles per second in a vacuum. However, the speed
of light in an optical fiber is about one third the speed of light in a vacuum. Thus, the propagation delay
for most fiber circuits is approximately 6.3 microseconds per km or 8.2 microseconds per mile.
Another point to keep in mind when calculating propagation delay is that optical fibers are not always
physically placed over the shortest path between two geographic points, especially over transoceanic
links. Due to installation convenience, circuits may be hundreds or even thousands of miles longer than
theoretically necessary.
Nonetheless, the G.114 real-time communications network latency budget of 150 ms allows for nearly
24,000 km or 15,000 miles worth of propagation delay (which is approximately 60% of the earth's
circumference). The theoretical worst-case scenario (exactly half of the earth's circumference) would
require only 126 ms of latency. Therefore, this latency target is usually achievable for virtually any two
locations (via a terrestrial path), given relatively direct transmission paths; however, in some scenarios
meeting this latency target might simply not be possible due to the distances involved and the relative
directness of their respective transmission paths. In such scenarios, if the G.114 150 ms one-way latency
target cannot be met due to the distances involved, administrators should be aware that both the ITU and
Cisco Technical Marketing have shown that real-time communication quality does not begin to degrade
significantly until one-way latency exceeds 200 ms, as is illustrated in the ITU G.114 graph of real-time
speech quality versus absolute delay, which is reproduced in Figure 3-12.
3-38
Chapter 3
Network Infrastructure
WAN Infrastructure
Note
348825
Figure 3-12
This discussion so far has focused on WAN circuits over terrestrial paths. For satellite circuits, the
expected latency can be in the range of 250 to 900 ms. For example, signals being relayed via
geostationary satellites will need to be sent to an altitude of 35,786 km (22,236 miles) above sea level
(from the equator) out into space and then back to Earth again. There is nothing an administrator can do
to decrease latency in such scenarios because they can do nothing about increasing the speed of light or
radio waves. All that can be done to address the effect of latency in these scenarios is to educate the
user-base so that realistic performance expectations are set.
The final network latency component to be considered is queuing delay, which is variable (variable delay
is also known as jitter). Queuing delay is a function of whether a network node is congested and, if so,
what scheduling policies have been applied to resolve congestion events. Real-time applications are
often more sensitive to jitter than latency, because packets need to be received in de-jitter buffers prior
to being played out. If a packet is not received within the time allowed by the de-jitter buffer, it is
essentially lost and can affect the overall voice or video call quality.
3-39
Chapter 3
Network Infrastructure
WAN Infrastructure
Given that the majority of factors contributing to network latency are fixed, careful attention has to be
given to queuing delay, since this is the only latency factor that is directly under the network
administrator's control via queuing policies. Therefore, a close examination of the Cisco IOS queuing
system, including the Tx-Ring and LLQ/CBWFQ operation, will assist network administrators to
optimize these critical policies.
Tx-Ring
The Tx-Ring is the final Cisco IOS output buffer for a WAN interface (a relatively small FIFO queue),
and it maximizes physical link bandwidth utilization by matching the outbound packet rate on the router
with the physical interface rate. The Tx-Ring is illustrated in Figure 3-13.
Figure 3-13
Packets
Out
348826
Packets
In
The Tx-Ring also serves to indicate interface congestion to the Cisco IOS software. Prior to interface
congestion, packets are sent on a FIFO basis to the interface via the Tx-Ring. However, when the
Tx-Ring fills to its queue limit, then it signals to the Cisco IOS software to engage any LLQ or CBWFQ
policies that have been attached to the interface. Subsequent packets are then queued within Cisco IOS
according to these LLQ and CBWFQ policies, dequeued into the Tx-Ring, and then sent out the interface
in a FIFO manner.
The Tx-Ring can be configured on certain platforms with the tx-ring-limit interface configuration
command. The default value of the Tx-Ring varies according to platform and link type and speed. For
further details, refer to Understanding and Tuning the tx-ring-limit Value, available at
http://www.cisco.com/c/en/us/support/docs/asynchronous-transfer-mode-atm/ip-to-atm-class-of-se
rvice/6142-txringlimit-6142.html
Changing the Tx-Ring Default Setting
During Cisco Technical Marketing design validation, it was observed that the default Tx-Ring limit on
some interfaces caused somewhat higher jitter values to some real-time application classes, particularly
HD video-based real-time applications such as Cisco TelePresence traffic. The reason for this is the
bursty nature of HD video traffic. For example, consider a fully-congested T3 WAN link (using a Cisco
PA-T3+ port adapter interface) with active LLQ and CBWFQ policies. The default Tx-Ring depth in this
case is 64 packets. Even if TelePresence traffic is prioritized via an LLQ, if there are no TelePresence
packets to send, the FIFO Tx-Ring is filled with other traffic to a default depth of 64 packets. When a
new TelePresence packet arrives, even if it gets priority treatment from the Layer 3 LLQ/CBWFQ
queuing system, the packets are dequeued into the FIFO Tx-Ring when space is available. However, with
the default settings, there can be as many as 63 packets in the Tx-Ring in front of that TelePresence
3-40
Chapter 3
Network Infrastructure
WAN Infrastructure
packet. In such a worst-case scenario it could take as long as 17 ms to transmit these non-real-time
packets out of this (45 Mbps) T3 interface. This 17 ms of instantaneous and variable delay (jitter) can
affect the video quality for TelePresence to the point of being visually apparent to the end user. However,
lowering the value of the Tx-Ring on this link will force in the Cisco IOS software engaging congestion
management policies sooner and more often, resulting in lower overall jitter values for real-time
applications such as TelePresence.
On the other hand, setting the value of the Tx-Ring too low might result in significantly higher CPU
utilization rates because the processor is continually being interrupted to engage queuing policies, even
when congestion rates are just momentary bursts and not sustained rates. Thus, when tuning the Tx-Ring,
a trade-off setting is required so that jitter is minimized, but not at the expense of excessive CPU
utilization rates.
Therefore, explicit attention needs be given to link types and speeds when the Tx-Ring is tuned away
from default values.
3-41
Chapter 3
Network Infrastructure
WAN Infrastructure
Figure 3-14
OAM CBWFQ
Packets
In
FQ
Tx-Ring
FQ
Packets
Out
CBWFQ
FQ
Pre-Sorters
348827
Each CBWFQ class is guaranteed bandwidth via a bandwidth policy-map class configuration statement.
CBWFQ derives the weight for packets belonging to a class from the bandwidth allocated to the class.
CBWFQ then uses the weight to ensure that the queue for the class is serviced fairly, via WRR
scheduling.
An important point regarding bandwidth assigned to a given CBWFQ class is that the bandwidth
allocated is not a static bandwidth reservation, but rather represents a minimum bandwidth guarantee to
the class, provided there are packets offered to the class. If there are no packets offered to the class, then
the scheduler services the next queue and can dynamically redistribute unused bandwidth allocations to
other queues as necessary.
Additionally, a fair-queuing pre-sorter may be applied to specific CBWFQ queues with the fair-queue
policy-map class configuration command. It should be noted that this command enables a flow-based
fair-queuing pre-sorter, and not a weighted fair-queuing pre-sorter, as the name for this feature implies
(and as such, the fair-queuing pre-sorter does not take into account the IP Precedence values of any
packets offered to a given class). For example, if a CBWFQ class was assigned 1 Mbps of bandwidth
and there were 4 competing traffic flows contending for this class, a fair-queuing pre-sorter would ensure
that each flow receives (1 / (total-number-of-flows)) of bandwidth, or in this example (1/4 of 1 Mpbs)
250 kbps of bandwidth.
3-42
Chapter 3
Network Infrastructure
WAN Infrastructure
Note
Prior to Cisco IOS Release 12.4(20)T, a fair-queue pre-sorter could be applied only to class-default;
however, subsequent Cisco IOS releases include the support of the Hierarchical Queuing Framework
(HQF) which, among many other QoS feature enhancements, allows for a fair-queue pre-sorter to be
applied to any CBWFQ class. HQF details are documented at
http://www.cisco.com/c/en/us/td/docs/ios-xml/ios/qos_hrhqf/configuration/15-mt/qos-hrhqf-15-mt-boo
k.html.
The depth of a CBWFQ is defined by its queue limit, which varies according to link speeds and
platforms. This queue limit can be modified with the queue-limit policy-map class configuration
command. In some cases, such as provisioning (bursty) TelePresence traffic in a CBWFQ, it is
recommended to increase the queue limit from the default value. This is discussed in more detail in the
section on Weighted-Random Early Detect, page 3-44.
Older (pre-HQF and pre-12.4(20)T) versions of Cisco IOS software include a legacy feature that
disallows LLQ/CBWFQ policies from being attached to an interface if those policies explicitly allocate
more than 75% of the interface's bandwidth to non-default traffic classes. This was intended as a safety
feature that would always allow the default class as well as control-traffic classes to receive adequate
bandwidth, and it allowed provisioning for Layer 2 bandwidth overhead. This feature can be overridden
by applying the max-reserved-bandwidth interface command, which takes as a parameter the total
percentage of interface bandwidth that can be explicitly provisioned (typically this value is set to 100).
However, if this safety feature is overridden, then it is highly recommended that the default class be
explicitly assigned no less than 25% of the link's bandwidth.
Low-Latency Queuing
Low-Latency Queuing (LLQ) is essentially CBWFQ combined with a strict priority queue. Basic LLQ
operation is illustrated in Figure 3-15.
Figure 3-15
LLQ
Packets
In
Packets
Out
CBWFQ
Scheduler
Tx-Ring
CBWFQ
348828
FQ
Pre-Sorters
3-43
Chapter 3
Network Infrastructure
WAN Infrastructure
As shown in Figure 3-15, LLQ adds a strict-priority queue to the CBWFQ subsystem. The amount of
bandwidth allocated to the LLQ is set by the priority policy-map class configuration command An
interesting facet of Cisco IOS LLQ is the inclusion of an implicit policer that admits packets to the
strict-priority queue. This implicit policer limits the bandwidth that can be consumed by servicing the
real-time queue, and it thus prevents bandwidth starvation of the non-real-time flows serviced by the
CBWFQ scheduler. The policing rate for this implicit policer is always set to match the bandwidth
allocation of the strict-priority queue. If more traffic is offered to the LLQ class than it has been
provisioned to accommodate, then the excess traffic will be dropped by the policer. And like the
LLQ/CBWFQ systems, the implicit policer is active only during the event of congestion (as signaled to
the Cisco IOS software by means of a full Tx-Ring).
Tail
of
Queue
Front
of
Queue
FairQueue
PreSorter
Direction of
Packet Flow
AF43 Minimum WRED Threshold:
Begin randomly dropping AF43 Packets
AF42 Minimum WRED Threshold:
Begin randomly dropping AF42 Packets
Maximum WRED Thresholds for AF41, AF42, and AF43 are set
to the tail of the queue in this example.
348829
3-44
Chapter 3
Network Infrastructure
WAN Infrastructure
As shown in Figure 3-16, packets marked with a given Drop Precedence (AF43, AF42, or AF41) will
begin to be dropped only when the queue fills beyond the minimum WRED threshold for the Drop
Precedence value. Packets are always dropped randomly, but their probability of being dropped increases
as the queue fills nearer the maximum WRED threshold for the Drop Precedence value. The maximum
WRED thresholds are typically set at 100% (the tail of the queue), as shown in Figure 3-16; but the
thresholds are configurable, and some advanced administrators may tune these WRED thresholds
according to their needs, constraints, and preferences.
Additionally, the WRED thresholds on the AF class may be optimized. By default the minimum WRED
thresholds for each AF class are 24, 28, and 32 packets for Drop-Precedence values 3, 2, and 1
respectively. These thresholds represent 60%, 70%, and 80% respectively of the default queue-depth of
64 packets. Also, by default the maximum WRED thresholds are set to 40 packets for all
Drop-Precedence values for each AF class. Considering that the default queue-limit or depth is
64 packets, these default settings are inefficient on links experiencing sustained congestion that can
cause a queue-depth of 40 packets (at which point all code points will be tail-dropped, despite the queue
having the capacity to accommodate another 24 packets). Thus, an administrator may choose to tune
these WRED thresholds so that each AF class has a minimum WRED threshold of 40, 45, and 50 packets
for Drop-Precedence values 3, 2, and 1 respectively, which represent approximately 60%, 70%, and 80%
of the default queue-depth of 64 packets, and/or the administrator may choose to tune the maximum
WRED thresholds for each Drop-Precedence value for each AF class to the default queue-depth of
64 packets.
An example design is presented in the chapter on Bandwidth Management, page 13-1.
3-45
Chapter 3
Network Infrastructure
WAN Infrastructure
Table 3-5
QoS Features and Tools Required to Support Unified Communications for Each WAN Technology and Link
Speed
WAN Technology
Leased Lines
Asynchronous Transfer
Mode (ATM)
Multiprotocol Label
Switching (MPLS)
LLQ
Traffic Shaping
Traffic Shaping
LFI (FRF.12)
LLQ
LLQ
Optional: VATS
Optional: cRTP
LLQ
MLP LFI
LLQ
MLP LFI
LLQ
LLQ
Class-based marking is
generally required to re-mark
flows according to service
provider specifications
The following sections highlight some of the most important features and techniques to consider when
designing a WAN to support voice, video, and data traffic:
3-46
Chapter 3
Network Infrastructure
WAN Infrastructure
Traffic Prioritization
In choosing from among the many available prioritization schemes, the major factors to consider include
the type of traffic involved and the type of media on the WAN. For multi-service traffic over an IP WAN,
Cisco recommends low-latency queuing (LLQ) for all links. This method supports up to 64 traffic
classes, with the ability to specify, for example, priority queuing behavior for voice and interactive
video, minimum bandwidth class-based weighted fair queuing for voice control traffic, additional
minimum bandwidth weighted fair queues for mission critical data, and a default best-effort queue for
all other traffic types.
Figure 3-17 shows an example prioritization scheme.
Packets in
Link fragmentation
and interleave
1 1 1 1
2
PQ Voice
Interleave
Class = X
3 3
Class = Y CBWFQ
4 4 4
0 0 0 0
Police
PQ Voice
WFQ
TX
ring
Packets out
0 4 3 2 1 1
Fragment
Packets out
Default
77295
Figure 3-17
The criterion for voice to be placed into a priority queue is a DSCP value of 46 (EF).
The criterion for video conferencing traffic to be placed into a class-based weighted fair queue
(CBWFQ) is a DSCP value of 34 (AF41). Due to the larger packet sizes of video traffic, link speeds
below 768 Kbps require packet fragmentation, which can happen only when video is placed in a
separate CBWFQ. Video in a priority queue (PQ) is not fragmented.
As the WAN links become congested, it is possible to starve the voice control signaling protocols,
thereby eliminating the ability of the IP phones to complete calls across the IP WAN. Therefore,
voice control protocols, such as H.323, MGCP, and Skinny Client Control Protocol (SCCP), require
their own class-based weighted fair queue. The entrance criterion for this queue is a DSCP value of
24 (CS3).
In some cases, certain data traffic might require better than best-effort treatment. This traffic is
referred to as mission-critical data, and it is placed into one or more queues that have the required
amount of bandwidth. The queuing scheme within this class is first-in-first-out (FIFO) with a
minimum allocated bandwidth. Traffic in this class that exceeds the configured bandwidth limit is
placed in the default queue. The entrance criterion for this queue could be a Transmission Control
Protocol (TCP) port number, a Layer 3 address, or a DSCP/PHB value.
All remaining enterprise traffic can be placed in a default queue for best-effort treatment. If you
specify the keyword fair, the queuing algorithm will be weighted fair queuing (WFQ).
3-47
Chapter 3
Network Infrastructure
WAN Infrastructure
Scavenger Class
The Scavenger class is intended to provide less than best-effort services to certain applications.
Applications assigned to this class have little or no contribution to the organizational objectives of the
enterprise and are typically entertainment oriented in nature. Assigning Scavenger traffic to a minimal
bandwidth queue forces it to be squelched to virtually nothing during periods of congestion, but it allows
it to be available if bandwidth is not being used for business purposes, such as might occur during
off-peak hours.
Scavenger traffic should be assigned the lowest configurable queuing service. For instance, in
Cisco IOS, this means assigning a CBWFQ of 1% to Scavenger class.
You can increase link efficiency by using Compressed Real-Time Transport Protocol (cRTP). This
protocol compresses a 40-byte IP, User Datagram Protocol (UDP), and RTP header into approximately
two to four bytes. cRTP operates on a per-hop basis. Use cRTP on a particular link only if that link meets
all of the following conditions:
Voice traffic represents more than 33% of the load on the specific link.
No other real-time application (such as video conferencing) is using the same link.
If the link fails to meet any one of the preceding conditions, then cRTP is not effective and you should
not use it on that link. Another important parameter to consider before using cRTP is router CPU
utilization, which is adversely affected by compression and decompression operations.
cRTP on ATM and Frame Relay Service Inter-Working (SIW) links requires the use of Multilink
Point-to-Point Protocol (MLP).
Note that cRTP compression occurs as the final step before a packet leaves the egress interface; that is,
after LLQ class-based queueing has occurred. Beginning in Cisco IOS Release 12.(2)2T and later, cRTP
provides a feedback mechanism to the LLQ class-based queueing mechanism that allows the bandwidth
in the voice class to be configured based on the compressed packet value. With Cisco IOS releases prior
to 12.(2)2T, this mechanism is not in place, so the LLQ is unaware of the compressed bandwidth and,
therefore, the voice class bandwidth has to be provisioned as if no compression is taking place. Table 3-6
shows an example of the difference in voice class bandwidth configuration given a 512-kbps link with
G.729 codec and a requirement for 10 calls.
Note that Table 3-6 assumes 24 kbps for non-cRTP G.729 calls and 10 kbps for cRTP G.729 calls. These
bandwidth numbers are based on voice payload and IP/UDP/RTP headers only. They do not take into
consideration Layer 2 header bandwidth. However, actual bandwidth provisioning should also include
Layer 2 header bandwidth based on the type WAN link used.
Table 3-6
LLQ Voice Class Bandwidth Requirements for 10 Calls with 512 kbps Link Bandwidth
and G.729 Codec
Prior to 12.2(2)T
240 kbps
240 kbps1
12.2(2)T or later
240 kbps
100 kbps
3-48
Chapter 3
Network Infrastructure
WAN Infrastructure
1. 140 kbps of unnecessary bandwidth must be configured in the LLQ voice class.
It should also be noted that, beginning in Cisco IOS Release 12.2(13)T, cRTP can be configured as part
of the voice class with the Class-Based cRTP feature. This option allows cRTP to be specified within a
class, attached to an interface via a service policy. This new feature provides compression statistics and
bandwidth status via the show policy interface command, which can be very helpful in determining the
offered rate on an interface service policy class given the fact that cRTP is compressing the IP/RTP
headers.
For additional recommendations about using cRTP with a Voice and Video Enabled IPSec VPN (V3PN),
refer to the V3PN documentation available at
http://www.cisco.com/en/US/solutions/ns340/ns414/ns742/ns817/landing_voice_video.html
Link Fragmentation and Interleaving (LFI)
For low-speed links (less than 768 kbps), use of link fragmentation and interleaving (LFI) mechanisms
is required for acceptable voice quality. This technique limits jitter by preventing voice traffic from being
delayed behind large data frames, as illustrated in Figure 3-18. The two techniques that exist for this
purpose are Multilink Point-to-Point Protocol (MLP) LFI (for Leased Lines, ATM, and SIW) and FRF.12
for Frame Relay.
Figure 3-18
Before
Data
Voice
Data
Data
Voice
Data
77296
After
In addition to the LFI mechanisms mentioned above, voice-adaptive fragmentation (VAF) is another LFI
mechanism for Frame Relay links. VAF uses FRF.12 Frame Relay LFI; however, once configured,
fragmentation occurs only when traffic is present in the LLQ priority queue or when H.323 signaling
packets are detected on the interface. This method ensures that, when voice traffic is being sent on the
WAN interface, large packets are fragmented and interleaved. However, when voice traffic is not present
on the WAN link, traffic is forwarded across the link unfragmented, thus reducing the overhead required
for fragmentation.
VAF is typically used in combination with voice-adaptive traffic shaping (see Voice-Adaptive Traffic
Shaping (VATS), page 3-51). VAF is an optional LFI tool, and you should exercise care when enabling
it because there is a slight delay between the time when voice activity is detected and the time when the
3-49
Chapter 3
Network Infrastructure
WAN Infrastructure
LFI mechanism engages. In addition, a configurable deactivation timer (default of 30 seconds) must
expire after the last voice packet is detected and before VAF is deactivated, so during that time LFI will
occur unnecessarily. VAF is available in Cisco IOS Release 12.2(15)T and later.
Traffic Shaping
Traffic shaping is required for multiple-access, non-broadcast media such as ATM and Frame Relay,
where the physical access speed varies between two endpoints and several branch sites are typically
aggregated to a single router interface at the central site.
Figure 3-19 illustrates the main reasons why traffic shaping is needed when transporting voice and data
on the same IP WAN.
Figure 3-19
Central Site
over-subscription
To prevent bursting above
Committed Rate (CIR)
T1
Frame Relay
or ATM
CIR = 64 kbps
T1
T1
2
Remote Sites
T1
253922
64kbps
3-50
Chapter 3
Network Infrastructure
WAN Infrastructure
2.
Oversubscription of the link between the central site and the remote sites
It is common practice in Frame Relay or ATM networks to oversubscribe bandwidth when
aggregating many remote sites to a single central site. For example, there may be multiple remote
sites that connect to the WAN with a T1 interface, yet the central site has only a single T1 interface.
While this configuration allows the deployment to benefit from statistical multiplexing, the router
interface at the central site can become congested during traffic bursts, thus degrading voice quality.
3.
Traffic shaping provides a solution to these issues by limiting the traffic sent out an interface to a rate
lower than the line rate, thus ensuring that no congestion occurs on either end of the WAN. Figure 3-20
illustrates this mechanism with a generic example, where R is the rate with traffic shaping applied.
Figure 3-20
Line
Rate
77298
VATS is an optional dynamic mechanism that shapes traffic on Frame Relay permanent virtual circuits
(PVCs) at different rates based on whether voice is being sent across the WAN. The presence of traffic
in the LLQ voice priority queue or the detection of H.323 signaling on the link causes VATS to engage.
Typically, Frame Relay shapes traffic to the guaranteed bandwidth or CIR of the PVC at all times.
However, because these PVCs are typically allowed to burst above the CIR (up to line speed), traffic
shaping keeps traffic from using the additional bandwidth that might be present in the WAN. With VATS
enabled on Frame Relay PVCs, WAN interfaces are able to send at CIR when voice traffic is present on
the link. However, when voice is not present, non-voice traffic is able to burst up to line speed and take
advantage of the additional bandwidth that might be present in the WAN.
When VATS is used in combination with voice-adaptive fragmentation (VAF) (see Link Fragmentation
and Interleaving (LFI), page 3-49), all non-voice traffic is fragmented and all traffic is shaped to the CIR
of the WAN link when voice activity is detected on the interface.
As with VAF, exercise care when enabling VATS because activation can have an adverse effect on
non-voice traffic. When voice is present on the link, data applications will experience decreased
throughput because they are throttled back to well below CIR. This behavior will likely result in packet
drops and delays for non-voice traffic. Furthermore, after voice traffic is no longer detected, the
deactivation timer (default of 30 seconds) must expire before traffic can burst back to line speed. It is
3-51
Chapter 3
Network Infrastructure
WAN Infrastructure
important, when using VATS, to set end-user expectations and make them aware that data applications
will experience slowdowns on a regular basis due to the presence of voice calls across the WAN. VATS
is available in Cisco IOS Release 12.2(15)T and later.
For more information on the Voice-Adaptive Traffic Shaping and Fragmentation features and how to
configure them, refer to the documentation at
http://www.cisco.com/en/US/docs/ios/12_2t/12_2t15/feature/guide/ft_vats.html
Bandwidth Provisioning
Properly provisioning the network bandwidth is a major component of designing a successful IP
network. You can calculate the required bandwidth by adding the bandwidth requirements for each major
application (for example, voice, video, and data). This sum then represents the minimum bandwidth
requirement for any given link, and it should not exceed approximately 75% of the total available
bandwidth for the link. This 75% rule assumes that some bandwidth is required for overhead traffic, such
as routing and Layer 2 keep-alives. Figure 3-21 illustrates this bandwidth provisioning process.
Voice
Video
Voice/Video
Control
Data
Routing
etc.
Reserved
Link capacity
77291
Figure 3-21
In addition to using no more than 75% of the total available bandwidth for data, voice, and video, the
total bandwidth configured for all LLQ priority queues should typically not exceed 33% of the total link
bandwidth. Provisioning more than 33% of the available bandwidth for the priority queue can be
problematic for a number of reasons. First, provisioning more than 33% of the bandwidth for voice can
result in increased CPU usage. Because each voice call will send 50 packets per second (with 20 ms
samples), provisioning for large numbers of calls in the priority queue can lead to high CPU levels due
to high packet rates. In addition, if more than one type of traffic is provisioned in the priority queue (for
example, voice and video), this configuration defeats the purpose of enabling QoS because the priority
queue essentially becomes a first-in, first-out (FIFO) queue. A larger percentage of reserved priority
bandwidth effectively dampens the QoS effects by making more of the link bandwidth FIFO. Finally,
allocating more than 33% of the available bandwidth can effectively starve any data queues that are
provisioned. Obviously, for very slow links (less than 192 kbps), the recommendation to provision no
more than 33% of the link bandwidth for the priority queue(s) might be unrealistic because a single call
could require more than 33% of the link bandwidth. In these situations, and in situations where specific
business needs cannot be met while holding to this recommendation, it may be necessary to exceed the
33% rule.
3-52
Chapter 3
Network Infrastructure
WAN Infrastructure
The voice and video bearer streams, which consists of Real-Time Transport Protocol (RTP) packets
that contain the actual voice samples.
The call control signaling, which consists of packets belonging to one of several protocols,
according to the endpoints involved in the call (for example, H.323, MGCP, SCCP, or (J)TAPI). Call
control functions are, for instance, those used to set up, maintain, tear down, or redirect a call.
Bandwidth provisioning should include not only the bearer traffic but also the call control traffic. In fact,
in multisite WAN deployments, the call control traffic (as well as the bearer traffic) must traverse the
WAN, and failure to allocate sufficient bandwidth for it can adversely affect the user experience.
The next three sub-sections describe the bandwidth provisioning recommendations for the following
types of traffic:
Voice and video bearer traffic in all multisite WAN deployments (see Provisioning for Bearer
Traffic, page 3-53)
Call control traffic in multisite WAN deployments with centralized call processing (see Provisioning
for Call Control Traffic with Centralized Call Processing, page 3-57)
Call control traffic in multisite WAN deployments with distributed call processing (see Provisioning
for Call Control Traffic with Distributed Call Processing, page 3-60)
Voice
payload
RTP
Header
UDP
Header
IP
Header
Link
Header
X bytes
12 bytes
8 bytes
20 bytes
X bytes
77292
VoIP Packet
The bandwidth consumed by VoIP streams is calculated by adding the packet payload and all headers (in
bits), then multiplying by the packet rate per second, as follows:
Layer 2 bandwidth in kbps = [(Packets per second) (X bytes for voice payload + 40 bytes for
RTP/UDP/IP headers + Y bytes for Layer 2 overhead) 8 bits] / 1000
Layer 3 bandwidth in kbps = [(Packets per second) (X bytes for voice payload + 40 bytes for
RTP/UDP/IP headers) 8 bits] / 1000
Packets per second = [1/(sampling rate in msec)] 1000
3-53
Chapter 3
Network Infrastructure
WAN Infrastructure
Voice payload in bytes = [(codec bit rate in kbps) (sampling rate in msec)] / 8
Table 3-7 details the Layer 3 bandwidth per VoIP flow. Table 3-7 lists the bandwidth consumed by the
voice payload and IP header only, at a default packet rate of 50 packets per second (pps) and at a rate of
33.3 pps for both non-encrypted and encrypted payloads. Table 3-7 does not include Layer 2 header
overhead and does not take into account any possible compression schemes, such as compressed
Real-Time Transport Protocol (cRTP). You can use the Service Parameters menu in Unified CM
Administration to adjust the codec sampling rate.
Table 3-7
CODEC
Sampling Rate
Voice Payload
in Bytes
Packets per
Second
Bandwidth per
Conversation
20 ms
160
50.0
80.0 kbps
20 ms
164
50.0
81.6 kbps
30 ms
240
33.3
74.7 kbps
30 ms
244
33.3
75.8 kbps
iLBC
20 ms
38
50.0
31.2 kbps
iLBC (SRTP)
20 ms
42
50.0
32.8 kbps
iLBC
30 ms
50
33.3
24.0 kbps
iLBC (SRTP)
30 ms
54
33.3
25.1 kbps
G.729A
20 ms
20
50.0
24.0 kbps
G.729A (SRTP)
20 ms
24
50.0
25.6 kbps
G.729A
30 ms
30
33.3
18.7 kbps
G.729A (SRTP)
30 ms
34
33.3
19.8 kbps
A more accurate method for provisioning is to include the Layer 2 headers in the bandwidth calculations.
Table 3-8 lists the amount of bandwidth consumed by voice traffic when the Layer 2 headers are included
in the calculations.
3-54
Chapter 3
Network Infrastructure
WAN Infrastructure
Table 3-8
Ethernet
14 Bytes
PPP
6 Bytes
ATM
53-Byte Cells
with a 48-Byte
Payload
85.6 kbps
82.4 kbps
106.0 kbps
81.6 kbps
84.0 kbps
81.6 kbps
89.6 kbps
87.2 kbps
84.0 kbps
106.0 kbps
83.2 kbps
85.6 kbps
83.2 kbps
N/A
78.4 kbps
76.3 kbps
84.8 kbps
75.7 kbps
77.3 kbps
75.7 kbps
81.1 kbps
79.5 kbps
77.4 kbps
84.8 kbps
76.8 kbps
78.4 kbps
76.8 kbps
N/A
36.8 kbps
33.6 kbps
42.4 kbps
32.8 kbps
35.2 kbps
32.8 kbps
40.8 kbps
iLBC (SRTP) at
50.0 pps
38.4 kbps
35.2 kbps
42.4 kbps
34.4 kbps
36.8 kbps
34.4 kbps
42.4 kbps
27.7 kbps
25.6 kbps
28.3 kbps
25.0 kbps
26.6 kbps
25.0 kbps
30.4 kbps
iLBC (SRTP) at
33.3 pps
28.8 kbps
26.6 kbps
42.4 kbps
26.1 kbps
27.7 kbps
26.1 kbps
31.5 kbps
29.6 kbps
26.4 kbps
42.4 kbps
25.6 kbps
28.0 kbps
25.6 kbps
33.6 kbps
G.729A (SRTP) at
50.0 pps
31.2 kbps
28.0 kbps
42.4 kbps
27.2 kbps
29.6 kbps
27.2 kbps
35.2 kbps
22.4 kbps
20.3 kbps
28.3 kbps
19.7 kbps
21.3 kbps
19.8 kbps
25.1 kbps
G729A (SRTP) at
33.3 pps
23.5 kbps
21.4 kbps
28.3 kbps
20.8 kbps
22.4 kbps
20.8 kbps
26.2 kbps
CODEC
Frame Relay
4 Bytes
MLPPP
10 Bytes
MPLS
4 Bytes
WLAN
24 Bytes
While it is possible to configure the sampling rate above 30 ms, doing so usually results in very poor
voice quality. As illustrated in Figure 3-23, as sampling size increases, the number of packets per second
decreases, resulting in a smaller impact to the CPU of the device. Likewise, as the sample size increases,
IP header overhead is lower because the payload per packet is larger. However, as sample size increases,
so does packetization delay, resulting in higher end-to-end delay for voice traffic. The trade-off between
packetization delay and packets per second must be considered when configuring sample size. While this
trade-off is optimized at 20 ms, 30 ms sample sizes still provide a reasonable ratio of delay to packets
per second; however, with 40 ms sample sizes, the packetization delay becomes too high.
3-55
Chapter 3
Network Infrastructure
WAN Infrastructure
Figure 3-23
Trade off
50
120
45
100
35
80
30
25
60
20
40
15
40
10
20
5
0
0
20 ms
30 ms
40 ms
Sample Size
Packetization Delay
114470
10 ms
3-56
Chapter 3
Network Infrastructure
WAN Infrastructure
Each time a remote branch phone places a call, the control traffic traverses the IP WAN to reach the
Unified CM at the central site, even if the call is local to the branch.
The signaling protocols that may traverse the IP WAN in this deployment model are SCCP
(encrypted and non-encrypted), SIP (encrypted and non-encrypted), H.323, MGCP, and CTI-QBE.
All the control traffic is exchanged between a Unified CM at the central site and endpoints or
gateways at the remote branches.
As a consequence, you must provision bandwidth for control traffic that traverses the WAN between the
branch routers and the WAN aggregation router at the central site.
The control traffic that traverses the WAN in this scenario can be split into two categories:
Quiescent traffic, which consists of keep-alive messages periodically exchanged between the branch
endpoints (phones and gateways) and Unified CM, regardless of call activity. This traffic is a
function of the quantity of endpoints.
Call-related traffic, which consists of signaling messages exchanged between the branch endpoints
and the Unified CM at the central site when a call needs to be set up, torn down, forwarded, and so
forth. This traffic is a function of the quantity of endpoints and their associated call volume.
To obtain an estimate of the generated call control traffic, it is necessary to make some assumptions
regarding the average number of calls per hour made by each branch IP phone. In the interest of
simplicity, the calculations in this section assume an average of 10 calls per hour per phone.
Note
If this average number does not satisfy the needs of your specific deployment, you can calculate the
recommended bandwidth by using the advanced formulas provided in Advanced Formulas, page 3-58.
Given the assumptions made, and initially considering the case of a remote branch with no signaling
encryption configured, the recommended bandwidth needed for call control traffic can be obtained from
the following formula:
Equation 1A: Recommended Bandwidth Needed for SCCP Control Traffic without Signaling
Encryption.
Bandwidth (bps) = 265 (Number of IP phones and gateways in the branch)
Equation 1B: Recommended Bandwidth Needed for SIP Control Traffic without Signaling Encryption.
Bandwidth (bps) = 538 (Number of IP phones and gateways in the branch)
3-57
Chapter 3
Network Infrastructure
WAN Infrastructure
If a site features a mix of SCCP and SIP endpoints, the two equations above should be employed
separately for the quantity of each type of phone used, and the results added.
Equation 1 and all other formulas within this section include a 25% over-provisioning factor. Control
traffic has a bursty nature, with peaks of high activity followed by periods of low activity. For this reason,
assigning just the minimum bandwidth required to a control traffic queue can result in undesired effects
such as buffering delays and, potentially, packet drops during periods of high activity. The default queue
depth for a Class-Based Weighted Fair Queuing (CBWFQ) queue in Cisco IOS equals 64 packets. The
bandwidth assigned to this queue determines its servicing rate. Assuming that the bandwidth configured
is the average bandwidth consumed by this type of traffic, it is clear that, during the periods of high
activity, the servicing rate will not be sufficient to "drain" all the incoming packets out of the queue, thus
causing them to be buffered. Note that, if the 64-packet limit is reached, any subsequent packets are
either assigned to the best-effort queue or are dropped. It is therefore advisable to introduce this 25%
over-provisioning factor to absorb and smooth the variations in the traffic pattern and to minimize the
risk of a temporary buffer overrun. This is equivalent to increasing the servicing rate of the queue.
If encryption is configured, the recommended bandwidth is affected because encryption increases the
size of signaling packets exchanged between Unified CM and the endpoints. The following formula
takes into account the impact of signaling encryption:
Equation 2A: Recommended Bandwidth Needed for SCCP Control Traffic with Signaling Encryption.
Bandwidth with signaling encryption (bps) = 415 (Number of IP phones and gateways in the
branch)
Equation 2B: Recommended Bandwidth Needed for SIP Control Traffic with Signaling Encryption.
Bandwidth with signaling encryption (bps) = 619 (Number of IP phones and gateways in the
branch)
If we now take into account the fact that the smallest bandwidth that can be assigned to a queue on a
Cisco IOS router is 8 kbps, we can summarize the values of minimum and recommended bandwidth for
various branch office sizes, as shown in Table 3-9.
Table 3-9
Recommended Layer 3 Bandwidth for Call Control Traffic With and Without Signaling Encryption
Recommended
Bandwidth for SCCP
Control Traffic (no
encryption)
Recommended
Bandwidth for SCCP
Control Traffic (with
encryption)
Recommended
Bandwidth for SIP
Control Traffic (no
encryption)
Recommended
Bandwidth for SIP
Control Traffic (with
encryption)
1 to 10
8 kbps
8 kbps
8 kbps
8 kbps
20
8 kbps
9 kbps
11 kbps
12 kbps
30
8 kbps
13 kbps
16 kbps
19 kbps
40
11 kbps
17 kbps
22 kbps
25 kbps
50
14 kbps
21 kbps
27 kbps
31 kbps
100
27 kbps
42 kbps
54 kbps
62 kbps
150
40 kbps
62 kbps
81 kbps
93 kbps
Advanced Formulas
The previous formulas presented in this section assume an average call rate per phone of 10 calls per
hour. However, this rate might not correspond to your deployment if the call patterns are significantly
different (for example, with call center agents at the branches). To calculate call control bandwidth
requirements in these cases, use the following formulas, which contain an additional variable (CH) that
represents the average calls per hour per phone:
3-58
Chapter 3
Network Infrastructure
WAN Infrastructure
Equation 3A: Recommended Bandwidth Needed for SCCP Control Traffic for a Branch with No
Signaling Encryption.
Bandwidth (bps) = (53 + 21 CH) (Number of IP phones and gateways in the branch)
Equation 3B: Recommended Bandwidth Needed for SIP Control Traffic for a Branch with No Signaling
Encryption.
Bandwidth (bps) = (138 + 40 CH) (Number of IP phones and gateways in the branch)
Equation 4A: Recommended Bandwidth Needed for SCCP Control Traffic for a Remote Branch with
Signaling Encryption.
Bandwidth with signaling encryption (bps) = (73.5 + 33.9 CH) (Number of IP phones and
gateways in the branch)
Equation 4B: Recommended Bandwidth Needed for SIP Control Traffic for a Remote Branch with
Signaling Encryption.
Bandwidth with signaling encryption (bps) = (159 + 46 CH) (Number of IP phones and gateways
in the branch)
Note
Equations 3A and 4A are based on the default SCCP keep-alive period of 30 seconds, while equations
3B and 4B are based on the default SIP keep-alive period of 120 seconds.
Considerations for Shared Line Appearances
Calls placed to shared line appearances, or calls sent to line groups using the Broadcast distribution
algorithm, have two net effects on the bandwidth consumed by the system:
Because all the phones on which the line is configured ring simultaneously, they represent a load on
the system corresponding to a much higher calls-per-hour (CH) value than the CH of the line. The
corresponding bandwidth consumption is therefore increased. The network infrastructure's
bandwidth provisioning requires adjustments when WAN-connected shared line functionality is
deployed. The CH value employed for Equations 3 and 4 must be increased according to the
following formula:
CHS = CHL (Number line appearances) / (Number of lines)
Where CHS is the shared-line calls per hour to be used in Equations 3 and 4, and CHL is the
calls-per-hour rating of the line. For example, if a site is configured with 5 lines making an average
of 6 calls per hour but 2 of those lines are shared across 4 different phones, then:
Number of lines = 5
Number of line appearances = (2 lines appear on 4 phones, and 3 lines appear on only one
phone) = (24) + 3 = 11 line appearances
CHL = 6
CHS = 6 (11 / 5) = 13.2
Because each of the ringing phones requires a separate signaling control stream, the quantity of
packets sent from Unified CM to the same branch is increased in linear proportion to the quantity of
phones ringing. Because Unified CM is attached to the network through an interface that supports
100 Mbps or more, it can instantaneously generate a very large quantity of packets that must be
3-59
Chapter 3
Network Infrastructure
WAN Infrastructure
buffered while the queuing mechanism is servicing the signaling traffic. The servicing speed is
limited by the WAN interface's effective information transfer speed, which is typically two orders
of magnitude smaller than 100 Mbps.
This traffic may overwhelm the queue depth of the central site's WAN router. By default, the queue
depth available for each of the classes of traffic in Cisco IOS is 64. In order to prevent any packets
from being dropped before they are queued for the WAN interface, you must ensure that the
signaling queue's depth is sized to hold all the packets from at least one full shared-line event for
each shared-line phone. Avoiding drops is paramount in ensuring that the call does not create a race
condition where dropped packets are retransmitted, causing system response times to suffer.
Therefore, the quantity of packets required to operate shared-line phones is as follows:
SCCP protocol: 13 packets per shared-line phone
SIP protocol: 11 packets per shared-line phone
For example, with SCCP and with 6 phones sharing the same line, the queue depth for the signaling
class of traffic must be adjusted to a minimum of 78. Table 3-10 provides recommended queue
depths based on the quantity of shared line appearances within a branch site.
Table 3-10
SIP
65
55
10
130
110
15
195
165
20
260
220
25
325
275
When using a Layer 2 WAN technology such as Frame Relay, this adjustment must be made on the
circuit corresponding to the branch where the shared-line phones are located.
When using a Layer 3 WAN technology such as MPLS, there may be a single signaling queue
servicing multiple branches. In this case, adjustment must be made for the total of all branches
serviced.
3-60
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
Assuming an average call duration of 2 minutes and 100 percent utilization of each virtual tie line, we
can derive that each tie line carries a volume of 30 calls per hour. This assumption allows us to obtain
the following formula that expresses the recommended bandwidth for call control traffic as a function of
the number of virtual tie lines.
Equation 6: Recommended Bandwidth Based on Number of Virtual Tie Lines.
Recommended Bandwidth (bps) = 116 (Number of virtual tie lines)
If we take into account the fact that 8 kbps is the smallest bandwidth that can be assigned to a queue on
a Cisco IOS router, we can deduce that a minimum queue size of 8 kbps can accommodate the call
control traffic generated by up to 70 virtual tie lines or 2,100 calls per hour. This amount of 8 kbps for
SIP signaling traffic between clusters should be sufficient for most large enterprise deployments.
High Availability for Voice and Video over WLAN, page 3-65
Capacity Planning for Voice and Video over WLAN, page 3-67
Design Considerations for Voice and Video over WLAN, page 3-67
For more information about voice and video over wireless LANs, refer to the Real-Time Traffic over
Wireless LAN Solution Reference Network Design Guide, available at
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/RToWLAN/CCVP_BK_R7805F2
0_00_rtowlan-srnd.html
3-61
Chapter 3
Network Infrastructure
Figure 3-24
Active Directory
(LADP)
Cisco Wireless
LAN Controller
Router
Cisco
Access Point
Switch
CAPWAP
CAPWAP
Switch
Cisco
Access Point
Cisco Unified
CM
Dual-Mode
Smart Phones
Cisco
Wireless
IP Phone
Wireless IP
Software
Voice and
Video Client
Switch
IP
Mobile Collaboration
Enterprise Tablets
IP Phone
284266
3-62
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
points for the wireless infrastructure include ease of management, dynamic network tuning, and high
availability. However, if you are using the managed mode instead of the autonomous mode in the access
points, you need to consider the network tunneling effect of the LWAP-WLC communication
architecture when designing your solution. This network tunneling effect is discussed in more depth in
the section on Wireless LAN Controller Design Considerations, page 3-72.
Figure 3-25
CAPWAP
Wireless LAN
Controller
CAPWAP
Wireless
IP Phone
Wireless
Access Point
Wireless
IP Phone
284267
CAPWAP
Authentication Database
The authentication database is a core component of the wireless networks, and it holds the credentials
of the users to be authenticated while the wireless association is in progress. The authentication database
provides the network administrators with a centralized repository to validate the credentials. Network
administrators simply add the wireless network users to the authentication database instead of having to
add the users to all the wireless access points with which the wireless devices might associate.
In a typical wireless authentication scenario, the WLC couples with the authentication database to allow
the wireless association to proceed or fail. Authentication databases commonly used are LDAP and
RADIUS, although under some scenarios the WLC can also store a small user database locally that can
be used for authentication purposes.
3-63
Chapter 3
Network Infrastructure
Call Control
Cisco wireless endpoints require a call control or call processing server to route calls efficiently and to
provide a feature-rich experience for the end users. The call processing entity resides somewhere in the
wired network, either in the LAN or across a WAN.
Call control for the Cisco wireless endpoints is achieved through a call control protocol, either SIP or
SCCP.
Media Termination
Media termination on wired endpoints occurs when the end users of the wireless endpoints communicate
with IP phones, PSTN users, or video endpoints. Voice gateways, IP phones, video terminals, PBX
trunks, and transcoders all serve as termination points for media when a user communicates through
them. This media termination occurs by means of coding and decoding of the voice or video session for
the user communication.
3-64
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
3-65
Chapter 3
Network Infrastructure
Channel 1
Channel 140
Channel 11
Channel 36
Channel 1
Channel 52
2.4 GHz
GH
Hz ch
channel
hannel cells
5 GHz channel cells
Ch
Channel
hanne
el 6
Channel 120
Channel 6
Channel 108
Channel 11
1
Channel 64
Channel 11
Channel 56
Channel 1
Channel 100
348679
Figure 3-26
Careful deployment of APs and channel configuration within the wireless infrastructure are imperative
for proper wireless network operation. For this reason, Cisco requires customers to conduct a complete
and thorough site survey before deploying wireless networks in a production environment. The survey
should include verifying non-overlapping channel configurations, Wi-Fi channel coverage, and required
data and traffic rates; eliminating rogue APs; and identifying and mitigating the impact of potential
interference sources.
Additionally, evaluate utilizing a 5 GHz frequency band, which is generally less crowded and thus
usually less prone to interference. If Bluetooth is used then 5 GHz 802.11a is highly recommended.
Similarly, the usage of Cisco CleanAir technology will increase the WLAN reliability by detecting radio
frequency interference in real time and providing a self-healing and self-optimizing wireless network.
For further information about Cisco CleanAir technology, refer to the product documentation available
at
http://www.cisco.com/en/US/netsol/ns1070/index.html
For further information on how to provide high availability in a WLAN that supports rich media, refer
to the Real-Time Traffic over Wireless LAN Solution Reference Network Design Guide, available at
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/RToWLAN/CCVP_BK_R7805F2
0_00_rtowlan-srnd.html
3-66
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
3-67
Chapter 3
Network Infrastructure
VLANs
Just as with a wired LAN infrastructure, when deploying voice or video in a wireless LAN, you should
enable at least two virtual LANs (VLANs) at the Access Layer. The Access Layer in a wireless LAN
environment includes the access point (AP) and the first-hop access switch. On the AP and access switch,
you should configure both a native VLAN for data traffic and a voice VLAN (under Cisco IOS) or
Auxiliary VLAN (under CatOS) for voice traffic. This auxiliary voice VLAN should be separate from
all the other wired voice VLANs in the network. However, when the wireless clients (for example, smart
phones or software rich-media clients) do not support the concept of an auxiliary VLAN, alternative
packet marking strategies (for example, packet classification per port) must be applied to segregate the
important traffic such as voice and video and treat it with priority. When deploying a wireless
infrastructure, Cisco also recommends configuring a separate management VLAN for the management
of WLAN APs. This management VLAN should not have a WLAN appearance; that is, it should not
have an associated service set identifier (SSID) and it should not be directly accessible from the WLAN.
Roaming
To improve the user experience, Cisco recommends designing the cell boundary distribution with a 20%
to 30% overlap of non-adjacent channels to facilitate seamless roaming of the wireless client between
access points. Furthermore, when devices roam at Layer 3, they move from one AP to another AP across
native VLAN boundaries. When the WLAN infrastructure consists of autonomous APs, a Cisco Wireless
LAN Controller allows the Cisco Unified Wireless endpoints to keep their IP addresses and roam at
Layer 3 while still maintaining an active call. Seamless Layer 3 roaming occurs only when the client is
roaming within the same mobility group. For details about the Cisco Wireless LAN Controller and
Layer 3 roaming, refer to the product documentation available at
http://www.cisco.com/en/US/products/hw/wireless/index.html
Seamless Layer 3 roaming for clients across a lightweight access point infrastructure is accomplished
by WLAN controllers that use dynamic interface tunneling. Cisco Wireless Unified Communications
endpoints that roam across WLAN controllers and VLANs can keep their IP address when using the
same SSID and therefore can maintain an active call.
Note
In dual-band WLANs (those with 2.4 GHz and 5 GHz bands), it is possible to roam between 802.11b/g
and 802.11a with the same SSID, provided the client is capable of supporting both bands. However, this
can cause gaps in the voice path. If Cisco Unified Wireless IP Phones 7921 or 7925 are used, make sure
that firmware version 1.3(4) or higher is installed on the phones to avoid these gaps; otherwise use only
one band for voice. (The Cisco Unified Wireless IP Phone 7926 provides seamless inter-band roaming
from its first firmware version.)
Wireless Channels
Wireless endpoints and APs communicate by means of radios on particular channels. When
communicating on one channel, wireless endpoints typically are unaware of traffic and communication
occurring on other non-overlapping channels.
Optimal channel configuration for 2.4 GHz 802.11b/g/n requires a minimum of five-channel separation
between configured channels to prevent interference or overlap between channels. Non-overlapping
channels have 22 MHz of separation. Channel 1 is 2.412 GHz, channel 6 is 2.437 GHz, and channel 11
is 2.462 GHz. In North America, with allowable channels of 1 to 11, channels 1, 6, and 11 are the three
usable non-overlapping channels for APs and wireless endpoint devices. However, in Europe where the
allowable channels are 1 to 13, multiple combinations of five-channel separation are possible. Multiple
combinations of five-channel separation are also possible in Japan, where the allowable channels are 1
to 14.
3-68
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
Optimal channel configuration for 5 GHz 802.11a and 802.11n requires a minimum of one-channel
separation to prevent interference or overlap between channels. In North America, there are 20 possible
non-overlapping channels: 36, 40, 44, 48, 52, 56, 60, 64, 100, 104, 108, 112, 116, 132, 136, 140, 149,
153, 157, and 161. Europe and Japan allow 16 possible non-overlapping channels: 36, 40, 44, 48, 52, 56,
60, 64, 100, 104, 108, 112, 116, 132, 136, and 140. Because of the larger set of non-overlapping
channels, 802.11a and 5 Ghz 802.11n allow for more densely deployed WLANs; however, Cisco
recommends not enabling all channels but using a 12-channel design instead.
Note that the 802.11a and 802.11n bands (when using channels operating at 5.25 to 5.725 GHz, which
are 15 of the 24 possible channels) do require support for Dynamic Frequency Selection (DFS) and
Transmit Power Control (TPC) on some channels in order to avoid interference with radar (military,
satellite, and weather). Regulations require that channels 52 to 64, 100 to 116, and 132 to 140 support
DFS and TPC. TPC ensures that transmissions on these channels are not powerful enough to cause
interference. DFC monitors channels for radar pulses and, when it detects a radar pulse, DFC stops
transmission on the channel and switches to a new channel.
AP coverage should be deployed so that no (or minimal) overlap occurs between APs configured with
the same channel. Same- channel overlap should typically occur at 19 dBm of separation. However,
proper AP deployment and coverage on non-overlapping channels requires a minimum overlap of 20%.
This amount of overlap ensures smooth roaming for wireless endpoints as they move between AP
coverage cells. Overlap of less than 20% can result in slower roaming times and poor voice quality.
Deploying wireless devices in a multi-story building such as an office high-rise or hospital introduces a
third dimension to wireless AP and channel coverage planning. Both the 2.4 GHz and 5.0 GHz wave
forms of 802.11 can pass through floors and ceilings as well as walls. For this reason, not only is it
important to consider overlapping cells or channels on the same floor, but it is also necessary to consider
channel overlap between adjacent floors. With the 2.4 GHz wireless spectrum limited to only three
usable non-overlapping channels, proper overlap design can be achieved only through careful
three-dimensional planning.
Note
Careful deployment of APs and channel configuration within the wireless infrastructure are imperative
for proper wireless network operation. For this reason, Cisco requires that a complete and thorough site
survey be conducted before deploying wireless networks in a production environment. The survey
should include verifying non-overlapping channel configurations, AP coverage, and required data and
traffic rates; eliminating rogue APs; and identifying and mitigating the impact of potential interference
sources.
Wireless Interference and Multipath Distortion
Interference sources within a wireless environment can severely limit endpoint connectivity and channel
coverage. In addition, objects and obstructions can cause signal reflection and multipath distortion.
Multipath distortion occurs when traffic or signaling travels in more than one direction from the source
to the destination. Typically, some of the traffic arrives at the destination before the rest of the traffic,
which can result in delay and bit errors in some cases. You can reduce the effects of multipath distortion
by eliminating or reducing interference sources and obstructions, and by using diversity antennas so that
only a single antenna is receiving traffic at any one time. Interference sources should be identified during
the site survey and, if possible, eliminated. At the very least, interference impact should be alleviated by
proper AP placement and the use of location-appropriate directional or omni-directional diversity radio
antennas.
3-69
Chapter 3
Network Infrastructure
Other 2.4 GHz and 5 Ghz devices, such as 2.4 GHz cordless phones, personal wireless network
devices, sulphur plasma lighting systems, microwave ovens, rogue APs, and other WLAN
equipment that takes advantage of the license-free operation of the 2.4 GHz and 5 Ghz bands
Metal equipment, structures, and other metal or reflective surfaces such as metal I-beams, filing
cabinets, equipment racks, wire mesh or metallic walls, fire doors and fire walls, concrete, and
heating and air conditioning ducts
Because Bluetooth-enabled devices use the same 2.4 GHz radio band as 802.11b/g/n devices, it is
possible that Bluetooth and 802.11b/g/n devices can interfere with each other, thus resulting in
connectivity issues. Due to the potential for Bluetooth devices to interfere with and disrupt 802.11b/g/n
WLAN voice and video devices (resulting in poor voice quality, de-registration, call setup delays, and/or
reduce per-channel-cell call capacity), Cisco recommends, when possible, that you deploy all WLAN
voice and video devices on the 5 GHz Wi-Fi band using 802.11a and/or 802.11n protocols. By deploying
wireless clients on the 5 Ghz radio band, you can avoid interference caused by Bluetooth devices.
Additionally, Cisco CleanAir technology is recommended within the wireless infrastructure because it
enables real-time interference detection. For more information about Cisco CleanAir technology, refer
to the product documentation available at
http://www.cisco.com/en/US/netsol/ns1070/index.html
Note
802.11n can operate on both the 2.4 GHz and 5 GHz bands; however, Cisco recommends using 5 GHz
for Unified Communications.
Multicast on the WLAN
By design, multicast does not have the acknowledgement level of unicast. According to 802.11
specifications, the access point must buffer all multicast packets until the next Delivery Traffic Indicator
Message (DTIM) period is met. The DTIM period is a multiple of the beacon period. If the beacon period
is 100 ms (typical default) and the DTIM value is 2, then the access point must wait up to 200 ms before
transmitting a single buffered multicast packet. The time period between beacons (as a product of the
DTIM setting) is used by battery-powered devices to go into power save mode temporarily. This power
save mode helps the device conserve battery power.
Multicast on WLAN presents a twofold problem in which administrators must weigh multicast traffic
quality requirements against battery life requirements. First, delaying multicast packets will negatively
affect multicast traffic quality, especially for applications that multicast real-time traffic such as voice
and video. In order to limit the delay of multicast traffic, DTIM periods should typically be set to a value
of 1 so that the amount of time multicast packets are buffered is low enough to eliminate any perceptible
delay in multicast traffic delivery. However, when the DTIM period is set to a value of 1, the amount of
time that battery-powered WLAN devices are able to go into power save mode is shortened, and
therefore battery life is shortened. In order to conserve battery power and lengthen battery life, DTIM
periods should typically be set to a value of 2 or more.
For WLAN networks with no multicast applications or traffic, the DTIM period should be set to a value
of 2 or higher. For WLAN networks where multicast applications are present, the DTIM period should
be set to a value of 2 with a 100 ms beacon period whenever possible; however, if multicast traffic quality
3-70
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
suffers or if unacceptable delay occurs, then the DTIM value should be lowered to 1. If the DTIM value
is set to 1, administrators must keep in mind that battery life of battery-operated devices will be
shortened significantly.
Before enabling multicast applications on the wireless network, Cisco recommends testing these
applications to ensure that performance and behavior are acceptable.
For additional considerations with multicast traffic, see the chapter on Media Resources, page 7-1.
For recommends on deploying access points for wireless voice, refer to the documentation at
http://www.cisco.com/en/US/products/ps5678/Products_Sub_Category_Home.html.
AP Deployment
The number of devices active with an AP affects the amount of time each device has access to the
transport medium, the Wi-Fi channel. As the number of devices increases, the traffic contention
increases. Associating more devices to the AP and the bandwidth of the medium can result in poor
performance and slower response times for all the endpoint devices associated to the AP.
While there is no specific mechanism prior to Cisco Wireless LAN Controller release 7.2 to ensure that
only a limited number of devices are associated to a single AP, system administrators can manage
device-to-AP ratios by conducting periodic site surveys and analyzing user and device traffic patterns.
If additional devices and users are added to the network in a particular area, additional site surveys
should be conducted to determine whether additional APs are required to handle the number of endpoints
that need to access the network.
Additionally, APs that support Cisco CleanAir technology should be considered because they provide
the additional function of remote monitoring of the Wi-Fi channel.
AP Configuration
When deploying wireless voice, observe the following specific AP configuration requirements:
Assign a Service Set Identifier (SSID) to each VLAN configured on the AP.
SSIDs enable endpoints to select the wireless VLAN they will use for sending and receiving traffic.
These wireless VLANs and SSIDs map to wired VLANs. For voice endpoints, this mapping ensures
priority queuing treatment and access to the voice VLAN on the wired network.
3-71
Chapter 3
Network Infrastructure
Configure two QoS policies on the AP, and apply them to the VLANs and interfaces.
To ensure that voice traffic is given priority queuing treatment, configure a voice policy and a data
policy with default classifications for the respective VLANs. (See Interface Queuing, page 3-75, for
more information).
3-72
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
Figure 3-27
CAPWAP
Wireless LAN
Controller
Switch
Switch
Switch
Wireless
Access Point
Mobile Collaboration
Enterprise Tablet
CAPWAP
Wireless
Access Point
Mobile Collaboration
Enterprise Tablets
CAPWAP
Wireless
Access Point
Mobile Collaboration
Enterprise Tablet
284268
CAPWAP
Traffic Flow
Additionally, the switch interface and switch platform egress buffer levels should match the maximum
combined burst you plan to support in your wireless network.
Failure to select adequate buffer levels could lead to packet drops and severely affect the user experience
of video over a wireless LAN, while lack of bandwidth coverage would cause packets to be queued and
in extreme cases cause delayed packets
3-73
Chapter 3
Network Infrastructure
Traffic Classification
As with the wired network infrastructure, it is important to classify or mark pertinent wireless traffic as
close to the edge of the network as possible. Because traffic marking is an entrance criterion for queuing
schemes throughout the wired and wireless network, marking should be done at the wireless endpoint
device whenever possible. Marking or classification by wireless network devices should be identical to
that for wired network devices, as indicated in Table 3-11.
In accordance with traffic classification guidelines for wired networks, the Cisco wireless endpoints
mark voice media traffic or voice RTP traffic with DSCP 46 (or PHB EF), video media traffic or video
RTP traffic with DSCP 34 (or PHB AF41), and call control signaling traffic (SCCP or SIP) with
DSCP 24 (or PHB CS3). Once this traffic is marked, it can be given priority or better than best-effort
treatment and queuing throughout the network. All wireless voice and video devices that are capable of
marking traffic should do it in this manner. All other traffic on the wireless network should be marked
as best-effort or with some intermediary classification as outlined in wired network marking guidelines.
If the wireless voice or video devices are unable to do packet marking, alternate methods such as
port-based marking should be implemented to provide priority to video and voice traffic.
Traffic Type
DSCP (PHB)
802.1p UP
IEEE 802.11e UP
Voice
46 (EF)
Video
34 (AF41)
For further information about 802.11e and its configuration, refer to your corresponding product
documentation available at
http://www.cisco.com/en/US/products/ps6302/Products_Sub_Category_Home.html
3-74
Chapter 3
Network Infrastructure
Wireless LAN Infrastructure
Interface Queuing
Once traffic marking has occurred, it is necessary to enable the wired network APs and devices to
provide QoS queuing so that voice and video traffic types are given separate queues to reduce the
chances of this traffic being dropped or delayed as it traverses the wireless LAN. Queuing on the wireless
network occurs in two directions, upstream and downstream. Upstream queuing concerns traffic
traveling from the wireless endpoint up to the AP, and from the AP up to the wired network. Downstream
queuing concerns traffic traveling from the wired network to the AP and down to the wireless endpoint.
For upstream queuing, devices that support Wi-Fi Multimedia (WMM) are able to take advantage of
queueing mechanisms, including priority queueing.
As for downstream QoS, Cisco APs currently provide up to eight queues for downstream traffic being
sent to wireless clients. The entrance criterion for these queues can be based on a number of factors,
including DSCP, access control lists (ACLs), and VLAN. Although eight queues are available, Cisco
recommends using only two queues when deploying wireless voice. All voice media and signaling traffic
should be placed in the highest-priority queue, and all other traffic should be placed in the best-effort
queue. This ensures the best possible queuing treatment for voice traffic.
In order to set up this two-queue configuration for autonomous APs, create two QoS policies on the AP.
Name one policy Voice, and configure it with the class of service Voice < 10 ms Latency (6) as the
Default Classification for all packets on the VLAN. Name the other policy Data, and configure it with
the class of service Best Effort (0) as the Default Classification for all packets on the VLAN. Then
assign the Data policy to the incoming and outgoing radio interface for the data VLAN(s), and assign
the Voice policy to the incoming and outgoing radio interfaces for the voice VLAN(s). With the QoS
policies applied at the VLAN level, the AP is not forced to examine every packet coming in or going out
to determine the type of queuing the packet should receive.
For lightweight APs, the WLAN controller has built-in QoS profiles that can provide the same queuing
policy. Voice VLAN or voice traffic is configured to use the Platinum policy, which sets priority
queueing for the voice queue. Data VLAN or data traffic is configured to use the Silver policy, which
sets best-effort queuing for the Data queue. These policies are then assigned to the incoming and
outgoing radio interfaces based on the VLAN.
The above configurations ensure that all voice and video media and signaling are given priority queuing
treatment in a downstream direction.
Note
Because Wi-Fi Multimedia (WMM) access is based on Enhanced Distributed Channel Access (EDCA),
it is important to assign the right priorities to the traffic to avoid Arbitration Inter-Frame Space (AIFS)
alteration and delivery delay. For further information on Cisco Unified Wireless QoS, refer to the latest
version of the Enterprise Mobility Design Guide, available at
http://www.cisco.com/en/US/netsol/ns820/networking_solutions_design_guidances_list.html.
3-75
Chapter 3
Network Infrastructure
Note
Currently there is no call admission control support for video. The QoS Basic Service Set (QBSS)
information element is sent by the AP only if QoS Element for Wireless Phones has been enable on the
AP. (Refer to Wireless AP Configuration and Design, page 3-71.)
3-76
CH A P T E R
Securing the various components in a Cisco Collaboration Solution is necessary for protecting the
integrity and confidentiality of voice and video calls.
This chapter presents security guidelines pertaining specifically to collaboration applications and the
voice and video network. For more information on data network security, refer to the Cisco SAFE
Blueprint documentation available at
http://www.cisco.com/en/US/netsol/ns744/networking_solutions_program_home.html
Following the guidelines in this chapter does not guarantee a secure environment, nor will it prevent all
penetration attacks on a network. You can achieve reasonable security by establishing a good security
policy, following that security policy, staying up-to-date on the latest developments in the hacker and
security communities, and maintaining and monitoring all systems with sound system administration
practices.
This chapter addresses centralized and distributed call processing, including clustering over the WAN
but not local failover mechanisms such as Survivable Remote Site Telephony (SRST). This chapter
assumes that all remote sites have a redundant link to the head-end or local call-processing backup in
case of head-end failure. The interaction between Network Address Translation (NAT) and IP Telephony,
for the most part, is not addressed here. This chapter also assumes that all networks are privately
addressed and do not contain overlapping IP addresses.
Described in:
Revision Date
4-1
Chapter 4
General Security
General Security
This section covers general security features and practices that can be used to protect the voice data
within a network.
Security Policy
Cisco Systems recommends creating a security policy associated with every network technology
deployed within your enterprise. The security policy defines which data in your network is sensitive so
that it can be protected properly when transported throughout the network. Having this security policy
helps you define the security levels required for the types of data traffic that are on your network. Each
type of data may or may not require its own security policy.
If no security policy exists for data on the company network, you should create one before enabling any
of the security recommendations in this chapter. Without a security policy, it is difficult to ascertain
whether the security that is enabled in a network is doing what it is designed to accomplish. Without a
security policy, there is also no systematic way of enabling security for all the applications and types of
data that run in a network.
Note
While it is important to adhere to the security guidelines and recommendations presented in this chapter,
they alone are not sufficient to constitute a security policy for your company. You must define a corporate
security policy before implementing any security technology.
This chapter details the features and functionality of a Cisco Systems network that are available to
protect the Unified Communications data on a network. It is up to the security policy to define which
data to protect, how much protection is needed for that type of data, and which security techniques to
use to provide that protection.
One of the more difficult issues with a security policy that includes voice and video traffic is combining
the security policies that usually exist for both the data network and the traditional voice network. Ensure
that all aspects of the integration of the media onto the network are secured at the correct level for your
security policy or corporate environment.
The basis of a good security policy is defining how important your data is within the network. Once you
have ranked the data according to its importance, you can decide how the security levels should be
established for each type of data. You can then achieve the correct level of security by using both the
network and application features.
In summary, you can use the following process to define a security policy:
4-2
Chapter 4
Security in Layers
This chapter starts with hardening the IP phone endpoints in a Cisco Unified Communications Solution
and works its way through the network from the phone to the access switch, to the distribution layer, into
the core, and then into the data center. (See Figure 4-1.) Cisco recommends building layer upon layer of
security, starting at the access port into the network itself. This design approach gives a network architect
the ability to place the devices where it is both physically and logically easy to deploy Cisco Unified
Communications applications. But with this ease of deployment, the security complexity increases
because the devices can be placed anywhere in a network as long as they have connectivity.
Figure 4-1
Layers of Security
Unified CM Cluster
M
M
Core
Distribution
Si
Si
Access
IP
IP
IP
148489
IP
4-3
Chapter 4
General Security
Secure Infrastructure
As the IP Telephony data crosses a network, that data is only as safe and secure as the devices that are
transporting the data. Depending on the security level that is defined in your security policy, the security
of the network devices might have to be improved or they might already be secure enough for the
transportation of IP Telephony traffic.
There are many best practices within a data network that, if used, will increase the entire security of your
network. For example, instead of using Telnet (which sends passwords in clear text) to connect to any of
the network devices, use Secure Shell (SSH, the secure form of Telnet) so that an attacker would not be
able to see a password in clear text.
Cisco Routers configured as gateways, Cisco Unified Border Element, and media resources can be
configured with Cisco IOS feature sets that provide the required media functionality but support only
Telnet and not Secure Shell (SSH). Cisco recommends that you use access control lists (ACLs) to control
who is permitted to connect to the routers using Telnet. It is more secure to connect to the gatekeeper
from a host that is in a secure segment of the network, because user names and passwords are sent over
Telnet in clear text.
You should also use firewalls, access control lists, authentication services, and other Cisco security tools
to help protect these devices from unauthorized access.
Physical Security
Just as a traditional PBX is usually locked in a secure environment, the IP network should be treated in
a similar way. Each of the devices that carries media traffic is really part of an IP PBX, and normal
general security practices should be used to control access to those devices. Once a user or attacker has
physical access to one of the devices in a network, all kinds of problems could occur. Even if you have
excellent password security and the user or attacker cannot get into the network device, that does not
mean that they cannot cause havoc in a network by simply unplugging the device and stopping all traffic.
For more information on general security practices, refer to the documentation at the following
locations:
http://www.cisco.com/en/US/netsol/ns744/networking_solutions_program_home.html
http://www.cisco.com/en/US/products/svcs/ps2961/ps2952/serv_group_home.html
IP Addressing
IP addressing can be critical for controlling the data that flows in and out of the logically separated IP
Telephony network. The more defined the IP addressing is within a network, the easier it becomes to
control the devices on the network.
As stated in other sections of this document (see Campus Access Layer, page 3-4), you should use IP
addressing based on RFC 1918. This method of addressing allows deployment of an IP Telephony
system into a network without redoing the IP addressing of the network. Using RFC 1918 also allows
for better control in the network because the IP addresses of the voice endpoints are well defined and
easy to understand. If the voice and video endpoints are all addressed within a 10.x.x.x. network, access
control lists (ACLs) and tracking of data to and from those devices are simplified.
If you have a well defined IP addressing plan for your voice deployments, it becomes easier to write
ACLs for controlling the IP Telephony traffic and it also helps with firewall deployments.
4-4
Chapter 4
Using RFC 1918 enables you easily to deploy one VLAN per switch, which is a best practice for campus
design, and also enables you to keep the Voice VLAN free of any Spanning Tree Protocol (STP) loops.
If deployed correctly, route summarization could help to keep the routing table about the same as before
the voice and video deployment, or just slightly larger.
IPv6 Addressing
The introduction of IPv6 addressing has extended the network address space and increased the options
for privacy and security of endpoints. Though both IPv4 and IPv6 have similar security concerns, IPv6
provides some advantages. For example, one of the major benefits with IPv6 is the enormous size of the
subnets, which discourages automated scanning and reconnaissance attacks.
When considering IPv6 as your IP addressing method, adhere to the best practices documented in the
following campus and branch office design guides:
Access Security
This section covers security features at the Access level that can be used to protect the voice and data
within a network.
4-5
Chapter 4
Access Security
support for voice endpoints. LLDP-MED defines how a switch port transitions from LLDP to
LLDP-MED if it detects an LLDP-MED-capable endpoint. Support for both LLDP and LLDP-MED on
IP phones and LAN switches depends on the firmware and device models. To determine if LLDP-MED
is supported on particular phone or switch models, check the specific product release notes or bulletins
available at:
Note
http://www.cisco.com/en/US/products/hw/phones/ps379/prod_release_notes_list.html
http://www.cisco.com/en/US/products/sw/iosswrel/ps5012/prod_bulletins_list.html
If an IP phone with LLDP-MED capability is connected to a Cisco Catalyst switch running an earlier
Cisco IOS release that does not support LLDP, the switch might indicate that an extra device has been
connected to the switch port. This can happen if the Cisco Catalyst switch is using Port Security to count
the number of devices connected. The appearance of an LLDP packet might cause the port count to
increase and cause the switch to disable the port. Verify that your Cisco Catalyst switch supports LLDP,
or increase the port count to a minimum of three, before deploying Cisco IP Phones with firmware that
supports LLDP-MED Link Layer protocol.
H.323 clients, Multipoint Control Units (MCUs), and gateways communicate with Unified CM using the
H.323 protocol. Unified CM H.323 trunks (such as H.225 and intercluster trunk variants as well as the
RASAggregator trunk type) use a random port range rather than the well-known TCP port 1720.
Therefore, you must permit a wide range of TCP ports between these devices and the Unified CM
servers. For port usage details, refer to the latest version of the Cisco Unified Communications Manager
TCP and UDP Port Usage guide, available at:
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
MCUs and gateways are considered infrastructure devices, and they typically reside within the
datacenter adjacent to the Unified CM servers. H.323 clients, on the other hand, typically reside in the
data VLAN.
Cisco TelePresence MCUs configured to run in SCCP mode communicate with the TFTP server(s) to
download their configuration, with the Unified CM servers for signaling, and with other endpoints for
RTP media traffic. Therefore, TFTP must be permitted between the MCU and the TFTP server(s), TCP
port 2000 must be permitted between the MCUs and the Unified CM server(s), and UDP ports for RTP
media must be permitted between the MCUs and the voice, data, and gateway VLANs.
Switch Port
There are many security features within a Cisco switch infrastructure that can be used to secure a data
network. This section describes some of the features that can be used in Cisco Access Switches to protect
the IP Telephony data within a network. (See Figure 4-2.) This section does not cover all of the security
features available for all of the current Cisco switches, but it does list the most common security features
used across many of the switches that Cisco manufactures. For additional information on the security
features available on the particular Cisco gear deployed within your network, refer to the appropriate
product documentation available at
http://www.cisco.com
4-6
Chapter 4
Figure 4-2
Access
IP
IP
IP
148493
IP
4-7
Chapter 4
Access Security
for every port. However, with dynamic port security, network administrators can merely specify the
number of MAC addresses they would like the switch to learn and, assuming the correct devices are the
first devices to connect to the port, allow only those devices access to that port for some period of time.
The period of time can be determined by either a fixed timer or an inactivity timer (non-persistent
access), or it can be permanently assigned. In the latter case, the MAC address learned will remain on
the port even in the event of a reload or reboot of the switch.
No provision is made for device mobility by static port security or persistent dynamic port security.
Although it is not the primary requirement, MAC flooding attacks are implicitly prevented by port
security configurations that aim to limit access to certain MAC addresses.
From a security perspective, there are better mechanisms for both authenticating and authorizing port
access based on userid and/or password credentials rather than using MAC address authorization. MAC
addresses alone can easily be spoofed or falsified by most operating systems.
148494
If the number of MAC addresses is not defined correctly, there is a possibility of denying access to the
network or error-disabling the port and removing all devices from the network.
4-8
Chapter 4
to the phones, you should use the DHCP Snooping feature in the switches to secure DHCP messaging.
Rogue DHCP servers can attempt to respond to the broadcast messages from a client to give out incorrect
IP addresses, or they can attempt to confuse the client that is requesting an address.
When enabled, DHCP Snooping treats all ports in a VLAN as untrusted by default. An untrusted port is
a user-facing port that should never make any reserved DHCP responses. If an untrusted DHCP-snooping
port makes a DHCP server response, it will be blocked from responding. Therefore, rogue DHCP servers
will be prevented from responding. However, legitimately attached DHCP servers or uplinks to
legitimate servers must be trusted.
Figure 4-4 illustrates the normal operation of a network-attached device that requests an IP address from
the DHCP server.
Figure 4-4
IP
DHCP Discover (Broadcast)
148894
However, an attacker can request not just a single IP address but all of the IP addresses that are available
within a VLAN. (See Figure 4-5.) This means that there would be no addresses for a legitimate device
trying to get on the network, and without an IP address the phone cannot connect to Unified CM.
4-9
Chapter 4
Access Security
Figure 4-5
DHCP
Server
Client
Denial of Service
Gobbler
148895
Untrusted
Rogue server
Bad DHCP
responses:
offer, ack, nak
Untrusted
Trusted
OK DHCP
responses:
offer, ack, nak
148495
Figure 4-6
DHCP Snooping prevents any single device from capturing all the IP addresses in any given scope, but
incorrect configurations of this feature can deny IP addresses to approved users.
4-10
Chapter 4
4-11
Chapter 4
Access Security
Dynamic ARP Inspection (DAI) is used to inspect all ARP requests and replies (gratuitous or
non-gratuitous) coming from untrusted (or user-facing) ports to ensure that they belong to the ARP
owner. The ARP owner is the port that has a DHCP binding which matches the IP address contained in
the ARP reply. ARP packets from a DAI trusted port are not inspected and are bridged to their respective
VLANs.
Using DAI
Dynamic ARP Inspection (DAI) requires that a DHCP binding be present to legitimize ARP responses
or Gratuitous ARP messages. If a host does not use DHCP to obtain its address, it must either be trusted
or an ARP inspection access control list (ACL) must be created to map the host's IP and MAC address.
(See Figure 4-7.) Like DHCP Snooping, DAI is enabled per VLAN, with all ports defined as untrusted
by default. To leverage the binding information from DHCP Snooping, DAI requires that DHCP
Snooping be enabled on the VLAN prior to enabling DAI. If DHCP Snooping is not enabled before you
enable DAI, none of the devices in that VLAN will be able to use ARP to connect to any other device in
their VLAN, including the default gateway. The result will be a self-imposed denial of service to any
device in that VLAN.
Figure 4-7
ARP 10.1.1.1
saying
10.1.1.2 is MAC C
ARP 10.1.1.2
saying
10.1.1.1 is MAC C
10.1.1.2
MAC B
148496
10.1.1.3
MAC C
None matching
ARPs in the
bit bucket
10.1.1.1
MAC A
Because of the importance of the DHCP Snooping binding table to the use of DAI, it is important to back
up the binding table. The DHCP Snooping binding table can be backed up to bootflash, File Transfer
Protocol (FTP), Remote Copy Protocol (RCP), slot0, and Trivial File Transfer Protocol (TFTP). If the
DHCP Snooping binding table is not backed up, the Cisco Unified IP Phones could lose contact with the
default gateway during a switch reboot. For example, assume that the DHCP Snooping binding table is
not backed up and that you are using Cisco Unified IP Phones with a power adapter instead of line
power. When the switch comes back up after a reboot, there will be no DHCP Snooping binding table
entry for the phone, and the phone will not be able to communicate with the default gateway unless the
DHCP Snooping binding table is backed up and loads the old information before traffic starts to flow
from the phone.
Incorrect configurations of this feature can deny network access to approved users. If a device has no
entry in the DHCP Snooping binding table, then that device will not be able to use ARP to connect to
the default gateway and therefore will not be able to send traffic. If you use static IP addresses, those
addresses will have to be entered manually into the DHCP Snooping binding table. If you have devices
that do not use DHCP again to obtain their IP addresses when a link goes down (some UNIX or Linux
machines behave this way), then you must back up the DHCP Snooping binding table.
4-12
Chapter 4
Note
Cisco recommends authenticating the IP phone before the attached data device is authenticated.
The multiple-authentication mode assigns authenticated devices to either a data or voice VLAN,
depending on the attributes received from the authentication server when access is approved. The 802.1X
port is divided into a data domain and a voice domain.
In multiple-authentication mode, a guest VLAN can be enabled on the 802.1x port. The switch assigns
end clients to a guest VLAN when the authentication server does not receive a response to its EAPOL
identity frame or when EAPOL packets are not sent by the client. This allows data devices attached to a
Cisco IP Phone, that do not support 802.1X, to be connected to the network.
A voice VLAN must be configured for the IP phone when the switch port is in a multiple-host mode.
The RADIUS server must be configured to send a Cisco Attribute-Value (AV) pair attribute with a value
of device-traffic-class=voice. Without this value, the switch treats the IP phone as a data device.
Dynamic VLAN assignment from a RADIUS server is supported only for data devices.
When a data or a voice device is detected on a port, its MAC address is blocked until authorization
succeeds. If the authorization fails, the MAC address remains blocked for 5 minutes.
When the 802.1x authentication is enabled on an access port on which a voice VLAN is configured and
to which a Cisco IP Phone is already connected, the phone loses connectivity to the switch for up to
30 seconds.
Most Cisco IP Phones support authentication by means of X.509 certificates using the EAP-Transport
Layer Security (EAP-TLS) or EAP-Flexible Authentication with Secure Tunneling (EAP-FAST)
methods of authentication. Some of the older models that do not support either method can be
authenticated using MAC Authentication Bypass (MAB), which enables a Cisco Catalyst Switch to
check the MAC address of the connecting device as the method of authentication.
To determine support for the 802.1X feature configuration, refer to the product guides for the Cisco
Unified IP Phones and the Cisco Catalyst Switches, available at http://www.cisco.com.
For configuration information, refer to the IP Telephony for 802.1x Design Guide, available at
http://www.cisco.com/en/US/docs/solutions/Enterprise/Security/TrustSec_1.99/IP_Tele/IP_Teleph
ony_DIG.html
4-13
Chapter 4
Endpoint Security
Endpoint Security
Cisco Unified IP Phones contain built-in features to increase security on an IP Telephony network.
These features can be enabled or disabled on a phone-by-phone basis to increase the security of an IP
Telephony deployment. Depending on the placement of the phones, a security policy will help determine
if these features need to be enabled and where they should be enabled. (See Figure 4-8.)
Figure 4-8
Access
IP
IP
IP
148490
IP
Before attempting to configure the security features on a phone, check the documentation at the
following link to make sure the features are available on that particular phone model:
http://www.cisco.com/en/US/products/sw/voicesw/index.html
4-14
Chapter 4
Unified Communications Manager (Unified CM) can disable the PC port on the back of the phone.
Before attempting to enable this feature, check the documentation at the following link to verify that this
features is supported on your particular model of Cisco Unified IP Phone:
http://www.cisco.com/en/US/products/hw/phones/ps379/tsd_products_support_series_home.html
PC sends data
tagged with 802.1q
as Voice VLAN 20 or
the PC sends any data
tagged with 802.1q,
and it is dropped.
Data VLAN 10
Voice VLAN 20
148893
IP
4-15
Chapter 4
Endpoint Security
Settings Access
Each Cisco Unified IP Phone has a network settings page that lists many of the network elements and
detailed information that is needed for the phone to operate. This information could be used by an
attacker to start a reconnaissance on the network with some of the information that is displayed on the
phone's web page. For example, an attacker could look at the settings page to determine the default
gateway, the TFTP server, and the Unified CM IP address. Each of these pieces of information could be
used to gain access to the voice network or to attack a device in the voice network.
This access can be disabled on individual phones or by using bulk management to prevent end users or
attackers from obtaining the additional information such as Unified CM IP address and TFTP server
information. With access to the phone settings page disabled, end users lose the ability to change many
of the settings on the phone that they would normally be able to control, such as speaker volume,
contrast, and ring type. It might not be practical to use this security feature because of the limitations it
places on end users with respect to the phone interface. The settings access can also be set as restricted,
which prevents access to network configuration information but allows users to configure volume, ring
tones, and so forth.
For more information on the phone settings page, refer to the latest version of the Cisco Unified
Communications Manager Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
4-16
Chapter 4
Administrative passwords
Device access
Cisco TelePresence endpoints support management through Secure Shell (SSH) and Hyper-Text
Transfer Protocol over Secure Sockets Layer (HTTPS). Access to the endpoints using HTTP, HTTPS,
SSH, or Telnet can be configured in the Network Services setting on the endpoint itself.
The endpoints ship with default administrative passwords, and Cisco recommends changing the
passwords at the time of installation. Access to management functions should be restricted to authorized
users with administrative privileges. If the default administrative passwords are used, then the video
stream can be viewed by anyone accessing the administrative page with the password.
The endpoints can be assigned to users who are given access based on defined roles and privileges.
Passwords and PINs can be specified for those users to enable SSH or Telnet and web-based access. A
credential management policy should be implemented to expire and change passwords periodically and
to time-out logins when idle. This is necessary for limiting access to the devices to verified users.
The Transport Layer Security (TLS) protocol is designed to provide authentication, data integrity, and
confidentiality for communications between two applications. TLS is based on Secure Sockets Layer
(SSL) version 3.0, although the two protocols are not compatible. TLS operates in a client/server mode
with one side acting as the "server" and the other side acting as the "client." TLS requires TCP as the
reliable transport layer protocol to operate over.
Cisco Collaboration devices use TLS to secure SIP or SCCP signaling in the following scenarios:
Between Cisco TelePresence Management Suite (TMS), Unified CM, and/or Cisco TelePresence
Video Communication Server (VCS)
Secure RTP (SRTP), defined in IETF RFC 3711, details the methods of providing confidentiality and
data integrity for both Real-time Transport Protocol (RTP) voice and video media, as well as their
corresponding Real-time Transport Control Protocol (RTCP) streams. SRTP accomplishes this through
the use of encryption and message authentication headers.
In SRTP, encryption applies only to the payload of the RTP packet. Message authentication, however, is
applied to both the RTP header and the RTP payload. Because message authentication applies to the RTP
sequence number within the header, SRTP indirectly provides protection against replay attacks as well.
4-17
Chapter 4
Endpoint Security
SRTP uses Advanced Encryption Standards (AES) with a 128-bit encryption key as the encryption
cipher. It also uses Hash-based Message Authentication Code Secure Hash Algorithm-1 (HMAC-SHA1)
as the authentication method.
Voice and Video System
Unified CM can be configured to provide multiple levels of security to the phones within a voice system,
if those phones support those features. This includes device authentication and media and signaling
encryption using X.509 certificates. Depending on your security policy, phone placement, and phone
support, the security can be configured to fit the needs of your company.
For information on which Cisco Unified IP Phone models support specific security features, refer to the
documentation available at
http://www.cisco.com/en/US/products/hw/phones/ps379/tsd_products_support_series_home.html
To enable security on the phones and in the Unified CM cluster, refer to the Cisco Unified
Communications Manager Security Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
When the Public Key Infrastructure (PKI) security features are properly configured in Unified CM, all
supported phones will have the following capabilities:
Integrity Does not allow TFTP file manipulation but does allow Transport Layer Security (TLS)
signaling to the phones when enabled.
Authentication The image for the phone is authenticated from Unified CM to the phone, and the
device (phone) is authenticated to Unified CM. All signaling messages between the phone and
Unified CM are verified as being sent from the authorized device.
Encryption For supported devices, signaling and media can be encrypted to prevent
eavesdropping.
Secure Real-time Transport Protocol (SRTP) Is supported to Cisco IOS gateways and on
phone-to-phone communications. Cisco Unity also supports SRTP for voicemail.
Unified CM supports authentication, integrity, and encryption for calls between two
Cisco Unified IP Phones but not for all devices or phones. To determine if your device supports these
features, refer to the documentation available at
http://www.cisco.com/en/US/products/hw/phones/ps379/tsd_products_support_series_home.html
Unified CM uses certificates for securing identities and enabling encryption. The certificates can be
either Manufacturing Installed Certificates (MIC) or Locally Significant Certificates (LSC). MICs are
already pre-installed and LSCs are installed by Unified CM's Cisco Certificate Authority Proxy Function
(CAPF). Unified CM creates self-signed certificates, but signing of certificates by a third-party
certificate authority (CA) using PKCS #10 Certificate Signing Request (CSR) is also supported. When
using third-party CAs, the CAPF can be signed by the CA, but the phone LSCs are still generated by the
CAPF. When MICs are used, the Cisco CA and the Cisco Manufacturing CA certificates act as the root
certificates. When LSCs are generated for natively registered endpoints, the CAPF certificate is the root
certificate.
Auto-registration does not work if you configure the cluster for mixed mode, which is required for device
authentication. The cluster mixed-mode information is included in the CTL file downloaded by the
endpoints. The CTL file configuration requires using a CTL client to sign the file. The CTL client is a
separate application that is installed on a Windows PC, and it uses the Cisco Security Administrator
Security Token (SAST), USB hardware device, to sign the CTL file.
Cisco TelePresence Management Suite (TMS) provides TLS certificates to verify its identity when
generating outbound connections.
4-18
Chapter 4
Application layer protocol inspection and Application Layer Gateways (ALGs) that allow IP Telephony
traffic to traverse firewalls and Network Address Translation (NAT) also do not work with signaling
encryption. Not all gateways, phones, or conference are supported with encrypted media.
Encrypting media makes recording and monitoring of calls more difficult and expensive. It also makes
troubleshooting VoIP problems more challenging.
Third-Party CA Certificates
The certificates generated by default are Cisco Unified CM self-signed certificates. In a deployment
where a third-party CA is implemented, Certificate Signing Requests (CSR) can be used for establishing
the third-party CA as the root CA for the Cisco Unified devices. This requires that both the signed
application certificate and the CA root certificate from the CA or PKCS#7 Certificate Chain (DER
format) containing both the application certificate and CA certificates must be obtained.
The CAPF LSCs used by the Cisco endpoints are locally signed. However, third-party CA signed LSCs
are also supported. Implementing such support involves importing the third-party CA certificate into the
Unified CM trust store and configuring Unified CM's CAPF service to use the off-system CA as the
certificate issuer for the endpoints.
Multiserver Certificates
Cisco Unified Communications operating system supports generating multiserver certificates with
Subject Alternative Names (SAN) extensions for the Tomcat service, CallManager service, and IM and
Presence Service. This functionality adds support for a single SAN certificate per application service
(Tomcat, XMPP, CallManager) across multiple nodes in a cluster. The multiserver certificate can contain
multiple FQDNs or domains present in SAN extensions.
Implementation of a multiserver certificate requires using a third-party CA in the deployment. The
multiserver Certificate Signing Request (CSR) can be generated on any server, and the corresponding
Certificate can be uploaded from any other server in the cluster. The parent domain field and the
Common Name (CN) fields are editable during the CSR generation process for both the single and
multiserver CSR.
Common Criteria Requirements
Elliptical Curve Cryptography (ECC) support for Cisco Unified Communications Manager certificates
has been introduced in version 11.x. Both self-signed and CA signed certificates, CTL and ITL files, SIP
signaling, and bulk certificate management functions can be configured to support ECC.
The certificates generated using ECC are required to have a common name with the EC suffix to
differentiate them from the default certificates. The multiserver certificates use the EC-ms suffix. These
certificates can use a keypair of 256, 384, or 521 bits and hash algorithms SHA256, SHA384, and
SHA521. The CallManager and TFTP certificates generated with ECC are stored as
callmanager-ECDSA in the Unified Communications Manager Trust Store.
The above mentioned EC key sizes are also supported on CAPF options for phone security profiles. The
CAPF function can use one of the Elliptical Curve key pair sizes to generate an ECDSA LSC certificate
for the endpoints. But not all phone models support these key sizes. In cases where it is not supported
on all endpoints, it is preferable to use EC preferred and RSA backup in the security profile.
4-19
Chapter 4
Endpoint Security
Quality of Service
Quality of Service (QoS) is a vital part of any security policy for an enterprise network. Even though
most people think of QoS as setting the priority of traffic in a network, it also controls the amount of
data that is allowed into the network. In the case of Cisco switches, that control point is at the port level
when the data comes from the phone to the Ethernet switch. The more control applied at the edge of the
network at the access port, the fewer problems will be encountered as the data aggregates in the network.
QoS can be used to control not only the priority of the traffic in the network but also the amount of traffic
that can travel through any specific interface. Cisco Smartports templates have been created to assist in
deploying voice QoS in a network at the access port level.
A rigorous QoS policy can control and prevent denial-of-service attacks in the network by throttling
traffic rates.
As mentioned previously in the lobby phone example, Cisco recommends that you provide enough flow
control of the traffic at the access port level to prevent any attacker from launching a denial-of-service
(DoS) attack from that port in the lobby. The configuration for that example was not as aggressive as it
could be because the QoS configuration allowed traffic sent to the port to exceed the maximum rate, but
the traffic was remarked to the level of scavenger class. Given a more aggressive QoS policy, any amount
4-20
Chapter 4
of traffic that exceeded that maximum limit of the policy could just be dropped at the port, and that
"unknown" traffic would never make it into the network. QoS should be enabled across the entire
network to give the IP Telephony data high priority from end to end.
For more information on QoS, refer to the chapter on Network Infrastructure, page 3-1, and the QoS
design guides available at
http://www.cisco.com/c/en/us/solutions/enterprise/design-zone-ipv6/design-guide-listing.html
4-21
Chapter 4
Firewalls
Figure 4-10
Distribution
Si
Si
Unified CM Cluster
Access
IP
IP
IP
148498
IP
There are many types of ACLs that can be deployed at Layer 3. For descriptions and examples of the
most common types, refer to Configuring Commonly Used IP ACLs, available (with Cisco partner login
required) at
http://cisco.com/en/US/partner/tech/tk648/tk361/technologies_configuration_example09186a0080
100548.shtml
Depending on your security policy, the Layer 3 ACLs can be as simple as not allowing IP traffic from
the non-voice VLANS to access the voice gateway in the network, or the ACLs can be detailed enough
to control the individual ports and the time of the day that are used by other devices to communicate to
IP Telephony devices. As the ACLs become more granular and detailed, any changes in port usage in a
network could break not only voice but also other applications in the network.
If there are software phones in the network, if web access to the phone is allowed, or if you use the
Attendant Console or other applications that need access to the voice VLAN subnets, the ACLs are much
more difficult to deploy and control.
For IP phones restricted to specific subnets and limited to a voice VLAN, ACLs can be written to block
all traffic (by IP address or IP range) to Unified CMs, voice gateways, phones, and any other voice
application that is being used for voice-only services. This method simplifies the ACLs at Layer 3
compared to the ACLs at Layer 2 or VLAN ACLs.
Firewalls
Firewalls can be used in conjunction with ACLs to protect the voice servers and the voice gateways from
devices that are not allowed to communicate with IP Telephony devices. Because of the dynamic nature
of the ports used by IP Telephony, having a firewall does help to control opening up a large range of ports
needed for IP Telephony communications. Given the complexities that firewalls introduce into a network
design, you must take care in placing and configuring the firewalls and the devices around the firewalls
to allow the traffic that is considered correct to pass while blocking the traffic that needs to be blocked.
IP Telephony networks have unique data flows. The phones use a client/server model for signaling for
call setup, and Unified CM controls the phones through that signaling. The data flows for the IP
Telephony RTP streams are more like a peer-to-peer network, and the phones or gateways talk directly
4-22
Chapter 4
to each other via the RTP streams. If the signaling flows do not go through the firewall so that the firewall
can inspect the signaling traffic, the RTP streams could be blocked because the firewall will not know
which ports need to be opened to allow the RTP streams for a conversation.
A firewall placed in a correctly designed network can force all the data through that device, so capacities
and performance need to be taken into account. Performance includes the amount of latency, which can
be increased by a firewall if the firewall is under high load or even under attack. The general rule in an
IP Telephony deployment is to keep the CPU usage of the firewalls to less than 60% for normal usage.
If the CPU runs over 60%, it increases the chance of impacting IP phones, call setup, and registration.
If the CPU usage stays at a sustained level above 60%, the registered IP phones will be affected, quality
of calls in progress will degrade, and call setup for new calls will suffer. In the worst case, if the sustained
CPU usage stays above 60%, phones will start to unregister. When this happens, they will attempt to
re-register with Unified CM, thus increasing the load on the firewalls even more. If this were to happen,
the effect would be a rolling blackout of phones unregistering and attempting to re-register with
Unified CM. Until the CPU usage of the firewall decreases to under 60% sustained load, this rolling
blackout would continue and most (if not all) of the phones would be affected. If you are currently using
a Cisco firewall in your network, you should monitor the CPU usage carefully when adding IP Telephony
traffic to your network so that you do not adversely affect that traffic.
There are many ways to deploy firewalls. This section concentrates on the Cisco Adaptive Security
Appliance (ASA) in the active/standby mode in both routed and transparent scenarios. Each of the
configurations in this section is in single-context mode within the voice sections of the firewall
configurations.
All of the Cisco firewalls can run in either multiple-context or single-context mode. In single-context
mode, the firewall is a single firewall that controls all traffic flowing through it. In multiple-context
mode, the firewalls can be turned into many virtual firewalls. Each of these contexts or virtual firewalls
have their own configurations and can be controlled by different groups or administrators. Each time a
new context is added to a firewall, it will increase the load and memory requirements on that firewall.
When you deploy a new context, make sure that the CPU requirements are met so that voice RTP streams
are not adversely affected.
Adaptive Security Appliances have limited support for application inspection of IPv6 traffic for Unified
Communications application servers and endpoints. Cisco recommends not using IPv6 for Unified
Communications if ASAs are deployed in your network.
Note
4-23
Chapter 4
Firewalls
Note
The timers for failover on the firewalls are set quite high by default. To keep from affecting voice RTP
streams as they go through the firewall if there is a failover, Cisco recommends reducing those timer
settings to less than one second. If this is done, and if there is a failover, the amount of time that the RTP
streams could be affected will be less because the firewalls will fail-over quicker and there will be less
impact on the RTP streams during the failover time.
When firewalls are placed between different Unified Communications components, the application
inspection must be enabled for all protocols used for communications between the components.
Application inspection can fail in call flow scenarios used by features such as Silent Monitoring by
Unified Communications Manager, when the firewall is between the remote agent phones and the
supervisor phones.
Unified Communications devices using TCP, such as Cisco Unified Communications Manager, support
the TCP SACK option to speed up data transfer in case of packet loss. But not all firewalls support the
TCP SACK option. In that case, TCP sessions established between Unified Communications devices
through such a firewall will encounter problems if they attempt to use the TCP SACK option, and the
TCP session might fail. Therefore, the firewalls should provide full support for the TCP SACK option.
If support is not available, then the firewalls should be able to modify the TCP packets during the
three-way handshake and to disable TCP SACK option support so that the endpoints will not attempt to
use this option.
To determine if the applications running on your network are supported with the version of firewall in
the network or if ACLs have to be written, refer to the appropriate application documentation available at
http://www.cisco.com
Routed ASA
The ASA firewall in routed mode acts as a router between connected networks, and each interface
requires an IP address on a different subnet. In single-context mode, the routed firewall supports Open
Shortest Path First (OSPF) and Routing Information Protocol (RIP) in passive mode. Multiple-context
mode supports static routes only. ASA version 8.x also supports Enhanced Interior Gateway Routing
Protocol (EIGRP). Cisco recommends using the advanced routing capabilities of the upstream and
downstream routers instead of relying on the security appliance for extensive routing needs. For more
information on the routed mode, refer to the Cisco Security Appliance Command Line Configuration
Guide, available at
http://www.cisco.com/en/US/products/ps6120/products_installation_and_configuration_guides_lis
t.html
The routed ASA firewall supports QoS, NAT, and VPN termination to the box, which are not supported
in the transparent mode (see Transparent ASA, page 4-25). With the routed configuration, each interface
on the ASA would have an IP address. In the transparent mode, there would be no IP address on the
interfaces other then the IP address to manage the ASA remotely.
The limitations of this mode, when compared to the transparent mode, are that the device can be seen in
the network and, because of that, it can be a point of attack. In addition, placing a routed ASA firewall
in a network changes the network routing because some of the routing can be done by the firewall. IP
addresses must also be available for all the interfaces on the firewall that are going to be use, so changing
the IP addresses of the routers in the network might also be required. If a routing protocol or RSVP is to
be allowed through the ASA firewall, then an ACL will have to be put on the inside (or most trusted)
interface to allow that traffic to pass to the outside (or lesser trusted) interfaces. That ACL must also
define all other traffic that will be allowed out of the most trusted interface.
4-24
Chapter 4
Transparent ASA
The ASA firewall can be configured to be a Layer 2 firewall (also known as "bump in the wire" or
"stealth firewall"). In this configuration, the firewall does not have an IP address (other than for
management purposes), and all of the transactions are done at Layer 2 of the network. Even though the
firewall acts as a bridge, Layer 3 traffic cannot pass through the security appliance unless you explicitly
permit it with an extended access list. The only traffic allowed without an access list is Address
Resolution Protocol (ARP) traffic.
This configuration has the advantage that an attacker cannot see the firewall because it is not doing any
dynamic routing. Static routing is required to make the firewall work even in transparent mode.
This configuration also makes it easier to place the firewall into an existing network because routing does
not have to change for the firewall. It also makes the firewall easier to manage and debug because it is
not doing any routing within the firewall. Because the firewall is not processing routing requests, the
performance of the firewall is usually somewhat higher with inspect commands and overall traffic than
the same firewall model and software that is doing routing.
With transparent mode, if you are going to pass data for routing, you will also have to define the ACLs
both inside and outside the firewall to allow traffic, unlike with the same firewall in routed mode. Cisco
Discovery Protocol (CDP) traffic will not pass through the device even if it is defined. Each directly
connected network must be on the same subnet. You cannot share interfaces between contexts; if you
plan on running multiple-context mode, you will have to use additional interfaces. You must define all
non-IP traffic, such as routing protocols, with an ACL to allow that traffic through the firewall. QoS is
not supported in transparent mode. Multicast traffic can be allowed to go through the firewall with an
extended ACL, but it is not a multicast device. In transparent mode, the firewall does not support VPN
termination other than for the management interface.
If a routing protocol or RSVP is to be allowed through the ASA firewall, then an ACL will have to be
put on the inside (or most trusted) interface to allow that traffic to pass to the outside (or lesser trusted)
interfaces. That ACL must also define all other traffic that will be allowed out of the most trusted
interface.
For more information on the transparent mode, refer to the Cisco Security Appliance Command Line
Configuration Guide, available at
http://www.cisco.com/en/US/products/ps6120/products_installation_and_configuration_guides_lis
t.html
Note
Using NAT in transparent mode requires ASA version 8.0(2) or later. For more information, refer to the
Cisco ASA 5500 Series Release Notes at
http://www.cisco.com/en/US/docs/security/asa/asa80/release/notes/asarn80.html.
4-25
Chapter 4
Data Center
packet appears to come from its assigned public address. When external devices send packets back to
the NAT at the public address, the NAT translates the IP addresses back to the internal private addresses
and then forwards the packets to the internal network.
The NAT functionality is often part of the firewall and is therefore sometimes referred to as a NAT/FW.
NATs map a large set of internal, private IP addresses into a smaller set of external, public IP addresses.
The current public IPv4 address space is limited, and until IPv6 emerges as a ubiquitous protocol, most
enterprises will have a limited number of IPv4 public addresses available. The NAT allows an enterprise
with a large number of endpoints to make use of a small pool of public IP addresses. The NAT
implements this functionality by dynamically mapping an internal IP address to an external IP address
whenever an internal endpoint makes a connection out through the NAT. Each of these mappings is called
a NAT binding.
The major complication in implementing NAT for voice and video devices occurs because the signaling
protocols for voice and video include source addresses and ports in the protocol signaling messages.
These source addresses provide the destination addresses that remote endpoints should use for return
packets. However, internal endpoints use addresses from the private address space, and a NAT without
an Application Layer Gateway (ALG) does not alter these internal addresses. When the remote endpoint
receives a message, it cannot route packets to the private IP address in the message. Fixing this problem
requires enabling an ALG, for example a SIP, H.323, or SCCP 'fixup', on the NAT device that can inspect
the contents of the packet and implement address translation for the media IP addresses and port numbers
encapsulated in the signaling messages.
A NAT ALG is similar to a firewall ALG, but a NAT ALG actually changes (maps) the addresses and
ports in the signaling messages. The NAT ALG cannot inspect the contents of encrypted signaling
messages.
Data Center
Within the data center, the security policy should define what security is needed for the IP Telephony
applications servers. Because the Cisco Unified Communications servers are based on IP, the security
that you would put on any other time-sensitive data within a data center could be applied to those servers
as well.
If clustering over the WAN is being used between data centers, any additional security that is applied
both within and between those data centers has to fit within the maximum round-trip time that is allowed
between nodes in a cluster. In a multisite or redundant data center implementation that uses clustering
over the WAN, if your current security policy for application servers requires securing the traffic
between servers across data center firewalls, then Cisco recommends using IPSec tunnels for this traffic
between the infrastructure security systems already deployed.
To design appropriate data center security for your data applications, Cisco recommends following the
guidelines presented in the Data Center Networking: Server Farm Security SRND (Server Farm Security
in the Business Ready Data Center Architecture), available at
http://www.cisco.com/go/designzone
4-26
Chapter 4
Securing Gateways and Media Resources with IPSec, ACLs, and QoS
ACLs
Core
Distribution
Si
Si
Access
IP
IP
IP
148499
IP
4-27
Chapter 4
Some gateways and media resources support Secure RTP (SRTP) to the gateways and media resources
from the phones, if the phone is enabled for SRTP. To determine if a gateway or media resource supports
SRTP, refer to the appropriate product documentation at:
http://www.cisco.com
For more information on IPSec tunnels, refer to the Site-to-Site IPSec VPN Solution Reference Network
Design (SRND), available at:
http://www.cisco.com/go/designzone
IP
PSTN
IP
Signaling
Media
148896
The second way to deploy the gateway is outside the firewall. Because the only type of data that is ever
sent to the gateway from the phones is RTP streams, the access switch's QoS features control the amount
of RTP traffic that can be sent to that gateway. The only thing that Unified CM sends to the gateway is
the signaling to set up the call. If the gateway is put in an area of the network that is trusted, the only
communication that has to be allowed between Unified CM and the gateway is that signaling. (See
Figure 4-12.) This method of deployment decreases the load on the firewall because the RTP streams are
not going through the firewall.
4-28
Chapter 4
Unlike an ACL, most firewall configurations will open only the RTP stream port that Unified CM has
told the phone and the gateway to use between those two devices as long as the signaling goes through
the firewall. The firewall also has additional features for DoS attacks and Cisco Intrusion Prevention
System (IPS) signatures to look at interesting traffic and determine if any attackers are doing something
they should not be doing.
As stated in the section on Firewalls, page 4-22, when a firewall is looking at all the signaling and RTP
streams from phones to a gateway, capacity could be an issue. Also, if data other than voice data is
running through the firewall, CPU usage must be monitored to make sure that the firewall does not affect
the calls that are running through the firewall.
4-29
Chapter 4
Figure 4-13
Unified CM
MCU
Video Endpoints
SIP/TLS
SRTP
348688
HTTPS
4-30
Chapter 4
Cisco Unified Border Element can register the enterprise network's E.164 DID numbers to the service
provider's SIP trunk on behalf of the endpoints behind it. If Cisco Unified Border Element is used to
proxy the network's E.164 DID numbers, the status of the actual endpoint is not monitored. Therefore
unregistered endpoints might still be seen as available.
Cisco Unified Border Element can connect RTP enterprise networks with SRTP over an external
network. This allows secure communications without the need to deploy SRTP within the enterprise. It
also supports RTP-SRTP interworking, but this is limited to a small number of codecs, including G.711
mulaw, G.711 alaw, G.729abr8, G.729ar8, G.729br8, and G.729r8.
Certain SIP service providers require SIP trunks to be registered before they allow call service. This
ensures that calls originate only from well-known endpoints, thus making the service negotiation
between the enterprise and the service provider more secure. Unified CM does not support registration
on SIP trunks natively, but this support can be accomplished by using a Cisco Unified Border Element.
The Cisco Unified Border Element registers to the service provider with the phone numbers of the
enterprise on behalf of Cisco Unified Communications Manager.
For configuration and product details about Cisco Unified Border Element, refer to the documentation at:
http://www.cisco.com/en/US/products/sw/voicesw/ps5640/index.html
http://www.cisco.com/en/US/products/sw/voicesw/ps5640/products_installation_and_configuration_
guides_list.html
4-31
Chapter 4
Figure 4-14
Enterprise
Network
Service Provider
Network
348782
Unified
CM
TFTP
Unified Border
Element
SIP/RTP
TLS/SRTP
Public Network
ers
Users
TFTP/HTTP
For information on supported devices and versions, refer to the Cisco Unified Border Element product
documentation available at
http://www.cisco.com/c/en/us/products/unified-communications/unified-border-element/index.htm
l
4-32
Chapter 4
On VCS, the neighbor zone to Unified CM must be configured to allow TLS. When the zone is
configured to use media encryption for calls, end-to-end TLS and SRTP are enabled for calls between
Unified CM and VCS endpoints (see Figure 4-15).
Figure 4-15
Cisco
TelePresence
System
VCS Control
VCS Expressway
SIP
Internet
TLS
SIP/TLS
SRTP (Video)
EX Series
EX Series
348689
9900
Series
C Series
Unified CM supports H.235 pass-through as a security mechanism when interacting with H.323 video
devices, and it has added support for Secure Real-Time Transport Protocol (SRTP) encryption of the
video and audio media streams of video calls of Cisco SIP video endpoints. However, interworking of
H.235 to SRTP is not currently supported in Unified CM. Whenever H.235 and SRTP are needed in a
video deployment, Cisco recommends registering the H.323 endpoints to a Cisco VCS as a gatekeeper
and using SIP-to-H.323 interworking, while providing SRTP for the SIP video endpoints in the
Unified CM side and a secure SIP trunk to the VCS. If the H.323 video endpoints are configured to use
H.235 with the VCS, the call can be encrypted end-to-end.
4-33
Chapter 4
Applications Servers
Applications Servers
For a list of the Unified CM security features and how to enable them, refer to the Cisco Unified
Communications Manager Security Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
Before enabling any of the Unified CM security features, verify that they will satisfy the security
requirements specified in your enterprise security policy for these types of devices in a network. For
more information, refer to the Cisco ASA 5500 Series Release Notes at
http://www.cisco.com/en/US/docs/security/asa/asa80/release/notes/asarn80.html
Single Sign-On
The Single Sign-On (SSO) feature allows end users to log into a Windows domain and have secure access
to the Unified Communication Manager's User Options page and the Cisco Unified Communications
Integration for Microsoft Office Communicator (CUCIMOC) application.
Configuring Single Sign-On requires integration of Cisco Unified CM with third-party applications,
including Microsoft Windows Servers, Microsoft Active Directory, and the ForgeRock Open Access
Manager (OpenAM). For configuration details, refer to the latest version of the Feature Configuration
Guide for Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
4-34
Chapter 4
ensure that work that is not related to configuring or managing the voice systems is not done on the IP
Telephony servers at any time. Activities that might be considered normal on application servers within
a network (for example, surfing the internet) should not take place on the IP Telephony servers.
In addition, Cisco provides a well defined patch system for the IP Telephony servers, and it should be
applied based on the patch policy within your IT organization. You should not patch the system normally
using the OS vendors patch system unless it is approved by Cisco Systems. All patches should be
downloaded from Cisco or from the OS vendor as directed by Cisco Systems, and applied according to
the patch installation process.
You should use the OS hardening techniques if your security policy requires you to lock down the OS
even more than what is provided in the default installation.
To receive security alerts, you can subscribe to the Cisco Notification service at:
http://www.cisco.com/cisco/support/notifications.html
Deployment Examples
This section presents examples of what could be done from a security perspective for a lobby phone and
a firewall deployment. A good security policy should be in place to cover deployments similar to these
types.
4-35
Chapter 4
Deployment Examples
A single VLAN could be used and Cisco Discovery Protocol (CDP) could be disabled on the port so that
attackers would not be able to see any information from the Ethernet port about that port or switch to
which it is attached. In this case, the phone would not have a CDP entry in the switch for E911 emergency
calls, and each lobby phone would need either a label or an information message to local security when
an emergency number is dialed.
A static entry in the DHCP Snooping binding table could be made because there would be no DHCP on
the port (see DHCP Snooping: Prevent Rogue DHCP Server Attacks, page 4-8). Once the static entry is
in the DHCP Snooping binding table, Dynamic ARP Inspection could be enabled on the VLAN to keep
the attacker from getting other information about one of the Layer 2 neighbors on the network (see
Requirement for Dynamic ARP Inspection, page 4-11).
With a static entry in the DHCP Snooping binding table, IP Source Guard could be used. If an attacker
got the MAC address and the IP address and then started sending packets, only packets with the correct
IP address could be sent.
A VLAN ACL could be written to allow only the ports and IP addresses that are needed for the phones
to operate (see VLAN Access Control Lists, page 4-21). The following example contains a very small
ACL that can be applied to a port at Layer 2 or at the first Layer 3 device to help control access into the
network (see Router Access Control Lists, page 4-21). This example is based on a Cisco 7960 IP Phone
being used in a lobby area, without music on hold to the phone or HTTP access from the phone.
4-36
Chapter 4
Figure 4-16
Data
Center
Cisco Unified CM
Access
IP
IP
M
DMZ
Voice
Gateways
V
V
IP
IP
PSTN
148996
Voice
Applications
4-37
Chapter 4
M
M
FUSION
Router
Data
Center
Call Signaling
Call Signaling
Inside
Outside
Campus
Core
271407
Media
VRF 1
IP
VRF 2
This scenario is the simplest to implement and is an incremental configuration change beyond the usual
network virtualization implementation. This design incorporates a data center router with the capability
to route packets to any VRF, and it is called the fusion router. (Refer to the Network Virtualization
documentation for details on the configuration of the fusion router.) The deployment scenario for
enabling peer-to-peer communications traffic utilizes the fusion router for routing between VRFs and
the firewall capabilities for securing access to the data center.
The following base requirements apply to this scenario:
Campus routers send packets for other campus VRFs toward the fusion router via default routing,
so all router hops must route by default to the fusion router. The data center shared VRF has route
information about each campus VRF. All VRFs other than the shared VRF have no direct
connectivity.
A Unified CM cluster is located in a shared VRF in the data center, and communication within that
shared VRF is totally unhindered.
4-38
Chapter 4
The shared VRF is located in the data center. If multiple data centers exist, the shared VRF spans
all the data centers.
The application layer gateway at the data center edge specifies access lists to open ports for TFTP and
SCCP or SIP sessions originated on the outside toward the Unified CM cluster in the data center. TFTP
is required to allow phones to download their configuration and software images from their TFTP server,
and SCCP or SIP is required to allow them to register with the Unified CM cluster. Refer to Unified CM
product documentation for a list of appropriate port numbers for the particular version of software used.
In this scenario, all call signaling from communication devices in each VRF passes through the
application layer gateway, and inspection of that signaling allows the application layer gateway to
dynamically open the necessary UDP pinholes for each VRF for the RTP traffic to pass from the outside
of the firewall toward the fusion router. Without the inspection occurring on the firewalls, each RTP
stream that originates from an endpoint on the outside is not allowed to pass through the firewall. It is
the inspection of the call control signaling that allows the UDP traffic to be forwarded through the
firewall.
This deployment model provides a method to allow communication devices on a VRF-enabled network
to have peer-to-peer connectivity. The application layer gateway provides secure access to the shared
VLAN and the fusion router. All media streams between different VRFs do not take the most direct path
between endpoints. The media is backhauled to the data center to be routed via the fusion router.
4-39
Chapter 4
Figure 4-18
M
M
M
M
Data
Center
FUSION
Router
FUSION
Router
Call
Signaling
Inside
Outside
STOP
Campus
Core
Call
Signaling
Media
Media
VRF 2
VRF 1
271408
IP
The solution is to utilize Trusted Relay Point (TRP) functionality. (See Figure 4-19.) Subscribers in each
data center can invoke TRPs that provide anchoring of the media and ensure that the media streams flow
through the appropriate firewall. A phone controlled by a subscriber in the left data center must invoke
a TRP in that data center, and a phone controlled by a subscriber in the right data center must invoke a
TRP located in the right data center. The TRP provides an IP address that enables a specific host route
for media that can ensure the exact same routing path as the call signaling. This is used to ensure that
signaling and media pass via the same firewall, thus solving the issue.
4-40
Chapter 4
Figure 4-19
M
M
M
M
FUSION
Router
FUSION
Router
Data
Center
Call
Signaling
TRPs
Inside
Outside
Campus
Core
Call
Signaling
Media
Media
VRF 2
VRF 1
271409
IP
TRPs are media termination point resources that are invoked at the device level for any call involving
that device. Each device has a configuration checkbox that specifies whether a TRP should be invoked.
Conclusion
This chapter did not cover all of the security that could be enabled to protect the voice and video data
within your network. The techniques presented here are just a subset of all the tools that are available to
network administrators to protect all the data within a network. On the other hand, even these tools do
not have to be enabled within a network, depending on what level of security is required for the data
within the network overall. Choose your security methods wisely. As the security within a network
increases, so do the complexity and troubleshooting problems. It is up to each enterprise to define both
the risks and the requirements of its organization and then to apply the appropriate security within the
network and on the devices attached to that network.
4-41
Chapter 4
Conclusion
4-42
CH A P T E R
Gateways
Revised: June 15, 2015
Gateways provide a number of methods for connecting a network of collaboration endpoints to the
Public Switched Telephone Network (PSTN), a legacy PBX, or external systems. Voice and video
gateways range from entry-level and standalone platforms to high-end, feature-rich integrated routers,
chassis-based systems, and virtualized applications.
This chapter explains important factors to consider when selecting a Cisco gateway to provide the
appropriate protocol and feature support for your voice and video network. The main topics discussed
in this chapter include:
Described in
Revision Date
Cisco Expressway
5-1
Chapter 5
Gateways
Cisco analog gateways are available on the following products and series:
Cisco Analog Voice Gateways VG204XM and VG300 Series (VG310, VG320, VG350) all support
SCCP.
Cisco Integrated Services Routers Generation 2 (ISR G2) 2900, 3900, 3900E, and 4000 Series (4300
and 4400) with appropriate PVDMs and service modules or cards. PVDM4s utilized by ISR 4000
Series do not support video today.
Cisco Analog Telephone Adapter (ATA) 190 (SIP only) provides a replacement for the ATA188.
5-2
Chapter 5
Gateways
Cisco TDM and Serial Gateways
Cisco Integrated Services Routers Generation 2 (ISR G2) 1900, 2900, 3900, 3900E, 4300, and 4400
Series with appropriate PVDMs and service modules or cards
H.323
Cisco Expressway Series and Cisco TelePresence Video Communication Server (VCS) support the
following IP protocols for gateways:
H.323
SIP is the recommended call signaling protocol because it aligns with the overall Cisco Collaboration
solution and the direction of new voice and video products. However, protocol selection might depend
on site-specific requirements and the current installed base of equipment. Existing deployments might
be limited by the gateway hardware or require a different signaling protocol for a specific feature.
5-3
Chapter 5
Gateways
For example, placement of certain Cisco video gateways within the network depends upon the existing
call control architecture. Both the Cisco ISDN and serial gateways are optimized for video calls and were
initially designed to work with the Cisco VCS. The Cisco TelePresence Serial Gateway 8330 and 3340
platforms are recommended to register with a Cisco VCS using H.323, as shown in Figure 5-1.
Figure 5-1
Unied CM
Serial
Gateway
VCS
PSTN
SIP
H.323
348975
TDM
Serial
The Cisco TelePresence ISDN Gateway 8321 and 3241 support SIP beginning with version 2.2. The
Cisco 8321 and 3241 gateways can either register to VCS using H.323 (as shown in Figure 5-2) or trunk
directly to Unified CM using SIP (as shown in Figure 5-3).
Figure 5-2
Unied CM
ISDN
Gateway
VCS
PSTN
SIP
TDM
Serial
348896
H.323
5-4
Chapter 5
Gateways
Cisco TDM and Serial Gateways
Figure 5-3
Unied CM
ISDN
Gateway
PSTN
SIP
H.323
348897
TDM
Serial
In addition, the Unified CM deployment model being used can influence gateway protocol selection.
(Refer to the chapter on Collaboration Deployment Models, page 10-1.)
Refer to the gateway product documentation to verify that any gateway you select for an enterprise
deployment can support the preceding core requirements. Additionally, every collaboration
implementation has its own site-specific feature requirements, such as analog or digital access, DID, and
capacity requirements.
DTMF Relay
Dual-Tone Multifrequency (DTMF) is a signaling method that uses specific pairs of frequencies within
the voice band for signals. A 64 kbps pulse code modulation (PCM) voice channel can carry these signals
without difficulty. However, when using a low bite-rate codec for voice compression, the potential exists
for DTMF signal loss or distortion. An out-of-band signaling method for carrying DTMF tones across
an IP infrastructure provides an elegant solution for these codec-induced symptoms.
SCCP Gateways
The Cisco VG300 Series carries DTMF signals out-of-band using Transmission Control Protocol (TCP)
port 2002. Out-of-band DTMF is the default gateway configuration mode for the VG310, VG320, and
VG350.
5-5
Chapter 5
Gateways
H.323 Gateways
H.323 gateways, such as the Cisco 4000 Series products, can communicate with Unified CM using the
enhanced H.245 capability for exchanging DTMF signals out-of-band. This capability is enabled
through the command line interface (CLI) of the 4000 Series gateway and the dtmf-relay command
available in its dial-peers.
MGCP Gateway
Cisco IOS-based platforms can use MGCP for Unified CM communication. Within the MGCP protocol
is the concept of packages. The MGCP gateway loads the DTMF package upon start-up. The MGCP
gateway sends symbols over the control channel to represent any DTMF tones it receives. Unified CM
then interprets these signals and passes on the DTMF signals, out-of-band, to the signaling endpoint.
The method used for DTMF can be configured using the gateway CLI command:
mgcp dtmf-relay voip codec all mode {DTMF method}
Note
An MGCP gateway cannot be forced to advertise only in-band DTMF. On enabling in-band DTMF relay,
the MGCP gateway will advertise both in-band and out-of-band (OOB) DTMF methods. Unified CM
determines which method should be selected and informs the gateway using MGCP signaling. If both
the endpoints are MGCP, there is no ability to invoke in-band for DTMF relay because after enabling
in-band DTMF, both sides will advertise in-band and OOB DTMF methods to Unified CM. Unified CM
will always select OOB if in-band and OOB capabilities are supported by the endpoints.
SIP Gateway
Cisco IOS and ISDN gateways can use SIP for Unified CM communication. They support various
methods for DTMF, but only the following methods can be used to communicate with Unified CM:
The method used for DTMF can be configured in Cisco IOS using the gateway CLI command
dtmf-relay under the respective dial-peer. The Cisco ISDN gateways support RFC 2833 and KPML for
DTMF.
For more details on DTMF method selection, see the section on Calls over SIP Trunks, page 7-9.
Supplementary Services
Supplementary services provide user functions such as hold, transfer, and conferencing. These are
considered basic telephony features and are more common in voice calls than in video calls.
SCCP Gateways
The Cisco SCCP gateways provide full supplementary service support. The SCCP gateways use the
Gateway-to-Unified CM signaling channel and SCCP to exchange call control parameters.
H.323 Gateways
H.323v2 implements Open/Close LogicalChannel and the emptyCapabilitySet features. The use of
H.323v2 by H.323 gateways eliminates the requirement for an MTP to provide supplementary services.
A transcoder is allocated dynamically only if required during a call to provide access to G.711-only
devices while still maintaining a G.729 stream across the WAN.
5-6
Chapter 5
Gateways
Cisco TDM and Serial Gateways
Once an H.323v2 call is set up between a Cisco IOS gateway and an IP endpoint, using the Unified CM
as an H.323 proxy, the endpoint can request to modify the bearer connection. Because the Real-Time
Transport Protocol (RTP) stream is directly connected to the endpoint from the Cisco IOS gateway, a
supported media codec can be negotiated.
Figure 5-4 and the following steps illustrate a call transfer between two IP phones:
1.
If IP Phone 1 wishes to transfer the call from the Cisco IOS gateway to Phone 2, it issues a transfer
request to Unified CM using SCCP.
2.
Unified CM translates this request into an H.323v2 CloseLogicalChannel request to the Cisco IOS
gateway for the appropriate SessionID.
3.
4.
Unified CM issues a request to Phone 2, using SCCP, to set up an RTP connection to the Cisco IOS
gateway. At the same time, Unified CM issues an OpenLogicalChannel request to the Cisco IOS
gateway with the new destination parameters, but using the same SessionID.
5.
After the Cisco IOS gateway acknowledges the request, an RTP voice bearer channel is established
between Phone 2 and the Cisco IOS gateway.
Figure 5-4
Cisco Unied CM
Step 1
Step 2
Phone 1
PSTN
H.323
gateway
Phone 2
Cisco Unied CM
Step 3
Phone 1
Phone 2
Step 4
PSTN
H.323
gateway
Step 5
Voice/RTP path
348898
H.323
SCCP
5-7
Chapter 5
Gateways
MGCP Gateway
The MGCP gateways provide full support for the hold, transfer, and conference features through the
MGCP protocol. Because MGCP is a master/slave protocol with Unified CM controlling all session
intelligence, Unified CM can easily manipulate MGCP gateway voice connections. If an IP telephony
endpoint (for example, an IP phone) needs to modify the session (for example, transfer the call to another
endpoint), the endpoint would notify Unified CM using SCCP. Unified CM then informs the MGCP
gateway, using the MGCP User Datagram Protocol (UDP) control connection, to terminate the current
RTP stream associated with the Session ID and to start a new media session with the new endpoint
information. Figure 5-5 illustrates the protocols exchanged between the MGCP gateway, endpoints, and
Unified CM.
Figure 5-5
Cisco Unied CM
Phone 1
PSTN
MGCP
gateway
Phone 2
Cisco Unied CM
Phone 1
PSTN
MGCP
gateway
Phone 2
MGCP
SCCP
348899
Voice/RTP path
5-8
Chapter 5
Gateways
Cisco TDM and Serial Gateways
SIP Gateway
The Unified CM SIP trunk interface to Cisco SIP gateways supports supplementary services such as
hold, blind transfer, and attended transfer. The support for supplementary services is achieved via SIP
methods such as INVITE and REFER. The corresponding SIP gateway must also support these methods
in order for supplementary services to work. For more details, refer to the following documentation:
Unified CM Redundancy
An integral piece of the collaboration solution architecture is the provisioning of low-cost, distributed
PC-based systems to replace expensive and proprietary legacy PBX systems. This distributed design
lends itself to the robust fault tolerant architecture of clustered Unified CMs. Even in its most simplistic
form (a two-system cluster), a secondary Unified CM should be able to pick up control of all gateways
initially managed by the primary Unified CM.
SCCP Gateways
Upon boot-up, the Cisco VG310, VG320, and VG350 gateways are provisioned with Unified CM server
information. When these gateways initialize, a list of Unified CMs is downloaded to the gateways. This
list is prioritized into a primary Unified CM and secondary Unified CM. In the event that the primary
Unified CM becomes unreachable, the gateway registers with the secondary Unified CM.
H.323 VoIP Call Preservation for WAN Link Failures
H.323 call preservation enhancements for WAN link failures sustain connectivity for H.323 topologies
where signaling is handled by an entity that is different from the other endpoint, such as a gatekeeper
that provides routed signaling or a call agent (such as Cisco Unified CM) that brokers signaling between
the two connected parties. Call preservation is useful when a gateway and the other endpoint are located
at the same site but the call agent is remote and therefore more likely to experience connectivity failures.
H.323 call preservation covers the following types of failures and connections.
Failure Types:
WAN failures that include WAN links flapping or degraded WAN links.
Cisco Unified CM software failure, such as when the ccm.exe service crashes on a Unified CM
server.
LAN connectivity failure, except when a failure occurs at the local branch.
5-9
Chapter 5
Gateways
Connection Types:
Calls between two Cisco Unified CM controlled endpoints under the following conditions:
During Unified CM reloads.
When a Transmission Control Protocol (TCP) connection used for signaling H.225.0 or H.245
Calls between a Cisco IOS gateway and an endpoint controlled by a softswitch, where the signaling
(H.225.0, H.245 or both) flows between the gateway and the softswitch and media flows between
the gateway and the endpoint:
When the softswitch reloads.
When the H.225.0 or H.245 TCP connection between the gateway and the softswitch is lost, and
Call flows involving a Cisco Unified Border Element running in media flow-around mode that
reload or lose connection with the rest of the network.
Note that, after the media is preserved, the call is torn down later when either one of the parties hangs
up or media inactivity is detected. In cases where there is a machine-generated media stream, such as
music streaming from a media server, the media inactivity detection will not work and then the call might
hang. Cisco Unified CM addresses such conditions by indicating to the gateway that such calls should
not be preserved, but third-party devices or the Cisco Unified Border Element would not do this.
Flapping is defined for this feature as the repeated and temporary loss of IP connectivity, which can be
caused by WAN or LAN failures. H.323 calls between a Cisco IOS gateway and Cisco Unified CM may
be torn down when flapping occurs. When Unified CM detects that the TCP connection is lost, it clears
the call and closes the TCP sockets used for the call by sending a TCP FIN, without sending an H.225.0
Release Complete or H.245 End Session message. This is called quiet clearing. The TCP FIN sent from
Unified CM could reach the gateway if the network comes up for a short duration, and the gateway will
tear down the call. Even if the TCP FIN does not reach the gateway, the TCP keepalives sent from the
gateway could reach Unified CM when the network comes up. Unified CM will send TCP RST messages
in response to the keepalives because it has already closed the TCP connection. The gateway will tear
down H.323 calls if it receives the RST message.
Configuration of H.323 call preservation enhancements for WAN link failures involves configuring the
call preserve command. If you are using Cisco Unified CM, you must enable the Allow Peer to Preserve
H.323 Calls parameter from the Service Parameters window.
The call preserve command causes the gateway to ignore socket closure or socket errors on H.225.0 or
H.245 connections for active calls, thus allowing the socket to be closed without tearing down calls using
those connections.
MGCP Gateway
MGCP gateways also have the ability to fail over to a secondary Unified CM in the event of
communication loss with the primary Unified CM. When the failover occurs, active calls are preserved.
Within the MGCP gateway configuration file, the primary Unified CM is identified using the call-agent
<hostname> command, and a list of secondary Unified CM is added using the ccm-manager
redundant-host command. Keepalives with the primary Unified CM are through the MGCP
5-10
Chapter 5
Gateways
Gateways for Video Telephony
application-level keepalive mechanism, whereby the MGCP gateway sends an empty MGCP notify
(NTFY) message to Unified CM and waits for an acknowledgement. Keepalive with the backup
Unified CMs is through the TCP keepalive mechanism.
If the primary Unified CM becomes available at a later time, the MGCP gateway can re-home, or
switch back to the original Unified CM. This re-homing can occur either immediately, after a
configurable amount of time, or only when all connected sessions have been released.
SIP Gateway
Redundancy with Cisco IOS SIP gateways can be achieved similarly to H.323. If the SIP gateway cannot
establish a connection to the primary Unified CM, it tries a second Unified CM defined under another
dial-peer statement with a higher preference.
By default the Cisco IOS SIP gateway transmits the SIP INVITE request 6 times to the Unified CM IP
address configured under the dial-peer. If the SIP gateway does not receive a response from that
Unified CM, it will try to contact the Unified CM configured under the other dial-peer with a higher
preference.
Cisco IOS SIP gateways wait for the SIP 100 response to an INVITE for a period of 500 ms. By default,
it can take up to 3 seconds for the Cisco IOS SIP gateway to reach the backup Unified CM. You can
change the SIP INVITE retry attempts under the sip-ua configuration by using the command retry invite
<number>. You can also change the period that the Cisco IOS SIP gateway waits for a SIP 100 response
to a SIP INVITE request by using the command timers trying <time> under the sip-ua configuration.
One other way to speed up the failover to the backup Unified CM is to configure the command monitor
probe icmp-ping under the dial-peer statement. If Unified CM does not respond to an Internet Control
Message Protocol (ICMP) echo message (ping), the dial-peer will be shut down. This command is useful
only when the Unified CM is not reachable. ICMP echo messages are sent every 10 seconds.
The Cisco ISDN Gateway can connect to Unified CM via SIP trunk starting with Unified CM release 9.0
and ISDN Gateway release 2.2 and later. The ISDN Gateway SIP configuration consists of entering an
IP address, hostname, DNS A record, or DNS SRV record for outbound SIP connections. Redundancy
can be achieved by utilizing DNS SRV records with appropriate weight and priority so that, if the
primary Unified CM fails, the ISDN Gateway will send outbound SIP calls to the secondary Unified CM.
Dial plan If the enterprise has the flexibility of a separate dial plan for video users, it can use
separate video gateways that allow it to keep existing enterprise dial plans.
Video users If the enterprise has a large number of users who primarily use voice rather than
video, then Cisco recommends using separate video gateways to service the video call users.
Locations If the enterprise has a large number of distributed locations with video users at many
locations, then Cisco recommends using an integrated gateway to reduce total cost of ownership
(TCO).
5-11
Chapter 5
Gateways
Additional video capabilities such as video IVR, auto attendant, and bonding across trunks
Dedicated video gateways support advanced features that integrated gateways do not support.
Protocol Gateway protocol can be an important factor to align with enterprise policies and
standards.
Unified CM System with Separate PSTN Lines for Voice and IP Video Telephony
Video
Gateway
Unied CM
PSTN
SCCP
SIP
MGCP
TDM
Serial
348900
Voice
Gateway
The Cisco video gateways, while excellent for video calls, do not support all of the telephony features
that Cisco voice gateways offer. Cisco video gateways have the following characteristics:
The ISDN Gateway supports H.323 and SIP (starting with release 2.2) for IP connectivity.
5-12
Chapter 5
Gateways
Gateways for Video Telephony
They support G.711, G.722, G.722.1, and G.728; but they do not support G.729 audio.
They support H.320, H.233, H.234, H.235 (AES), H.239, H.221, FTP, RTP, HTTP, HTTPS, DHCP,
SNMP, and NTP.
As a result of these differences in the products, the Cisco TDM and Serial Gateways are not
recommended as replacements for Cisco voice gateways. IP Telephony customers who want to add video
to their communications environment should deploy both types of gateways and use the Cisco voice
gateways for all voice calls and use the Cisco video gateways for video calls only. Customers might also
have to procure separate circuits for voice and video from their PSTN service provider, depending on
which model of Cisco gateway is deployed.
Also consider how calls will be routed across the IP network to a remote gateway for the purpose of
providing toll bypass, and how calls will be re-routed over the PSTN in the event that the IP network is
unavailable or does not have enough bandwidth to complete the call. More specifically, do you want to
invoke automated alternate routing (AAR) for video calls?
Provide H.323 and SIP support (except Serial Gateway, which is H.323 only)
The following considerations apply for deploying Cisco IOS, ISDN, and Serial Video gateways:
Consider the capacity needed on PSTN links for additional video calls.
Consider the need of devices to use content sharing such as Binary Floor Control Protocol (BFCP),
and the additional bandwidth that will be used on the IP network.
Consider if users need features such as far-end camera control or DTMF that is used for conferences
that the gateway needs to support.
Configure a SIP trunk pointing to the ISDN gateway (as shown in Figure 5-3), and add appropriate
Unified CM route patterns pointing to the SIP trunk.
Configure a SIP trunk from Unified CM to Cisco VCS. Have the ISDN gateway (or Serial gateway
in this case) register to the VCS using H.323 (as shown in Figure 5-2.
The Cisco TelePresence Serial Gateway cannot be trunked directly to Unified CM. It must register to
Cisco VCS, which in turn has a SIP trunk to Unified CM.
Either way, the goal is have all inbound calls received by the gateways sent to Unified CM so that
Unified CM can decide how to route the calls. See the chapter on Cisco Unified CM Trunks, page 6-1,
for more details on how to configure the SIP trunk between Unified CM and VCS.
5-13
Chapter 5
Gateways
H.245TCSTimeout
Cisco recommends that you increase each of these timers to 25 by modifying them under the Service
Parameters in Unified CM Administration. Note that these are cluster-wide service parameters, so they
will affect calls to all types of devices, including voice calls to existing Cisco voice gateways.
Whether the region in Unified CM is configured to allow video for calls between those devices
Unified CM supports retrying the video call as audio, and this feature can be enabled through
configuration. When Unified CM makes a video call with bearer-caps set to Unrestricted Digital and
the call fails, Unified CM then retries the same call as an audio call with the bearer-caps set to speech.
When using H.323, Cisco IOS gateways can service calls as voice or video, based on the bearer
capabilities it receives in the call setup. When using SIP, the gateway translates the ISDN capabilities
into SDP for call negotiations.
If the Cisco voice gateway uses MGCP to communicate with Unified CM, the problem will not occur
because Unified CM does not support video on its MGCP protocol stack and because, in MGCP mode,
Unified CM has complete control over the D-Channel signaling to the PSTN.
5-14
Chapter 5
Gateways
IP Gateways
IP Gateways
The Cisco IP gateways Include:
Session Control Call admission control, trunk routing, QoS, statistics, billing, redundancy,
scalability, voice quality monitoring
Security Encryption, authentication, registration, SIP protection, voice policy, toll fraud
prevention, telephony denial of service (TDoS) attack protection
Interworking Various SIP and H323 stack interoperability, sip normalization, dtmf, Transcoding,
Transrating, Codec Filtering
Demarcation Fault isolation, topology and address hiding, L5/L7 protocol demarcation, network
border
CUBE provides essential capabilities that ensure interoperability, security, and service assurance when
carrying IP traffic via SIP trunks across various enterprises and service provider networks. It is a
Back-to-Back User Agent (B2BUA) and is part of the Cisco IOS infrastructure on Cisco ISR G2 800
Series platforms, Cisco IOS-XE for the ASR 1000 Series, Cisco ISR 4000 Series, and CUBE on the
Cisco Cloud Services Router (CSR) 1000V Series (virtual CUBE, or vCUBE). Figure 5-7 illustrates the
enterprise CUBE deployment.
5-15
Chapter 5
Gateways
IP Gateways
Figure 5-7
Unied CM
TDM Backup
(Not available in vCUBE)
PSTN
CUBE
SIP
H.323
MediaSense
348976
RTP
For more information about Cisco Unified Border Element, refer to the documentation at
http://www.cisco.com/go/cube
Cisco Expressway
Use of the Internet for collaboration services continues to increase in popularity and is quickly replacing
existing legacy ISDN video systems and gateways. The two primary protocols leveraged for Internet
based collaboration services are SIP and H.323. The Internet is also used to connect remote and mobile
users to voice, video, IM and presence, and content sharing services without the use of a virtual private
network (VPN).
The Expressway-C and Expressway-E pair performs the following functions:
Mobile and remote access, as well as business-to-business services, can be enabled as part of the
same Cisco Expressway-C and Expressway-E solution pair.
Interworking The capability to interconnect H.323-to-SIP calls for voice, video, and content
sharing.
Security The capability to provide authentication and encryption for both mobile and remote
access and business-to-business communications.
Expressway-C and Expressway-E are designed to work together to form a firewall traversal solution that
is the core component for business-to-business communications over the Internet. Expressway-C sits on
the inside (trusted side) of the enterprise network and serves the role of providing a secure, trusted, and
standards-based way of connecting to Expressway-E. It acts as a traversal client to all devices behind it.
This solves the problem for devices using a large number of media ports by multiplexing all media to a
very small number of ports opened for outbound communications. It provides an authenticated and
trusted connection from inside the enterprise to outside by sending a keep-alive for the traversal zone
from Expressway-C to Expressway-E. Additionally, it provides a single point of contact for all Internet
communications, thus minimizing the security risk.
5-16
Chapter 5
Gateways
IP Gateways
Real-time and near real-time communication protocols such as SIP, H.323, and XMPP do not address
the need to communicate with devices that might be behind a firewall. Typical communications using
these protocols include the device IP address in the signaling and media, which becomes the payload of
the TCP and UDP packets, respectively. When these devices are on the same internally routable network,
they can successfully communicate directly with each other. The signaling IP address carried in the
payload of the TCP packet is routable back to the initiating device, and vice versa. However, when the
initiating device is on a different network behind a public or network edge firewall, two problems are
encountered. The first problem is that the receiving device, after decoding the packet, will respond to the
internal IP address carried in the payload. This IP address is typically a non-routable RFC 1918 address
and will never reach the return destination. The second problem is that, even if the return IP address is
routable, the media (which is RTP/UDP) is blocked by the external firewall. This applies to both
business-to-business and mobile and remote access communications.
Expressway-E sits at the network edge in the DMZ. It serves the role of solving both the signaling and
media routing problems for SIP, H323, and XMPP, while maintaining standards interoperability.
Expressway-E changes the appropriate headers and IP addresses to process the media and signaling on
behalf of the endpoints, devices, and application servers that are inside the network.
Expressway-E is the traversal server installed within the enterprise DMZ, and Expressway-C is the
traversal client installed inside the enterprise network.
2.
Expressway-C initiates traversal connections outbound through the firewall to specific ports on
Expressway-E, with secure login credentials. If the firewall allows outbound connections, as it does
in the vast majority of cases, no additional ports are required to be opened in the enterprise firewall.
For port details, refer to the latest version of the Unified Communications Mobile and Remote Access
via Cisco Expressway Deployment Guide, which includes all ports used by Expressway in
business-to-business and mobile and remote access scenarios. This guide is available at
http://www.cisco.com/c/en/us/support/unified-communications/expressway-series/products-in
stallation-and-configuration-guides-list.html
3.
Once the connection has been established, Expressway-C sends periodic keep-alive packets to
Expressway-E to maintain the connection.
5-17
Chapter 5
Gateways
IP Gateways
4.
When Expressway-E receives an incoming call or other collaboration service request, it issues an
incoming request to Expressway-C.
5.
Expressway-C then routes the request to Unified CM or other collaboration service applications.
6.
The connection is established, and application traffic (including voice and video media) traverses
the firewall securely over an existing traversal connection.
For firewall traversal to work, a traversal client zone has to be configured on Expressway-C and a
traversal server zone has to be configured on Expressway-E. Figure 5-8 summarizes the firewall traversal
process in a dual-interface deployment scenario for Expressway-E.
Figure 5-8
Enterprise Network
Intranet
Expressway-C
DMZ
Expressway-E
Internet
Signaling
Firewall Traversal
348830
Media
In the dual-interface deployment scenario, Expressway-E sits in the DMZ between two firewalls: the
Internet firewall provides for NAT services toward the Internet, and the intranet firewall provides access
to the corporate trusted network.
Expressway-E has two LAN interfaces: one toward the Internet firewall (also called the external
interface) and the other toward the intranet firewall (also called the internal interface). There is no need
for the external interface to be assigned a public IP address because the address can be translated
statically by NAT. In this case, the public IP address has to be configured on Expressway-E itself.
A connection from the Internet for business-to-business communications between Expressway-C and
back-end application services may or may not be encrypted, based on the configuration and dictated by
the corporate policies. Note that in this case the communication will be encrypted end-to-end only if both
the corporate and the remote business-to-business party supports encryption with public certificates. In
all other cases, the video call will be sent unencrypted, or it will be dropped based on Expressway-E
configuration policies.
5-18
Chapter 5
Gateways
IP Gateways
Type of Communication
Domain
Port
Protocol
SIP business-to-business
_sips._tcp.domain
5061
TLS
_sip._tcp.domain
5060
TCP
_sip._udp.domain
5060
UDP
_h323ls._udp.domain
1719
RAS
_h323cs._tcp.domain
1720
H.225
_collab-edge._tls.domain
8443
Jabber login
_xmpp-server._tcp.domain
5269
XMPP Federation
H.323 business-to-business
For more information about configuring a DNS zone on Expressway-E, refer to latest version of the
Cisco Expressway Basic Configuration Deployment Guide, available at
http://www.cisco.com/c/en/us/support/unified-communications/expressway-series/products-install
ation-and-configuration-guides-list.html
Outbound calls use a SIP Route Pattern on Cisco Unified CM set to "*". Any SIP URI that does not find
a match inside the local Unified CM cluster or ILS table will be sent through this SIP Route Pattern,
according to the routing rules logic defined in the chapter on Dial Plan, page 14-1. Configure this SIP
Route Pattern to have a Route List to the Expressway-C cluster as a target.
Configure Expressway-C to have two rules for business-to-business communications:
Send any SIP URI with the domain portion matching the domain of the company to Cisco Unified
Communications Manager.
Send any SIP URI with the domain portion matching any other domain to Expressway-E.
5-19
Chapter 5
Gateways
IP Gateways
Send any SIP URI with the domain portion matching the domain of the company to the
Expressway-C cluster.
Send any SIP URI with the domain portion matching any other domain to the DNS Zone that is used
for DNS SRV resolution.
When a user dials a string followed by an external domain from an endpoint connected to Unified CM,
the SIP Route Pattern will be matched. Unified CM will send the call to Expressway-C, and
Expressway-C to will send it Expressway-E. Expressway-E will perform a DNS SRV lookup on a public
DNS. The DNS will resolve the SRV record, and Expressway-E will be able to direct the call to the
unknown remote edge.
Inbound calls will be received by the Expressway-E on the Default Zone, and based on the search rules
specified above, Expressway-E will send the call to the Expressway-C, which will send it to Cisco
Unified CM.
Note that any Cisco endpoint connected to Cisco Unified CM, regardless from model type or voice/video
capabilities, will be reachable.
If the endpoint does not have any associated SIP URI, it will be reachable through the string
<DN>@<domain>, where <DN> is the directory number configured on Cisco Unified CM and
<domain> is the company SIP domain.
In case the device has a corresponding alphanumeric SIP URI associated with its DN, the same device
can also be reached by dialing the alphanumeric SIP URI.
Outbound Calls
Outbound IP dialing is supported on Expressway-E and Expressway-C, but it does not have full native
support on Cisco Unified CM. However, it is possible to set up Unified CM to have IP-based dialing, as
described here.
Instead of dialing the IP address alone, users on Cisco Unified CM can dial a SIP URI-based IP address
as shown in this example: 10.10.10.10@ip, where "@ip" is literal and could be replaced with "external",
"offsite" or other meaningful terms.
Unified CM will match a SIP route pattern configured to route the "ip" fictional domain to
Expressway-C. Expressway-C strips off the domain "@ip" and sends the call to Expressway-E, which is
also configured for IP address dialing.
Calls to unknown IP addresses on Expressway -E should be set to Direct. Since IP-based address dialing
is mostly configured in H.323 endpoints when no call control is deployed, this allows Expressway-E to
send H.323 calls directly to an endpoint at a public IP address. The call will remain a SIP call until
interworked on Expressway-E.
Alternatively, instead of having to append the fictional domain, users might replace the dots with a star
character, as in this example: 10*10*10*10.
5-20
Chapter 5
Gateways
IP Gateways
Unified CM will match a Route Pattern defined as !*!*!*! and send the call to Expressway-C, which will
replace the "star" character with a dot. In this case, the search rule will match the following regex
expression: (\d\d?\d?)(\*)(\d\d?\d?)(\*)(\d\d?\d?)(\*)(\d\d?\d?), and will have \1\.\3\.\5\.\7 as the
replacement string.
Inbound Calls
IP-based inbound calls make use of a fallback alias configured in Expressway-E. When a user on the
Internet dials the IP address of the Expressway-E external LAN interface, Expressway-E receives the
call and sends the call to the alias configured in the fallback alias setting. As an example, if the fallback
alias is configured to send the call to conference number 80044123 or to the conference alias
meet@ent-pa.com, the inbound call will be sent to the TelePresence Server in charge of such
conferences.
If the static mapping between the IP address and the fallback alias is too limited, it is possible to set the
fallback alias to the pilot number of Cisco Unity Connection. In this way it is possible to use the Unity
Connection auto-attendant feature to specify the final destination through DTMF, or by speech
recognition if Unity Connection is enabled to support this feature. If Unity Connection is used as an
auto-attendant feature for external endpoints dialing the IP address of the Expressway-E, remember to
set the Rerouting Calling Search Space on the Unified CM trunk configuration for Unity Connection.
Expressway-C and Expressway-E node types cannot be mixed in the same cluster.
Configuration changes should be made only on the master node, and this will overwrite the
configuration on the other nodes in the cluster when replication occurs.
If a node becomes unavailable, the licenses it contributed to the cluster will become unavailable after
2 weeks.
5-21
Chapter 5
Gateways
IP Gateways
All nodes in a cluster need to be within 30 ms maximum round-trip time to all other cluster nodes.
Clustering over the WAN is therefore not recommended due to latency constraints.
You must use the same cluster preshared key for all nodes within the same cluster.
If mobile and remote access and business-to-business communications are enabled on the same
Expressway-C and Expressway-E pairs, the SIP port number used on the SIP trunk between
Unified CM and Expressway-C must be changed from the default 5060 or 5061 (for example, use
5560 and 5561).
A DNS SRV record must be available for the cluster and must contain A or AAAA records for each
node of the cluster.
Since Expressway-C is deployed in the internal network and Expressway-E is in the DMZ,
Expressway-C has to be connected to Expressway-E through a traversal zone for business-to-business
calls. Mobile and remote access requires a separate traversal zone, referred to as Unified
Communication traversal zone. The traversal server and traversal client zones include all the nodes of
Expressway-C and Expressway-E, so that if one of the nodes is not reachable, another node of the cluster
will be reached instead.
The traversal client zone configured on Expressway-C should contain the fully qualified domain names
of all of the cluster nodes of the corresponding Expressway-E cluster. Likewise, the traversal server zone
should connect to all Expressway-C cluster nodes. This is achieved by including, in the subject
alternative names of the Expressway-C certificate, the FQDNs of the Expressway-C cluster nodes and
by setting the TLS verify subject name equal to the FQDN of the Expressway-C cluster. This creates a
mesh configuration of cluster nodes across the traversal zone and provides continuous and high
availability of the traversal zone until the last cluster node is unavailable.
Expressway-C connects to Unified CM via a neighbor zone for routing inbound and outbound
business-to-business calls. Unified CM also trunks to Expressway-C. For high availability, the fully
qualified domain names of each Expressway-C cluster node should be listed in the trunk configuration
on Unified CM. If Unified CM is clustered, the fully qualified domain name (FQDN) of each member
of the cluster should be listed in the neighbor zone profile of Expressway-C.
A meshed trunk configuration is created here as well. Unified CM will check the status of the nodes in
the trunk configuration via a SIP OPTIONS Ping. If a node is not available, Unified CM will take that
node out of service and will not route calls to it. Expressway-C will also check the status of the trunk
from Unified CM via a SIP OPTIONS Ping. Calls will be routed only to nodes that are shown as active
and available. This provides high availability for both sides of the trunk configuration.
DNS SRV records can add to availability of Expressway-E for inbound business-to-business traffic. For
high availability, all nodes in the cluster should be listed with the same priority and weight in the SRV
record. This allows all nodes to be returned in the DNS query. A DNS SRV record helps to minimize the
time spent by a client on lookups because a DNS response can contain all of the nodes listed in the SRV
record. The far-end server or far-end endpoint will typically cache the DNS response and will try all
nodes returned in the DNS query until a response is received. This provides the best chance for a
successful call.
In addition, Expressway clusters support rich media license sharing across clusters. If a node is lost from
the cluster, its call licenses will continue to be shared for the next 2 weeks.
Any one Expressway cannot process any more rich media licenses than its physical capacity, even though
it can carry more licenses than its physical capacity.
5-22
Chapter 5
Gateways
IP Gateways
Specify the source IP address subnet from which to allow or deny traffic.
Configure well known services such as SSH and HTTP/HTTPS, or specify customized rules based
on transport protocols and port ranges.
Configure different rules for the LAN 1 and LAN 2 interfaces on Expressway-E.
The Automated Intrusion Protection feature should be used to detect and block malicious traffic and to
help protect the Expressway from dictionary-based attempts to breach login security. Automated
Intrusion Protection works by parsing the system log files to detect repeated failures to access specific
service categories such as SIP, SSH, and web/HTTPS. When the number of failures within a specified
time reaches the configured threshold, the source host IP address (the intruder) and destination port are
blocked for a specified period of time. The host address is automatically unblocked after that time period
so as not to lock out any genuine hosts that might have been temporarily mis-configured.
5-23
Chapter 5
Gateways
IP Gateways
Depending on the requirements, different media encryption policies might be configured. If a corporate
enforcing policy is not in place, the recommendation is to set up zones with auto specified as the media
encrypted mode. As an example, a customer might require that calls from specific endpoints always be
encrypted, and that they fall back to unencrypted if the remote party does not support encryption. A
setting of auto delegates the encryption decisions to endpoints, and Expressway does not perform any
sort of RTP-to-SRTP conversion. This setting preserves the consistency of the lock icon on Cisco
devices. If a call is encrypted, the device will show the closed lock; if unencrypted, it will show the open
lock.
When the encryption policy is enforced on Expressway, the lock icon might be inconsistent, as in the
following scenario:
If a Unified CM endpoint calls an unencrypted endpoint on the Internet, the call will be encrypted up to
Expressway-C and unencrypted between Expressway-C and the destination. The B2BUA is engaged on
the Expressway-C. Because the B2BUA terminates the encrypted call and makes another independent
call to the destination, the icon showing encryption status in the calling device will be a closed lock
(encrypted), even if the call is sent in clear on the Internet. Conversely, depending on the media
encryption settings on the Expressway zones, a call might show an open lock but could still be encrypted
on the Internet.
Every company has the control of encryption up to the other company's edge, thus allowing an endpoint
to establish an encrypted call from the endpoint to the remote edge. Encryption policy can protect media
on the Internet if force encrypted is configured on Expressway; but once the call hits the remote edge,
the call might be decrypted at the edge level before sending it to the called endpoint.
The only way to provide for true end-to-end encryption is by controlling the remote endpoint. This could
be done by deploying Jabber Guest or giving the customer the ability to download Jabber and login
through Mobile and Remote Access in the corporate Cisco Unified CM and IM and Presence servers.
5-24
Chapter 5
Gateways
IP Gateways
Expressway-E and Expressway-C nodes, the _sip._tls.ent-pa.com and _sip._tcp.ent-pa.com records will
include all six Expressway-E records at the same priority and weight. This distributes the registrations
and calls equally across the various Expressway-E and Expressway-C clusters.
However, if the Expressway clusters are deployed across geographical regions, some intelligent
mechanisms on top of the DNS SRV priority and weight record are needed to ensure that the endpoint
uses the nearest Expressway-E cluster. As an example, if an enterprise has two Expressway clusters, one
in the United States (US) and the other in Europe (EMEA), it is desirable for users located in the US to
be directed to the Expressway-E cluster in the US while users in Europe are directed to the Expressway-E
cluster in Europe. This is facilitated by implementing GeoDNS services. GeoDNS services are cost
effective and easy to configure. To show how GeoDNS services work, the example below uses an
Amazon Route 53 Geo DNS server. There are many GeoDNS services available in the market, including
Amazon Route 53, Edge Director, GeoScaling, Max Mind GeoIP2, and others.
With GeoDNS it is possible to route traffic based on different policies such as location (IP address
routing), latency (minimum latency), and others. As an example, Amazon Route 53 allows routing by
both latency and geographical location.
With latency-based routing, a client in the same site might access different data centers over time if
latency on the Internet changes. However, this does not happen as soon as latency changes; rather,
latency is measured as a mean value over a period of time. Spikes due to instant congestion of the Internet
are thus absorbed by the mean value.
In our scenario, two Internet edge Expressway clusters are deployed, one in the US and one in Europe,
each composed of two Expressway-C and Expressway-E servers. If the measured latency between the
calling endpoint and the European edge is less than the latency between the endpoint and the US edge,
the endpoint will be directed to the European edge for registration.
Following this scenario, business-to-business SRV records resolve into expe.ent-pa.com, a CNAME
record, which is an alias that resolves into the real A-records for that resource. There are two records for
expe.ent-pa.com; the first resolving into us-expe.ent-pa.com (DNS name for the US edge) and the second
resolves into emea-expe.ent-pa.com (DNS name for the EMEA edge). A-records us.expe.ent-pa.com and
emea-expe.ent-pa.com resolve to the IP addresses of Expressway-E server nodes for US and Europe.
While DNS SRV records are configured for standard routing, expe.ent-pa.com records have a routing
policy set to latency. As a result, the locations of the Expressway-E clusters have to be specified. If the
latency between the client and emea-expe.ent-pa.com is less than the latency between the client and
us-expe.ent-pa.com, the call will be sent to the European Expressway-E.
If for any reason latency changes over time and goes above the latency to the US, the us-expe.ent-pa.com
will be selected instead.
Both emea-expe.ent-pa.com and us-expe.ent-pa.com are A-records, and since they cannot set priority
and weight like SRV records can, another load-balancing and redundancy mechanism is needed to
specify the server of the Expressway-E cluster to which the calling device has to connect. This can be
done by using a round-robin mechanism. As an example, two emea-expe.ent-pa.com records can be
created, each one with the routing policy set to weighted.
Specifying the same weight for the two records assures that an equal load-balancing process takes place
between the servers of the cluster. The first record resolves into multiple Expressway-E servers of the
same cluster (in this case, two servers). The second record resolves into the same set of servers, but in
reverse order.
Figure 5-9 shows the DNS record structure for GeoDNS with latency-based routing between regional
Expressway-E clusters and round-robin inside the same cluster. As shown in the figure, both records
emea-expe.ent-pa.com resolve into the same set of Expressway-E nodes, but in different orders. This
provides both redundancy and load balancing.
5-25
Chapter 5
Gateways
IP Gateways
Figure 5-9
SRV Record
CNAME Record
A-Records
Expressway-E
expe.ent-pa.com
locaon: us-east-1
us-expe.ent-pa.com
Weight: 50
A1.B1.C1.D1
A2.B2.C2.D2
us-expe.ent-pa.com
Weight: 50
A2.B2.C2.D2
A1.B1.C1.D1
emea-expe.ent-pa.com
Weight: 50
X1.Y1.W1.Z1
X2.Y2.W2.Z2
emea-expe.ent-pa.com
Weight: 50
X2.Y2.W2.Z2
X1.Y1.W1.Z1
_collab-edge._tls.ent-pa.com
348831
expe.ent-pa.com
locaon: eu-west-1
5-26
Chapter 5
Gateways
IP Gateways
In this example, assume that the call is sent to the EMEA Expressway-E cluster. The EMEA
Expressway-E and Expressway-C will try to send the call to the destination, but the inbound calling
search space of the Expressway-C trunk will block the call. The EMEA Expressway-E will then forward
the call to the APJC Expressway-E. This time the call will be delivered to the destination because the
inbound calling search space of APJC Expressway-C contains the APJC endpoints partition.
In order to allow the Expressway-E in EMEA to remove itself from the signaling and media path, it is
important to make sure that there is no TCP-to-TLS or RTP-to-SRTP conversion on Expressway-E
EMEA clusters, and to make sure that the call signaling optimization parameter is set to On in all
Expressway-C and Expressway-E nodes.
Because this is not a deterministic process, in the case of three or more Expressway edges the searching
mechanism would require too much time. Therefore, this configuration is recommended for no more
than two Expressway edges.
Figure 5-10 shows the Expressway edge design that enables selection of the edge closest to the
destination endpoint.
Figure 5-10
EMEA
east@ent-pa.com
APJC
11
Expressway-C
EMEA
5
EMEA Inbound CSS trunk only
includes EMEA phones. It does
not include APJC phones like
east@ent-pa.com.
VCS
V
CS S
2. Send localdomain to E
APJC w/stop
Expressway-C
APJC
Expressway-E
EMEA
Search Rules
10
Expressway-E
APJC
Search Rules
1. Send Localdomain to
UCM w/connue
VCS
V
CS S
2. Send localdomain to E
EMEA w/stop
SIP
12
2
Internet
signaling
media
348832
1 Dial: east@ent
east@ent-pa.com
pa.c
5-27
Chapter 5
Gateways
IP Gateways
SJC_International
SJC_PSTNInternational
Partition
CUBE_US_PSTN1
SJC_PSTNInternaonal
CUBE_US_PSTN2
Trunk_To_
CUBE_US1
Phone in SJC
RCD_International
RCD_PSTNInternational
Partition
CUBE_US_PSTN2
RCD_PSTNInternaonal
CUBE_US_PSTN1
Trunk_To_
CUBE_US2
Phone in RCD
348833
CUBE2
The Unified CM Local Route Group feature helps scale this solution when multiple sites access two or
more Expressway-C clusters. This mechanism is also applied on ISDN gateways and Cisco Unified
Border Element. A full description of the configuration is documented in the next two sections, since it
also applies to Cisco Unified Border Element and voice gateways.
5-28
Chapter 5
Gateways
Best Practices for Gateways
Assign a single directory number to each user for both video and voice calls. This method is not
recommended because all calls would have to be received from the PSTN on a video gateway,
including audio-only calls. This would waste valuable video gateway resources and be hard to scale.
Assign at least two different directory numbers to each video-enabled device in the Unified CM
cluster, with one line for audio and another line for video. With this method, the outside (PSTN)
caller must dial the correct number to enable video.
For video calls, have outside callers dial the main number of the video gateway. Cisco ISDN and
Serial gateways offer an integrated auto-attendant that prompts the caller to enter the extension
number of the party they are trying to reach. Unified CM will then recognize that it is a video call
when ringing the destination device. This method relieves the caller from having to remember two
different DID numbers for each called party, but it adds an extra step to dialing an inbound video
call.
Note
The outside video endpoints must support DTMF in order to enter the extension of the called
party at the IVR prompt.
5-29
Chapter 5
Gateways
to that DID number through T1-PRI circuit(s) connected to a Cisco Voice Gateway. When the
call is received by the gateway, Unified CM knows that the gateway is capable of audio only, so
it negotiates only a single audio channel for that call. Conversely, for people to reach the user
from the PSTN for a video call, they must dial the main number of the video gateway and then
enter the users extension. For example, they might dial 1-408-555-1000. The CO would send
calls to that number through the T1-PRI circuit(s) connected to a Cisco ISDN video gateway.
When the call is received by the gateway, an auto-attendant prompt asks the caller to enter the
extension of the person they are trying to reach. When the caller enters the extension via DTMF
tones, Unified CM knows that the gateway is capable of video, so it negotiates both audio and
video channels for that call.
Place a call to a number (or IP address, hostname, or URI in the case of a PSTN-to-IP call)
When the action is to place a call to a number, the original called number or parts of it can be used in
the new number to call.
For more details, refer to the following documentation:
Assign different access codes (that is, different route patterns) for voice and video calls. For
example, when the user dials 9 followed by the PSTN telephone number they are trying to reach, it
could match a route pattern that directs the call out a voice gateway. Similarly, the digit 8 could be
used for the route pattern that directs calls out a video gateway.
Assign at least two different directory numbers on each video-enabled device in the Unified CM
cluster, with one line for audio and another line for video. The two lines can then be given different
calling search spaces. When users dial the access code (9, for example) on the first line, it could be
directed out a voice gateway, while dialing the same access code on the second line could direct the
call out a video gateway. This method alleviates the need for users to remember two different access
codes but requires them to press the correct line on their phones when placing calls. However not
all Cisco video endpoints support multiple lines at this time, in which case prefixes would be the
preferred method for routing outbound calls to the PSTN.
5-30
Chapter 5
Gateways
Best Practices for Gateways
128 kbps
192 kbps
256 kbps
320 kbps
384 kbps
512 kbps
768 kbps
1152 kbps
1472 kbps
Calls from an IP endpoint toward the PSTN can include the service prefix at the beginning of the called
number in order for the gateway to decide which service to use for the call. Optionally, you can configure
the default prefix to be used for calls that do not include a service prefix at the beginning of the number.
This method can become quite complex because users will have to remember which prefix to dial for the
speed of the call they wish to make, and you would have to configure multiple route patterns in
Unified CM (one for each speed).
Note
Two global settings on the Cisco TelePresence ISDN Gateway can be used to set a minimum or
maximum bandwidth value for incoming and outgoing ISDN calls. The dial plan cannot override this
value with a higher maximum bandwidth; however, a dial plan can impose a lower bandwidth for
particular calls.
Fail the call, playing busy tone to the caller and displaying a Bandwidth Unavailable message on the
callers screen
Use automated alternate routing (AAR) to re-route the call over an alternative path, such as a PSTN
gateway
The Retry Video Call as Audio option takes effect only on the terminating (called) device, thus allowing
the flexibility for the calling device to have different options (retry or AAR) for different destinations.
If a video call fails due to bandwidth limitations but automated alternate routing (AAR) is enabled,
Unified CM will attempt to reroute the failed call as a video call to the AAR destination. If AAR is not
enabled, the failed call will result in a busy tone and an error message will be sent to the caller. (See
Figure 5-12.)
5-31
Chapter 5
Gateways
Figure 5-12
Video call
initiated
Enough bandwidth
for video call?
Yes
Video call
No
Retry
Video as Audio
enabled on terminating
device?
Yes
Audio-only
call
No
No
AAR configured
and enabled on
both devices?
Yes
Video call
with AAR
successful?
No
Yes
119462
Call fails
To provide AAR for voice or video calls, you must configure the calling and called devices as members
of an AAR group and configure an External Phone Number Mask for the called device. The External
Phone Number Mask designates the fully qualified E.164 address for the users extension, and the AAR
group indicates what digits should be prepended to the External Phone Number Mask of the called device
in order for the call to route successfully over the PSTN.
For example, assume that user A is in the San Jose AAR group and user B is in the San Francisco AAR
group. User B's extension is 51212, and the External Phone Number Mask is 6505551212. The AAR
groups are configured to prepend 91 for calls between the San Jose and San Francisco AAR groups.
Thus, if user A dials 51212 and there is not enough bandwidth available to process the call over the IP
WAN between those two sites, Unified CM will take user B's External Phone Number Mask of
6505551212, prepend 91 to it, and generate a new call to 916505551212 using the AAR calling search
space for user A.
By default, all video-capable devices in Unified CM have the Retry Video Call as Audio option enabled
(checked). Therefore, to provide AAR for video calls, you must disable (uncheck) the Retry Video Call
as Audio option. Additionally, if a call admission control policy based on Resource Reservation Protocol
(RSVP) is being used between locations, the RSVP policy must be set to Mandatory for both the audio
and video streams.
Furthermore, Unified CM looks at only the called device to determine whether the Retry Video Call as
Audio option is enabled or disabled. So in the scenario above, user B's phone would have to have the
Retry Video Call as Audio option disabled in order for the AAR process to take place.
Finally, devices can belong to only one AAR group. Because the AAR groups determine which digits to
prepend, AAR groups also influence which gateway will be used for the rerouted call. Depending on
your choice of configuration for outbound call routing to the PSTN, as discussed in the previous section,
video calls that are rerouted by AAR might go out a voice gateway instead of a video gateway. Therefore,
carefully construct the AAR groups and the AAR calling search spaces to ensure that the correct digits
are prepended and that the correct calling search space is used for AAR calls.
While these considerations can make AAR quite complex to configure in a large enterprise environment,
AAR is easier to implement when the endpoints are strictly of one type or the other. When endpoints are
capable of both audio and video calls (such as Cisco Unified IP Phone 9971 or a Cisco TelePresence
5-32
Chapter 5
Gateways
Best Practices for Gateways
System EX90), the configuration of AAR can quickly become unwieldy. Therefore, Cisco recommends
that large enterprise customers who have a mixture of voice and video endpoints give careful thought to
the importance of AAR for each user, and use AAR only for select video devices such as dedicated
videoconference rooms or executive video systems. Table 5-3 lists scenarios when it is appropriate to
use AAR with various device types.
Table 5-3
Device Type
Enable AAR?
Comments
IP Phone
Yes
Other video-capable
devices only
Yes
No
Other video-capable
devices only
Yes
No
Least-Cost Routing
Least-cost routing (LCR) and tail-end hop-off (TEHO) are very popular in VoIP networks and can be
used successfully for video calls as well. In general, both terms refer to a way of configuring the call
routing rules so that calls to a long-distance number are routed over the IP network to the gateway closest
to the destination, in an effort to reduce toll charges. Unified CM supports this feature through its rich
set of digit analysis and digit manipulation capabilities, including:
Translation patterns
Configuring LCR for video calls is somewhat more complicated than for voice calls, for the following
reasons:
Video calls require their own dedicated gateways, as discussed previously in this chapter
With respect to dedicated gateways, the logic behind why you might or might not decide to use LCR for
video calls is very similar to that explained in the section on Automated Alternate Routing (AAR),
page 5-31. Due to the need to have different types of gateways for voice and video, it can become quite
complex to configure all the necessary partitions, calling search spaces, translation patterns, route
patterns, route filters, route lists, and route groups needed for LCR to route voice calls out one gateway
and video calls out another.
5-33
Chapter 5
Gateways
With respect to bandwidth requirements, the decision to use LCR depends on whether or not you have
enough available bandwidth on your IP network to support LCR for video calls to/from a given location.
If the current bandwidth is not sufficient, then you have to determine whether the benefits of video calls
are worth the cost of either upgrading your IP network to make room for video calls or deploying local
gateways and routing calls over the PSTN. For example, suppose you have a central site with a branch
office connected to it via a 1.544-Mbps T1 circuit. The branch office has twenty video-enabled users in
it. A 1.544-Mbps T1 circuit can handle at most about four 384-kbps video calls. Would it really make
sense in this case to route video calls up to the central site in order to save on toll charges? Depending
on the number of calls you want to support, you might have to upgrade your 1.544-Mbps T1 circuit to
something faster. Is video an important enough application to justify the additional monthly charges for
this upgrade? If not, it might make more sense to deploy a Cisco video gateway at the branch office and
not bother with LCR. However, placing local Cisco video gateways at each branch office is not
inexpensive either, so ultimately you must decide how important video-to-PSTN calls are to your
business. If video is not critical, perhaps it is not worth upgrading the bandwidth or buying video
gateways but, instead, using the Retry Video Call as Audio feature to reroute video calls as voice-only
calls if they exceed the available bandwidth. Once a call is downgraded to voice-only, local gateway
resources and bandwidth to perform LCR become more affordable and easier to configure.
The Gateways chapter of the Cisco Unified Communications System 9.0 SRND, available at
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/srnd/9x/gateways.html
5-34
CH A P T E R
A trunk is a communications channel on Cisco Unified Communications Manager (Unified CM) that
enables Unified CM to connect to other servers. Using one or more trunks, Unified CM can receive or
place voice, video, and encrypted calls, exchange real-time event information, and communicate in other
ways with call control servers and other external servers.
Trunks are an integral and crucial part of a Cisco Collaboration System deployment, hence it is important
to understand the types of trunks available, their capabilities, and design and deployment considerations
such as resiliency, capacity, load balancing, and so forth.
There are two basic types of trunks that can be configured in Unified CM:
SIP and H.323 trunks, both of which can be used for external communications
While H.323 trunks are still supported, SIP trunks are the recommended trunk type for Unified
Communication deployments because SIP trunks provide additional features and functionality not
available with H.323 trunks. This chapter provides a comparative overview of the capabilities of H.323
and SIP trunks, but the focus of this chapter is on SIP trunks, their operation, and features for Unified
Communications deployments. For detailed information on H.323 trunk capabilities and operation, refer
to the Cisco Unified CM Trunks chapter of the Cisco Collaboration 9.x SRND, available at
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/srnd/collab09/trunks.html
For more details on the applications of Unified CM trunks, refer to their respective sections in the
following chapters of this document:
6-1
Chapter 6
This chapter has been revised significantly for the current release of this document. Cisco recommends
that you read this entire chapter before attempting to deploy trunks in your Collaboration and Unified
Communications System.
Table 6-1 lists the topics that are new in this chapter or that have changed significantly from previous
releases of this document.
Table 6-1
Described in:
Revision Date
6-2
Chapter 6
Unified CME
Service Provider
IP PSTN
Service Provider
IP PSTN
Unified CM Cluster
Gateway
IP PBX
SIP Trunks
348639
Figure 6-1
Calls are directed toward trunks as defined by the dial plan, using the route pattern construct. A route
pattern can use a trunk either directly or through a route list. The route list, if used, consists of one or
more route groups, each of which contains one or more trunks. An individual trunk within a route group
may be configured to be selected in either a top-down or circular fashion. For outgoing calls, Unified CM
selects one of the trunks associated in this fashion with the route pattern. Before it accepts an incoming
call, Unified CM verifies whether a trunk is defined to the remote address from which the call is
received.
6-3
Chapter 6
Feature
SIP
H.323
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Alerting Name
Yes
No
Yes/Yes
Yes/Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
Call Hold/Resume
Yes
Yes
Yes
Yes
6-4
Chapter 6
Table 6-2
Feature
SIP
H.323
DTMF-relay
N/A
N/A
Yes
N/A
N/A
No
Codecs with MTP for SIP Early Offer/ H323 Fast All codecs supported
Start
when Early Offer
support for voice and
video calls Mandatory (insert
MTP if needed) or
Early Offer support
for voice and video
calls - Best Effort (no
MTP inserted) is
selected
Yes
Yes
Video codecs
Yes
No
Yes
Yes
T.38 Fax
Yes
Yes
Signaling Authentication
Digest, TLS
No
Signaling Encryption
TLS
No
SRTP
SRTP
Yes
No
Yes
No
Yes
Yes
Yes
Yes
6-5
Chapter 6
Table 6-2
Feature
SIP
H.323
Yes
Yes
Yes
Yes
Yes
Yes
OPTIONS Ping
Yes
No
Yes
No
Yes
Yes
Up to 16 Destination Addresses
Yes
Yes
Yes
No
Yes
Yes
1. H.323 trunks support signaling of RFC 2833 for certain connection types.
Early Offer support for voice and video calls Mandatory (insert MTP if needed)
Early Offer support for voice and video calls Best Effort (No MTP inserted) also known as Best
Effort Early Offer
Audio codec preference (and Accept Audio Codec Preference in Received Offer)
H.264 Video with Desktop Presentation (Binary Floor Control Protocol (BFCP)) and Far End
Camera Control (FECC)
The SIP trunk features available in the current release of Unified CM make SIP the preferred choice for
new and existing trunk connections. The QSIG over SIP feature provides parity with H.323 intercluster
trunks and can also be used to provide QSIG over SIP trunk connections to Cisco IOS gateways (and on
6-6
Chapter 6
to QSIG-based TDM PBXs). The ability to run on all Unified CM nodes and to handle up to
16 destination IP addresses improves outbound call distribution from Unified CM clusters and reduces
the number of SIP trunks required between clusters and devices. SIP OPTIONS ping provides dynamic
reachability detection for SIP trunk destinations, rather than per-call reachability determination. Early
Offer support for voice and video calls Mandatory (insert MTP if needed) and Best Effort Early
Offer eliminate the use of MTPs to create an Early Offer for voice, video, and encrypted calls over SIP
trunks. With Best Effort Early Offer, Unified CM sends only SIP Early Offer if the media
characteristics of the calling device can be determined (for example, a call from a SIP-based IP phone
over a Best Effort Early Offer trunk). If the media characteristics of the calling device cannot be
determined (for example, for an inbound SIP Delayed Offer call forwarded over a Best Effort SIP Early
Offer trunk), SIP Delayed Offer is sent instead.
SIP trunk normalization and transparency using Lua scripts improve native Unified CM interoperability
with third-party unified communications systems. Normalization allows inbound and outbound SIP
messages and SDP information to be modified on a per-SIP-trunk basis. Transparency allows
Unified CM to pass SIP headers, parameters, and content bodies from one SIP trunk call leg to another,
even if Unified CM does not understand or support the parts of the message that are being passed
through.
These features are discussed in detail later in this section.
For the complete list of new enhancements for SIP trunks, refer to the Cisco Unified Communications
Manager product release notes available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_release_notes_list.html
6-7
Chapter 6
Unified CM SIP trunks support both SIP Delayed Offer and SIP Early Offer. By default, SIP trunks are
configured as Delayed Offer and support voice, video and encrypted calls. For Early Offer calls, there
are three possible trunk configuration options:
MTP Required option selected on the SIP trunk An MTP is inserted for every call.
Early Offer support for voice and video calls Mandatory (insert MTP if needed) A SIP
Profile option, where Unified CM inserts a media termination point (MTP) if the media
characteristics of the calling device cannot be determined (for example, for an inbound Delayed
Offer call forwarded over an Early Offer SIP trunk).
Early Offer support for voice and video calls Best Effort (no MTP inserted) A SIP Profile
option, where an Early Offer is sent only if the media characteristics of the calling device can be
determined. If the media characteristics cannot be determined, a Delayed Offer is sent.
Unified CM Early Offer trunk configuration for Delayed Offer, Early Offer, and Best Effort Early Offer
is discussed in the section on Unified CM SIP Trunks Delayed Offer, Early Offer, and Best Effort Early
Offer, page 6-19.
INVITE
INVITE
100 Trying
100 Trying
180 Ringing
180 Ringing
348640
Figure 6-2
Note
SIP-based Cisco Unified IP Phones send Early Offer. (See Figure 6-3.)
6-8
Chapter 6
Figure 6-3
100 Trying
100 Trying
y g
180 Ringing
180 Ringing
g g
One-Way
Media
ACK
348641
Two-Way Media
ACK
y g
100 Trying
PRACK
PRACK
(
)
200 OK (PRACK)
200 OK (PRACK)
(
)
348642
Two-Way Media
6-9
Chapter 6
Figure 6-5
INVITE Supported:100rel
INVITE Supported:100rel
100 Trying
y g
100 Trying
(
)
PRACK with SDP (Answer)
(
)
200 OK (PRACK)
Note
348643
Two-Way Media
200 OK ((PRACK))
100 Trying Responses indicate that Unified CM has received the INVITE. 180 Ringing and 183 Session
in Progress Responses indicate that the user is being alerted of the call and are used to send information
about the called user in SIP header messages and, if PRACK is used, in the SDP content in SIP message
bodies.
Description
v=0
s=SIP Call
Session name
t=0 0
a=rtpmap:0 PCMU/8000
a=ptime:20
a=rtpmap:8 PCMA/8000
a=ptime:20
6-10
Chapter 6
Table 6-3
Description
a=rtpmap:18 G729/8000
G.729 codec
a=ptime:20
a=sendrecv
Media direction
a=rtpmap:101 telephone-event/8000
a=fmtp:101 0-15
The corresponding SDP Answer describes the media characteristics of the endpoint that receives the
Offer and the voice codec selected by the endpoint for two-way voice media (see Table 6-4).
Table 6-4
Description
v=0
s=SIP Call
Session name
t=0 0
a=rtpmap: 0 PCMU/8000
a=ptime:20
a=sendrecv
Media direction
a=rtpmap:101 telephone-event/8000
a=fmtp:101 0-15
Figure 6-6
SIP Trunk
10.10.199.250
10.10.199.130
RTP UDP Port 16444
G.711 Audio codec
Two-way Audio
RFC 2833 DTMF
10.10.199.251
Audio
10.10.199.179
RTP UDP Port 28668
G.71 Audio codec
Two-way Audio
RFC 2833 DTMF
6-11
Chapter 6
Description
v=0
Session name
t=0 0
Endpoint IP address
a=rtpmap:98 H264/90000
a=fmtp:99
QCIF=1;CIF=1;CIF4=1;CUSTOM=352,240,1
Notice that the H.264 and H.263 codecs offered in this SDP message contain a range of additional
parameters that describe the receive capabilities of the endpoint. As shown in Table 6-6 for the
negotiated H.264 codec in the SDP Answer, these parameters do not need to be symmetrical.
6-12
Chapter 6
Table 6-6
Description
v=0
Session name
t=0 0
Endpoint IP address
a=rtpmap:98 H264/90000
a=fmtp:98
profile-level-id=428016;packetization-mode=1;m
ax-mbps=108000;max-fs=3600;max-cpb=200;ma
x-br=5000;max-rcmd-nalu-size=1382400;max-s
mbps=108000;max-fps=6000
The Profile Level ID and Packetization Mode must be symmetrical for the negotiated video call. The
Profile Level ID describes a minimum subset of H.264 features, resolution, frame rate, and bit rate
supported by the endpoint. The Packetization Mode describes how video samples can be encapsulated
and sent in RTP packets. The media attributes, which follow the Profile Level ID and Packetization
Mode, need not be symmetrical and indeed are not all symmetrical for the negotiated video call shown
in Table 6-6 and Figure 6-7.
6-13
Chapter 6
Figure 6-7
SIP Trunk
10.10.199.250
10.10.199.130
RTP UDP Port 16444
G.711 Audio codec
Two-way Audio
RFC 2833 DTMF
10.10.199.130
RTP UDP Port 16446
H264 Video codec
Asymmetric Receive values
RTCP Feedback
10.10.199.251
Audio
Video
10.10.199.179
RTP UDP Port 28668
G.71 Audio codec
Two-way Audio
RFC 2833 DTMF
10.10.199.179
RTP UDP Port 28670
H264 Video codec
Asymmetric Receive values
RTCP Feedback
6-14
Chapter 6
Audio and Video Call with Presentation Sharing and Far End Camera Control
348646
Figure 6-8
SIP Trunk
10.10.199.250
10.10.199.251
10.10.199.179
10.10.199.130
Audio RTP UDP Port 16444
G711 Audio codec
RFC 2833 DTMF
Video RTP UDP Port 16446
H.264 Video codec
Video RTP UDP Port 16448
H.264 Video codec
BFCP UDP Port 5070
FECC UDP Port 16450
RTP Payload Type 107
Audio
Main Video
Slide Video
Binary Floor Control
Far End Camera Control
SIP Trunks Run on All Nodes and the Route Local Rule
When the Run on all Active Unified CM Nodes option is checked on a SIP trunk, Unified CM creates
an instance of the SIP trunk daemon on every call processing subscriber within the cluster, thus allowing
a SIP trunk call to be made or received on any call processing subscriber. (Prior to this feature, up to
three nodes could be selected per trunk by using Unified CM Groups.) With Run on all Active
Unified CM Nodes enabled, outbound SIP trunk calls originate from the same node on which the
inbound call (for example, from a phone or trunk) is received (based on the Route Local rule). The Run
on all Active Unified CM Nodes feature overrides the trunk's Unified CM Group configuration.
6-15
Chapter 6
Route Lists Run on All Nodes and the Route Local Rule
Although this is not specifically a SIP trunk feature, running route lists on all nodes provides benefits
for trunks in route lists and route groups. Running route lists on all nodes improves outbound call
distribution by using the Route Local rule to avoid unnecessary intra-cluster call setup traffic.
For route lists, the Route Local rule operates as follows:
For outbound calls that use route lists (and associated route groups and trunks), when a call from a
registered phone or inbound trunk arrives at the node with the route list instance, Unified CM checks
to see if an instance of the selected outbound trunk exists on the same node as the route list. If so,
Unified CM uses this node to establish the outbound trunk call.
If both the route list and the trunk have Run on all Active Unified CM Nodes enabled, outbound call
distribution will be determined by the node on which the inbound call arrives. If the selected outbound
trunk uses Unified CM Groups instead of running on all nodes, Unified CM applies the Route Local rule
if an instance of the selected outbound trunk exists on the same node on which the inbound call arrived.
If an instance of the trunk does not exist on this node, then Unified CM forwards the call (within the
cluster) to a node where the trunk is active.
If the route list does not have Run on all Active Unified CM Nodes enabled, an instance of the route
list will be active on one node within the cluster (the primary node of the route list's Unified CM Group).
If the selected outbound trunk is also active on the primary node of the route list's Unified CM Group,
the Route Local rule will apply, resulting in sub-optimal outbound call distribution because all outbound
trunk calls will originate from this node.
Cisco strongly recommends enabling Run on all Active Unified CM Nodes on all route lists and SIP
trunks.
6-16
Chapter 6
fields. Use a single SIP trunk with one or more destination addresses to connect a Unified CM cluster to
one other unified communications system. If trunk fail-over is required, create an additional trunk to the
fail-over unified communications system and use route lists and route groups to order trunk selection.
Unified CM randomly distributes outbound calls over the configured SIP trunk destination addresses.
Figure 6-9
SIP Trunks with Run on All Unified CM Nodes and Multiple Destination Addresses
348647
SIP
Figure 6-10
SIP Trunks and Route Lists with Run on All Unified CM Nodes Enabled
Trunk A
SIP
348648
Trunk B
6-17
Chapter 6
Note
If the configuration option Destination Address is an SRV is selected, only a single SRV entry can be
added as the trunk destination. (For example, Destination Address = cluster1.cisco.com. Port = 0.)
Figure 6-11 shows the call flow for a SIP trunk using DNS SRV to resolve the addresses to a destination
Unified CM cluster. However, this destination could also be a third-party unified communications
system.
Figure 6-11
DNS Server
6
Cluster 2 FQDN :
cluster2.foo.com
Cluster 1 FQDN :
cluster1.foo.com
CUCM-1
CUCM-A
CUCM-2
CUCM-B
CUCM-3
CUCM-C
CUCM-4
CUCM-D
Calls 87522001
86453001
87522001
348649
6-18
Chapter 6
Note
1.
2.
The call matches a route pattern of 8752XXXX that is pointing to the SIP trunk with DNS SRV of
cluster2.foo.com. CUCM-4 in Cluster 1 is the node handling this call because the phone and the SIP
trunk are both registered to it. CUCM-4 sends a DNS SRV lookup for cluster2.foo.com.
3.
4.
CUCM-4 sends a SIP Invite to 87522001@cluster2.foo.com, with destination address set to the IP
address of CUCM-D.
5.
Unified CM interprets this call as a local call because the host portion of the uniform resource
identifier (URI) matches the Cluster FQDN enterprise parameter. Cluster 2 does not have any SIP
trunk configured with a destination of CUCM-4, so it does a DNS SRV lookup for all domains
configured under the SIP trunks with DNS SRV. In this case, the example shows a single trunk with
a DNS SRV destination of cluster1.foo.com.
6.
The DNS server returns two entries, and one of them matches the source IP address of the Invite.
The cluster accepts the call and extends it to extension 87522001.
Unified CM SIP Trunks Delayed Offer, Early Offer, and Best Effort Early Offer
This section provides guidance on the use of Delayed Offer, Early Offer, and Best Effort Early Offer with
Unified CM SIP trunks.
6-19
Chapter 6
in the Answer sent in response to a received Offer. However, MTPs may be used to address DTMF
transport mismatches. Use this default configuration if you want all calls sent over the SIP trunk to send
Delayed Offer. Voice, video and encrypted calls are supported.
Early Offer support for voice and video calls Mandatory (insert MTP if needed)
SCCP Line
SIP Trunk
H.323 Trunk
MGCP Trunk
348650
Figure 6-12
Early Offer Using Early Offer support for voice and video calls Mandatory (insert MTP if needed)
Enabling Early Offer support for voice and video calls Mandatory (insert MTP if needed) on the
SIP Profile associated with the SIP trunk inserts an MTP only if the calling device cannot provide
Unified CM with the media characteristics required to create the Early Offer. In general, Early Offer
support for voice and video calls Mandatory (insert MTP if needed) is recommended over Media
Termination Point Required because this configuration option reduces MTP usage and can support
voice, video, and encrypted calls. (see Figure 6-13).
6-20
Chapter 6
For outbound calls over a SIP trunk configured as Early Offer support for voice and video calls
Mandatory (insert MTP if needed), Unified CM inserts an MTP to create an SDP Offer in the following
cases only:
As a general rule, Early Offer calls of this type that use MTPs support voice only, but they are not limited
to a single voice codec (as they are with Early Offer using MTP Required). These calls support only
audio in the initial call setup but can be escalated mid-call to support video and SRTP if the call media
is renegotiated (for example, after hold or resume).
Note
Figure 6-13
MTP resources are not required for incoming INVITE messages, whether or not they contain an initial
Offer SDP.
Early Offer support for voice and video calls - Mandatory (insert MTP if needed)
SIP Line
SCCP Line
SCCP Line
SIP Trunk
H.323 Trunk
MGCP Trunk
H.323 Trunk
SIP Trunk
MGCP Gateway
348651
6-21
Chapter 6
Unified CM does not need to insert an MTP to create an outbound Early Offer call over a SIP trunk if
the inbound call to Unified CM is received by any of the following means:
On an MGCP trunk
For the above devices, Unified CM uses the media capabilities of the endpoint and applies the codec
filtering rules based on the region-pair of the calling device and outgoing SIP trunk to create the offer
SDP for the outbound SIP trunk call. In most cases, the offer SDP will have the IP address and port
number of the endpoint initiating the call. This is assuming that Unified CM does not have to insert an
MTP for other reasons such as a DTMF mismatch, TRP requirements, or a transcoder requirement when
there is no common codec between the regions of the calling device and the SIP trunk.
When Early Offer support for voice and video calls Mandatory (insert MTP if needed) is configured
on a trunk's SIP Profile, calls from older SCCP-based phones, SIP Delayed Offer trunks, and H.323 Slow
Start trunks will cause Unified CM to allocate an MTP. The MTP is used to generate an offer SDP with
a valid media port and IP address. The MTP will be allocated from the media resources associated with
the calling device rather than from the outbound SIP trunk's media resources. (This prevents the media
path from being hair-pinned via the outbound SIP trunk's MTP). If the MTP cannot be allocated from
the calling device's media resource group list (MRGL), then the MTP allocation is attempted from the
SIP trunk's MRGL.
For calls from older SCCP phones registered to Unified CM, some of the media capabilities of the
calling device (for example, supported voice codecs, video codecs, and encryption keys if supported) are
available for media exchange through the Session Description Protocol (SDP). Unified CM will create
a superset of the endpoint and MTP codec capabilities and apply the codec filtering based on the
applicable region-pair settings. The outbound Offer SDP will use the MTP's IP address and port number
and can support voice, video, and encrypted media. Note that a Cisco IOS-based MTP should be used
and configured to support the pass-through codec.
Note
Older SCCP-based IP phones such as the Cisco Unified IP Phone 7902, 7905, 7910, 7912, 7920, 7935,
7940, and 7960 require the use of an MTP when they make calls over a SIP trunk with the Early Offer
for voice and video Mandatory (insert MTP if needed) feature enabled. If you have a significant
number of these phone types deployed in a cluster, consider deploying Delayed Offer trunks instead of
Early Offer for voice and video Mandatory (insert MTP if needed). If Early Offer for voice and
video Mandatory (insert MTP if needed) trunks are used, provision MTP resources in the cluster
equivalent to the number of busy hour calls over those SIP trunks that use this Early Offer feature.
When Unified CM receives an inbound call on an H.323 Slow Start or SIP Delayed Offer trunk, the
media capabilities of the calling device are not available when the call is initiated. In this case,
Unified CM must insert an MTP and will use its IP address and UDP port number to advertise all
supported audio codecs (after region pair filtering) in the Offer SDP of the initial INVITE sent over the
outbound SIP trunk. When the Answer SDP is received on the SIP trunk, if it contains a codec that is
supported by the calling endpoint, then no additional offer-answer transaction is needed. In case of codec
mismatch, Unified CM can either insert a transcoder to address the mismatch or send a Re-INVITE or
UPDATE to trigger media negotiation. Calls from H.323 Slow Start or SIP Delayed Offer trunks support
only audio in the initial call setup, but they can be escalated mid-call to support video and SRTP if the
call media is renegotiated (for example, after Hold or Resume).
6-22
Chapter 6
Best Effort Early Offer [Early Offer support for voice and video calls Best Effort (no MTP inserted)]
Best Effort Early Offer can be enabled on the SIP Profile associated with the SIP trunk, and it is the
recommended configuration for all Unified CM and Unified CM Session Management Edition (SME)
trunks. Best Effort Early Offer trunks never use MTPs to create an Early Offer and, depending on the
calling device, may initiate an outbound SIP trunk call using either Early Offer or Delayed Offer. Best
Effort Early Offer SIP trunks support voice, video, and encrypted calls.
Best Effort Early Offer SIP trunks send outbound calls with an Early Offer (INVITE with SDP content)
in the following situations:
An inbound call to Unified CM or SME is received over a SIP trunk using Early Offer.
An inbound call to Unified CM or SME is received over an H.323 trunk using Fast Start.
A call is initiated from a newer model SCCP-based Cisco Unified IP Phone registered to
Unified CM.
Best Effort Early Offer trunks send outbound calls with a Delayed Offer (INVITE without SDP
content) in the following situations:
An inbound call to Unified CM or SME is received over a Delayed Offer SIP trunk.
An inbound call to Unified CM or SME is received over an H.323 Slow Start trunk.
A call is initiated from an older model SCCP-based IP phone registered to Unified CM.
6-23
Chapter 6
Figure 6-14
SCCP Line
SCCP Line
SIP Trunk
H323 Trunk
MGCP Trunk
H323 Trunk
SIP Trunk
MGCP Gateway
348652
Media resources such as MTPs for DTMF translation, trusted relay points (TRPs), and transcoders for
codecs mismatches can still be associated with and used by a Best Effort Early Offer trunk. Note that
with Best Effort Early Offer, MTPs are never used to create an Early Offer or to create an Answer in
response to a received Offer.
Using Best Effort Early Offer for all SIP trunks in your enterprise simplifies Cisco Collaboration
System network design and deployments, and it eliminates the need to use MTPs to generate an Offer.
Note, however, that Cisco Collaboration call control systems, applications, and gateways may receive
either an Early Offer or Delayed Offer call over a Best Effort Early Offer trunk, and they should be
able to receive either. All Cisco Collaboration System applications support the receipt of either Early
Offer or Delayed Offer calls.
In certain cases (for example, calls via a Cisco Unified Border Element Session Border Controller (SBC)
to a service provider's IP PSTN), Early Offer must always be sent to the IP PSTN. In these situations,
use Cisco Unified Border Elements Delayed Offer to Early Offer feature to convert a received Delayed
Offer to Early Offer.
If your Cisco Collaboration System application must receive either Early Offer only or Delayed Offer
only, you can use a Unified CM SIP trunk configured for Early Offer (using Early Offer support for
voice and video calls Mandatory (insert MTP if needed) or MTP Required) or Delayed Offer,
respectively, to connect to this application. With single Unified CM cluster deployments, these trunk
6-24
Chapter 6
choices are straightforward. For multi-cluster deployments interconnected via Unified CM Session
Management Edition, where a single SIP trunk can be shared to reach many end Cisco Collaboration
Systems, Best Effort Early Offer is recommended for all SME trunks. For more information on the
design considerations for Best Effort Early Offer, see Summary of SIP Trunk Recommendations for
Multi-Cluster SME Deployments, page 6-48.
MTP-Less Early Offer, Best Effort Early Offer, and SME Media Transparency
MTP-Less Early Offer is a special SIP trunk configuration for Unified CM Session Manager Edition
(SME) cluster versions that do not support the Best Effort Early Offer feature. Best Effort Early Offer
provides the same functionality as MTP-Less Early Offer; but whereas MTP-Less Early Offer
deployments require that no media resources are configured on the SME cluster, with Best Effort Early
Offer, media resources can be configured if needed. SME deployments using only Best Effort Early
Offer or MTP-Less Early Offer SIP trunks allow you to deploy an SME cluster that is media
transparent (no media resources are required in the SME cluster) because all media negotiation takes
place in the leaf Unified Communications systems, which insert media resources (MTPs, transcoders,
and so forth) as required. (See Figure 6-15.)
MTP-Less Early Offer takes advantage the Unified CM SIP service parameter Fail Call Over SIP
Trunk if MTP Allocation Fails. The default setting for this service parameter is False, thus allowing
an inbound Delayed Offer call to proceed over the outbound SIP trunk (configured for Early Offer) as a
Delayed Offer call if no MTP resources are available.
Figure 6-15
Delayed Offer
Delayed Offer
SME
Early Offer
SIP
MTP-Less Early Offer
Early Offer
Leaf Cluster
SME Cluster
with no
Media Resources
348653
Leaf Cluster
Enable all trunks with Early Offer support for voice and video calls Mandatory (insert MTP if
needed).
Disable the IPVMS service on all SME nodes. This disables Unified CM media termination points,
conferencing, music on hold, and annunciator resources.
6-25
Chapter 6
Figure 6-16
Do not associate any Cisco IOS media resources with the SME cluster.
Enable Accept Audio Codec Preference in Received Offer on all SME SIP trunks.
Delayed Offer
Delayed Offer
SME
Early Offer
SIP
Best Effort Early Offer
Early Offer
Leaf Cluster
SME Cluster
348654
Leaf Cluster
Note
Enable Accept Audio Codec Preference in Received Offer on all SME SIP trunks.
Media resources can be deployed in an SME cluster where Best Effort Early Offer SIP trunks are
configured, but these resources will be used only if one or more SIP trunks are configured as Delayed
Offer or Early Offer. In these cases, calls to and from Early Offer or Delayed Offer trunks are not media
transparent and can invoke media resources if a DTMF or codec mismatch is encountered.
6-26
Chapter 6
Software MTPs in Cisco IOS gateways Available with any Cisco IOS T-train software release and
scaling up to 5,000 sessions (calls) on the Cisco Aggregation Services Routers (ASR) 1000 Series
with Route Processor RP2.
Hardware MTPs in Cisco IOS gateways Available with any Cisco IOS T-train software release,
hardware MTPs use on-board DSP resources and scale calls according to the number of DSPs
supported on the Cisco router platform.
Cisco Unified CM software MTPs using the Cisco IP Voice Media Streaming Application on a
Unified CM subscriber node
Cisco IOS MTPs are recommended over Unified CM MTPs because Cisco IOS MTPs provide additional
scalability and greater functionality, such as support for additional codec types, multiple media streams,
and the pass-through codec. (For details, see the section on Media Termination Point (MTP), page 7-7.)
The following example configuration is for a Cisco IOS software MTP:
!
sccp local Vlan5
sccp ccm 10.10.5.1 identifier 5 version 8.6.2
! Communications Manager IP address (10.10.5.1)
sccp
!
sccp ccm group 5
bind interface Vlan5
associate ccm 5 priority 1
associate profile 5 register MTP000E83783C50
! MTP name (MTP000E83783C50) ... must match the Unified CM MTP name.
!
dspfarm profile 5 mtp
description software MTP
codec g711ulaw
codec pass-through
maximum sessions software 500
associate application SCCP
6-27
Chapter 6
Because in-band signaling methods send DTMF tones in the RTP media stream, the SIP endpoints in a
session must either support the transport method used (for example, RFC 2833) or provide a method of
intercepting this in-band signaling and converting it. If the two endpoints are using a back-to-back user
agent (B2BUA) server for the call control (for example, Cisco Unified CM) and the endpoints negotiate
different DTMF methods between each device and call control agent, then the call control agent
determines how to handle the DTMF differences, either through MTP insertion or by OOB methods.
With Unified CM, a DTMF transport mismatch (for example, in-band to out-of-band DTMF) is resolved
by inserting a media termination point (MTP), which terminates the RTP stream with in-band DTMF
signaling (RFC 2833), extracts the DTMF tones from the RTP stream, and forwards these tones
out-of-band to Unified CM, where they are then forwarded to the endpoint supporting out-of-band
signaling. For DTMF mismatches, the inserted MTP is always in the media path between the two
endpoints. In-band DTMF packets are identified by their RTP Payload type, extracted by Unified CM,
and converted to out-of-band DTMF, while RTP media packets pass transparently through the MTP.
In-band DTMF tones can also be transported as raw (audible) tones in the RTP media stream. This
transport method is not widely supported by Cisco products and, in general, is not recommended as an
end-to-end DTMF transport mechanism. In-band audio DTMF tones can generally be reproduced
reliably when using high-bandwidth codecs such as G.711 a-law or mu-law, but they are not suitable for
use with low-bandwidth codecs such as G.729. In cases where in-band audio is the only available DTMF
transport mechanism, the Cisco Unified Border Element can be used to translate the in-band audio
DTMF signaling into RFC 2833 signaling.
Three DTMF options are available on Unified CM SIP trunks:
Cisco recommends configuring the DTMF Signaling Method to No Preference on Unified CM SIP
trunks. This setting allows Unified CM to make an optimal decision for DTMF and to minimize MTP
allocation.
6-28
Chapter 6
Region A
Codecs
Supported
by device
G.711
G.729
Region A-C
Maximum
Audio
Bitrate
64kbps
Region C
SIP INVITE
Codec Preference in Offer SDP
(1) G.711
(2) G.729
Figure 6-17
6-29
Chapter 6
Region A
Codecs
Supported
by device
G.711
G.729
Region C
SIP INVITE
Codec Preference in Offer SDP
(1) G.729
Region A-C
Maximum
Audio
Bitrate
48kbps
Figure 6-18
For calls between two Unified CM clusters over SIP intercluster trunks, audio codec preference lists
allow the codec to be selected for a call based upon the codec preferences of the calling and called
devices. By grouping devices in each cluster into regions based on their codec preferences, a single
intercluster trunk can be used to support multiple calls, with each call type using its preferred codec.
(See Figure 6-19.)
Region A
Audio Codec Preference Lists for Voice and Fax Calls Between Two Unified CM Clusters
Region A-C
Codec
Preference
List
(1) G.729
(2) G.711
(3) G.722
(4) ..
Region C
G.729 Voice Call
Phones
Region B
Fax
Region E
Region E-F
Codec
Preference
List
(1) G.729
(2) G.711
(3) G.722
(4) ..
Phones
SIP Trunk
Region B-C
Codec
Preference
List
(1) G.711
(2) G.722
(3) G.729
(4) ..
Cluster 1
Region F
Cluster 2
Region E-G
Codec
Preference
List
(1) G.711
(2) G.722
(3) G.729
(4) ..
Region G
Fax
348657
Figure 6-19
6-30
Chapter 6
Note
Configure equivalent inter-region audio codec preference lists for each device type in each cluster to
ensure that a common codec is selected for each device type, irrespective of call direction or trunk
configuration. If the audio codec preference lists in each cluster are not equivalent, the codecs used per
call can vary based on call direction and trunk configuration. (Ordinarily, the codec preference order is
not honoured by the cluster receiving the codec preference list.)
Note
Do not use SIP trunks configured for Early Offer with MTP Required enabled if codec preference is
required. This trunk configuration inserts an MTP for inbound and outbound calls, which is limited to a
single audio codec only, thereby overriding codec preference and selection.
SME Deployment Using "Accept Audio Codec Preferences in Received Offer" on SIP Trunks
Region A-C
Codec
Preference
List
(1) G.729
(2) G.711
(3) G.722
(4) ..
Region C
Fax
Region E
G.729 Call
Phones
Region B
Region D
Region E-F
Codec
Preference
List
(1) G.729
(2) G.711
(3) G.722
(4) ..
Phones
SIP Trunk
Region B-C
Codec
Preference
List
(1) G.711
(2) G.722
(3) G.729
(4) ..
G.711 Call
Cluster 1
SME
Cluster
Region F
Cluster 2
Region E-G
Codec
Preference
List
(1) G.711
(2) G.722
(3) G.729
(4) ..
Region G
Fax
348658
Region A
6-31
Chapter 6
Note
The Accept Audio Codec Preferences in Received Offer feature is available only on SIP trunks (a SIP
Profile feature). This feature does not offer consistent results if used in an SME deployment where the
SME cluster uses a combination of SIP, H.323, and/or MGCP trunks. Therefore, the Accept Audio
Codec Preferences in Received Offer feature should be used when the SME cluster is deployed using
only SIP trunks.
Cisco Unified CM and Cisco Unified Border Element SIP Trunk Codec Preference
Unified CM audio codec preference lists can be used in Unified Communications deployments with
Cisco Unified Border Element to simplify configuration of SIP trunks between Unified CM and Unified
Border Element. For example, instead of using dedicated SIP trunks to the Unified Border Element for
voice and fax calls, a single Unified CM SIP trunk can be used where the codec preference for each
device type is honored as calls pass through the Unified Border Element.
In Figure 6-21 the Voice Class Codec Preference lists defined on Cisco Unified Border Elements
inbound and outbound dial peers do not change the preference of the listed codecs in the received Offer.
Cisco Unified Border Element does codec filtering on the received Offer, both on the inbound and
outbound dial-peer, and passes across the common codecs in the same preference order as received in
the inbound Offer to the peer leg.
If codecs, in addition to those received in an Offer, are defined in the voice class codec list, then these
codecs will be appended to those received in the ordered list and sent out in the outbound Offer.
Thus, a single inbound and outbound dial-peer can be configured on Cisco Unified Border Element for
all device types. Cisco recommends using the same voice class codec preference list for both the inbound
and outbound dial-peer, with that list containing the codecs that you want to negotiate with the service
provider. As mentioned above, the order of the codecs will be dictated first by the order received in the
inbound Offer and then by the order defined in the voice class codec preference list.
Region A
Cisco Unified CM and Cisco Unified Border Element SIP Trunk Codec Preference
Region A-C
Codec
Preference
List
(1) G.729
(2) G.711
(3) G.722
(4) ..
Region C
Phones
SIP Trunk
Region B
Fax
Outbound
Dial Peer
Voice Class
Codec
Preference
List
(1) G.729
(2) G.711
(3) G.722
(4) ..
Outbound
Dial Peer
Voice Class
Codec
Preference
List
(1) G.729
(2) G.711
(3) G.722
(4) ..
Phones
Region B-C
Codec
Preference
List
(1) G.711
(2) G.722
(3) G.729
(4) ..
Cluster 1
Fax
348659
Figure 6-21
6-32
Chapter 6
Note
Although TCP is the recommended transport protocol within a Cisco Collaboration Systems network,
most service providers prefer to use UDP because it has a lower processing overhead than TCP. Cisco
Unified Border Element can be used to provide TCP-based SIP trunk connections to the Cisco
Collaboration Systems network and UDP-based SIP trunk connections to service provider networks.
Configuring the trunk to encrypt media (see Media Encryption, page 6-33)
Configuring the trunk to encrypt signaling (see Signaling Encryption, page 6-33)
Media Encryption
Media encryption can be configured on SIP trunks by checking the trunk's SRTP allowed check box. It
is important to understand that enabling SRTP allowed causes the media for calls to be encrypted, but
the trunk signaling will not be encrypted and therefore the session keys used to establish the secure
media stream will be sent unencrypted. It is therefore important that you ensure that signaling between
Unified CM and its destination SIP trunk device is also encrypted so that keys and other security-related
information do not get exposed during call negotiations.
Signaling Encryption
SIP trunks use TLS for signaling encryption. TLS is configured on the SIP Security Profile associated
with the SIP trunk, and it uses X.509 certificate exchanges to authenticate trunk devices and to enable
signaling encryption.
Certificates can be either of the following:
Imported to each Unified CM node from every device that wishes to establish a TLS connection to
that node's SIP trunk daemon
Signed by a Certificate Authority (CA), in which case there is no need to import the certificates of
the remote devices; only the CA certificate needs to be imported
6-33
Chapter 6
Unified CM provides a bulk certificate import and export facility. However, for SIP trunks using Run on
all Active Unified CM Nodes and up to 16 destination addresses, using a Certificate Authority provides
a centralized and less administratively burdensome approach to setting up signaling encryption on SIP
trunks.
For more information on TLS for SIP trunks, refer to the latest version of the Cisco Unified
Communications Manager Security Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
For information on certificate authorities, refer to the Certificate Authority (CA) information in the latest
version of the Cisco Unified Communications Operating System Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
If the system can establish a secure media or signaling path and if the end devices support SRTP, the
system uses an SRTP connection. If the system cannot establish a secure media or signaling path, or if
at least one device does not support SRTP, the system uses an RTP connection. SRTP-to-RTP fall-back
(and vice versa) may occur for transfers from a secure device to a non-secure device or for conferencing,
transcoding, music on hold, and so on.
For SRTP-configured devices, Unified CM classifies a call as encrypted if the SRTP Allowed check box
is checked for the device and if the SRTP capabilities for the devices are successfully negotiated for the
call. If these criteria are not met, Unified CM classifies the call as non-secure. If the device is connected
to a phone that can display security icons, the phone displays the lock icon when the call is encrypted.
Note
MTPs that are statically assigned to a SIP trunk by means of the MTP Required checkbox do not
support SRTP because they do not support the pass-through codec.
To ensure that SRTP is supported for all calls, configure the SIP trunk for Delayed Offer or Best Effort
Early Offer.
Where Early Offer support for voice and video calls Mandatory (insert MTP if needed) is
configured for devices that support encryption, all calls that do not need to use MTPs can support SRTP.
When an MTP is inserted into the call path, this dynamically inserted MTP supports the pass-through
codec, and encrypted calls are supported in the following cases:
If the calling device is an older SCCP-based phone registered to Unified CM, SRTP can be
negotiated in the initial call setup.
If the call arrives inbound to Unified CM on a Delayed Offer SIP trunk or an H.323 Slow Start trunk,
SRTP will not be negotiated in the initial call setup because no security keys are available, but the
call can be escalated mid-call to support SRTP if the call media is renegotiated (for example, after
hold or resume).
If Unified CM dynamically inserts an MTP for reasons other than Early Offer, such as for a Trusted
Relay Point or as an RSVP agent, then SRTP will be supported with an MTP that supports the
pass-through codec (Cisco IOS MTPs).
Note
In-band to out-of-band DTMF conversion using MTPs does not function for SRTP encrypted media
streams because the MTP is unable to decrypt the DTMF packets.
6-34
Chapter 6
Examples
From:
To:
P-Asserted-Identity:
Remote-Party-ID:
The From and To message headers sent in SIP Requests and Responses indicate the direction of the call.
(The From header represents the calling user and the To header represents the called user.) The From and
To headers remain the same in all SIP Requests and Responses for the call.
SIP allows the From header to be made anonymous so that the calling user information is not presented
to the called user.
The P-Asserted-Identity and Remote-Party-ID headers (if present) always contain the user's identity. The
user information contained in SIP messages with these identity headers is directional, so that the headers
contain the calling user's identity in an Initial INVITE and the called user's identity in Responses. The
P-Asserted-Identity and Remote-Party-ID headers can be used to trace the identity of an anonymous call.
By default, both the P-Asserted-Identity and Remote-Party-ID headers are sent over Unified CM SIP
trunks, but they can be disabled. The usage of P-Asserted-Identity and Remote-Party-ID headers will
depend upon the device that the Unified CM SIP trunk is connected to. P-Asserted-Identity is a more
recent standard and more commonly used than Remote-Party-ID. The P-Asserted-Identity standard
(RFC 3325) is considered to be more secure than Remote-Party-ID because it supports authentication
between untrusted SIP Realms. For SIP trunk connections to untrusted networks, configure Unified CM
to send a P-Preferred-Identity header instead of a P-Asserted-Identity header. Unified CM will respond
to a Digest authentication challenge for the sent the P-Preferred-Identity header.
By configuring calling name and calling number presentation or restriction in a translation pattern
associated with the calling device
By configuring calling name and calling number presentation or restriction on the Unified CM trunk
By configuring the P-Asserted-Identity related, SIP Privacy value on the Unified CM SIP trunk
These caller ID presentation and restriction configuration options operate in the following precedence
order (highest precedence first):
1.
2.
Trunk configuration
3.
Device configuration
6-35
Chapter 6
Calls within ISDN and Q.931 PSTN networks provide additional information in the Number Type fields
of call setup messages to classify called and calling numbers. Number-types can be one of four types:
Unknown, Subscriber, National, or International. For calls from the PSTN to the enterprise network, the
number-type parameter can be used by the enterprise to globalize the calling number to its +E.164 value
by prefixing it with the appropriate digits. Using a globalized PSTN calling number within the enterprise
allows calls to be returned to the PSTN caller with little or no additional digit manipulation. Depending
on the number format sent by the service provider, the enterprise called number might also have to be
modified to match that of the enterprise dial plan. Cisco recommends deploying a +E.164 dial plan
within the enterprise.
For more details and examples on how these number-types are used and dial plan recommendations, refer
to the chapter on Dial Plan, page 14-1.
SIP-Based IP PSTN Networks
Calls from SIP-based IP PSTN networks do not include number type information in SIP messages. In
this case, the IP PSTN service provider should present the PSTN calling number using a globally
routable international representation (for example, a +E.164 number). Depending on the number format
sent by the service provider, the enterprise called number might have to be modified to match that of the
enterprise dial plan. Cisco recommends deploying a +E.164 dial plan within the enterprise.
If the service provider sends the PSTN calling number in +E.164 format and the called number in a
format that matches that used by the enterprise dial plan (+E.164 recommended), then little or no
changes need to be made to these numbers within the enterprise.
The inability of SIP to transport the number type implies that the normalization of the calling number
must be performed before the call is presented to Unified CMs call routing process. One place where
the transformation can be performed is on the ingress SIP gateway. The following example configuration
shows the translation rules that can be defined on a Cisco IOS gateway to accomplish this
transformation:
voice translation-rule 1
rule 1 // /+4940/ type subscriber subscriber
rule 2 // /+49/ type national national
rule 3 // /+/ type international international
...
voice translation-profile 1
translate calling 1
...
dial-peer voice 300 voip
translation-profile outgoing 1
destination-pattern .T
session protocol sipv2
session target ipv4:9.6.3.12
...
When configured as in the example above, a Cisco IOS gateway using SIP to communicate with
Unified CM will send calling party information digits normalized to the E.164 format, including the +
sign. The Unified CM configuration will receive all calls from this gateway with a numbering type of
"unknown" and will not need to add any prefixes.
6-36
Chapter 6
For more details on configuring translation rules, refer to the Voice Translation Rules document,
available at
http://www.cisco.com/en/US/tech/tk652/tk90/technologies_tech_note09186a0080325e8e.shtml
Unified CM can set the calling party number of outgoing calls to the normalized global format. The
number-type in outgoing calls from the SIP trunk will be "unknown," and the Cisco IOS gateway should
change it to International if no stripping is done, or perform a combination of stripping and numbering
type change if required by the connected service provider.
SIP OPTIONS Ping which tracks the overall operational status of the trunk and the state of each
trunks destination nodes.
Codec Preference Lists and the ability to accept the codec preferences received in an SDP Offer.
Support for H.264 video with BFCP-based presentation sharing and Far End Camera Control.
SIP message normalization and transparency, which provides powerful script-based functionality
for SIP trunks that can be used to transparently forward and/or modify SIP messages and message
body (SDP) contents as they traverse Unified CM. Normalization and transparency scripts are
designed to address SIP interoperability issues, allowing Unified CM to interoperate with SIP-based
third-party PBXs, applications, and IP PSTN services.
Support for IPv4 only, IPv6 only, or Dual Stack (IPv4 and IPv6) ANAT-enabled SIP trunks.
Using only SIP trunks configured as Best Effort Early Offer, eliminates MTP usage for Early Offer
creation in leaf clusters and makes SME clusters transparent to media negotiation. With Best Effort
Early Offer, the SIP trunk sends an Early Offer only if it has enough information about the calling
device's media capabilities to create the Offer; if it does not have this information, it sends a Delayed
Offer instead.
Prior to Best Effort Early Offer, the decision to use Delay Offer or Early Offer on leaf cluster trunks
was typically based upon the number of older SCCP endpoints registered to the cluster. Because older
SCCP endpoints require the insertion of an MTP to create an Offer for calls over Early Offer SIP trunks,
where large numbers of SCCP endpoints exist within the cluster, Delayed Offer was preferred to avoid
MTP usage. Best Effort Early Offer removes the need to decide upon Early Offer or Delayed Offer SIP
trunk configuration based on the type of endpoints registered to the cluster.
6-37
Chapter 6
In Cisco Collaboration System deployments, receipt of Early Offer only may be required by non-Cisco
Unified Communications applications and services. There are two options to address the requirement
that Early Offer is always received:
Cisco Unified Border Element provides a SIP Delayed Offer to Early Offer feature for voice calls,
which converts inbound Delayed Offer calls to outbound Early Offer calls, thus allowing
Unified CM and SME to use Best Effort Early Offer trunks. A typical example of this use case is
service provider IP PSTN connections, which typically must always receive SIP Early Offer.
For enterprise Unified Communications applications that accept only SIP Early Offer, a dedicated
Early Offer SIP trunk can be used from the Unified CM Leaf cluster to the Unified Communications
application. If a large number of MTPs are required on the Early Offer SIP trunk, consider using the
Cisco Unified Border Element Delayed Offer to Early Offer conversion feature.
In SME clusters, Best Effort Early Offer performs the same role as MTP-Less Early Offer by making
the SME cluster transparent to media negotiation, which in turn forces media decisions to be made by
the end Unified Communications system where, if required, media resources can be inserted to address
DTMF or codec mismatch issues. Media resources must not be associated with MTP-Less Early Offer
SME trunks. If needed, media resources can be associated with Best Effort Early Offer trunks.
Run on All Unified CM Nodes
This feature is supported on SIP trunks and route lists, and it greatly simplifies call routing from and
through Unified CM and SME clusters. Cisco highly recommends enabling the Run on all Unified CM
nodes feature on all SIP trunks and route lists. Call routing is simplified through a combination of the
Run on all Unified CM nodes and the Route Local features, whereby phone calls over SIP trunks will
always originate from the Unified CM node where the phone is registered. Likewise for trunk to trunk
calls, the outbound SIP trunk call will always originate from the Unified CM node on which the inbound
Trunk call arrived. Enabling Run on all Unified CM nodes on all SIP trunks and route lists eliminates
the need to set up calls between call processing nodes within the cluster, which can be useful when
clustering over the WAN is deployed within a Unified CM or SME cluster.
Up to 16 SIP Trunk Destination IP Addresses
SIP trunks can be configured with up to 16 destination IP addresses, 16 fully qualified domain names,
or a single DNS SRV entry. Support for additional destination IP addresses reduces the need to create
multiple trunks associated with route lists and route groups for call distribution between two Unified
Communications systems, thus simplifying Unified CM trunk design. When IP addresses are used as
destinations on a SIP trunk, Unified CM randomly distributes calls across all defined destination IP
addresses.
SIP OPTIONS Ping
Enable the SIP OPTIONS Ping feature on the SIP Profile associated with a SIP trunk to dynamically
track the state of each the trunk's destinations and the overall state of the trunk.
PRACK
PRACK provides reliability of 1XX responses for interoperability scenarios with the PSTN, and it can
also be used to reduce the number of SIP messages that need to be exchanged before setting up two-way
media. Enable PRACK through the SIP Rel1XX Options parameter on the SIP Profile associated with
the trunk.
SIP Trunk DTMF Signaling Method No Preference
Using DTMF Signaling Method: No Preference is recommended on SIP trunks. In this mode
Unified CM attempts to minimize the usage of MTP resources by selecting the most appropriate DTMF
signaling method (in-band or out-of-band) for the call.
6-38
Chapter 6
Voice calls
Video calls
Encrypted calls
Fax calls
Unified CM Session Management Edition may also be used to connect to the PSTN and third-party
unified communications systems such as PBXs and centralized unified communications applications.
Figure 6-22
Multisite Distributed Call Processing Deployment with Unified CM Session Management Edition
348660
Cisco Unified CM
Session Management Edition
Centralized
PSTN
Centralized
Applications
SIP
Campus
Campus
Remote
Office
TDM
PBX
IP PBX
Unified
CME
6-39
Chapter 6
Note
Running ILS GDPR on SME and Unified CM leaf clusters further simplifies dial plan
administration because individual directory numbers, E.164 numbers corresponding to DNs,
route patterns (for internal and external number ranges), and URIs can be distributed using
the ILS service. This approach simplifies dial plan administration by reducing the required
number of route patterns to one SIP route pattern per call control system (Unified CM
cluster, for example), instead of a route pattern for each unique number range. For more
information on ILS and GDPR, see Intercluster Lookup Service (ILS) and Global Dial Plan
Replication (GDPR), page 10-29.
Centralize applications
The deployment of Unified CM Session Management Edition enables commonly used applications
such as conferencing or voice mail to connect directly to the SME cluster, thus reducing the
overhead of managing multiple trunks to leaf systems.
6-40
Chapter 6
It is important to size the Unified CM Session Management cluster correctly based on the expected
BHCA traffic load between leaf Unified Communications systems (for example, between Unified CM
clusters and PBXs), to and from any centralized PSTN connections, and to any centralized applications.
Determine the average BHCA and call holding time for users of your Unified Communications system,
and share this information with your Cisco account Systems Engineer (SE) or Cisco Partner to size your
Unified CM Session Management Edition cluster correctly. For more information on SME sizing, see
the chapter on Collaboration Solution Sizing Guidance, page 25-1.
SME Trunks
Although SME supports SIP, H.323, and MGCP trunks, Cisco highly recommends SIP as the trunk
protocol of choice for SME and Unified CM leaf clusters running Cisco Unified Communications
System Release 8.5 and later versions.
Using only SIP trunks in the SME cluster allows you deploy a "media transparent" cluster where media
resources (when required) are inserted by the end or leaf Unified Communications system and never by
SME. Using only SIP trunks also allows you use extended round-trip times (RTTs) between SME nodes
when clustering over the WAN.
SME SIP trunks should be configured as Best Effort Early Offer trunks. Leaf Unified CM cluster SIP
trunks should also be configured as Best Effort Early Offer.
SME Transparency for Media Negotiation
When a media resource such as an MTP or transcoder is needed to allow a call to proceed successfully,
these resources should be allocated by the edge or leaf Unified Communications systems. If SME trunk
media resources are used for a call traversing the SME cluster, the media path call will hairpin through
the SME media resource. By using SIP trunks only and Best Effort Early Offer (or MTP-less Early
Offer), an SME cluster can be deployed without media resources. If or when media resources are
required, they can be provided by the edge or leaf Unified Communications system.
Clustering over the WAN with SME CoW+
With Cisco Unified CM 9.1 and later releases, SME deployments support round-trip times (RTTs) of up
to 500 ms between SME cluster nodes. (See Figure 6-23.) This extended RTT applies only to SME
clusters (80 ms is the maximum RTT for standard Unified CM cluster designs) and is subject to the
following design restrictions:
SME deployments with extended clustering over the WAN (CoW+) round-trip times are supported
with SIP trunks only. All SIP trunks must be configured as either all Best Effort Early Offer
(preferred) or all MTP-less Early Offer and must use the Run on all Unified CM Nodes feature
so that calls are not routed between nodes within the SME cluster. H.323, MGCP, and SCCP
protocols are not supported for SME deployments with extended clustering over the WAN round-trip
times.
6-41
Chapter 6
Note
No media resources such as MTPs, trusted relay points (TRPs), RSVP agents, or transcoders are
configured or registered to the SME cluster. (To disable media resources hosted on Unified CM
nodes, deactivate the IPVMS service on each node within the cluster.)
A minimum of 1.544 Mbps (T1) bandwidth is required for Intra-Cluster Communication Signaling
(ICCS) traffic between sites.
In addition to the bandwidth required for Intra-Cluster Communication Signaling (ICCS) traffic, a
minimum of 1.544 Mbps (T1) bandwidth is required for database and other inter-server traffic
between the publisher node and every remote subscriber node.
As with all SME designs, your SME design must be reviewed and approved by the Cisco SME Team
prior to deployment.
The upgrade process for an SME cluster consists of two key parts:
Version switch-over The call processing node is rebooted and initialized with the new software
version (this takes approximately 45 minutes per server).
Database replication The subscriber's database is synchronized with that of the publisher node.
The time taken to complete this database replication phase depends on the number of subscribers nodes
in the cluster and the RTT between the publisher and subscriber nodes. The database replication process
has a minimal impact of the subscriber's call processing capability and typically can be run as a
background process during normal SME cluster operation. Avoid making changes to the SME cluster
configuration during the database replication phase because this increases the time it takes to complete
the replication.
For SME clusters deployed with extended RTTs, before upgrading the cluster, run the following
administrator-level CLI command on the publisher node:
utils dbreplication setprocess 40
This command improves replication setup performance and reduce database replication times.
6-42
Chapter 6
Figure 6-23
Unified CM Session Management Edition Clustering over the WAN with Extended Round Trip Times
North America
Europe
< 500 mS
RTT
Latin America
Publisher Node
Asia Pac
SIP Trunks
Best Effort Early Offer
348661
Cisco Unified CM
Session Management Edition
Cluster
Unified CM Versions
Using the latest Cisco Collaboration Systems Release and SIP trunks across all Unified CM leaf clusters
and the SME cluster enables your deployment to benefit from common cross-cluster features such as
codec preference lists, ILS, GDPR, and Enhanced Locations call admission control (CAC). If you do not
wish to upgrade to the latest Unified CM version on all clusters, the lowest recommended version is
Cisco Unified CM 8.5 using SIP trunks because this version includes features that improve and simplify
call routing through Unified CM and Session Management Edition clusters.
6-43
Chapter 6
Unityy
n
Connecon
Voicemaill
m
System
348817
Figure 6-24
Expressway-E and
Expressway-C
Session
Management
Edion Cluster
C
Cisco
EEmergency
R
Responder
CER
Operator
Console
SIP Line
Cisco Unied CM
Leaf Cluster 1
Cisco Unied CM
Leaf Cluster N
Enabling inbound and outbound Redirecting Number IE Delivery on MGCP gateways, H.323
gateways, and H.323 trunks
Enabling inbound and outbound Redirecting Diversion Header Delivery on SIP trunks
For QSIG-enabled SIP, MGCP, and H323 trunks, the original called party number is sent in QSIG
Diverting Leg Information APDUs. The diversion information sent in QSIG APDUs over QSIG-enabled
trunks does not pick up any calling party modifications and also does not honor the Voice Mail Box Mask
setting. QSIG diversion information sent by Unified CM is always set to the redirecting DN without
applying any transformations.
6-44
Chapter 6
If the redirecting DN is configured as +E.164, the leading "+" is removed and the QSIG diversion
information carries only the E.164 number without the "+" character.
Users with directory numbers in E.164 format should have a corresponding voicemail system
mailbox number using the same E.164 format.
Users with directory numbers in +E164 format should have a corresponding voicemail system
mailbox number using the same E164 format and an alternate voice mailbox number using the
+E.164 format.
6-45
Chapter 6
SIP Trunk
XML RPC over HTTP/HTTPS
348818
Figure 6-25
Permenent
Conferences
Session
Management
Edion Cluster
Cisco Unied CM
Leaf Cluster 1
Cisco Unied CM
Leaf Cluster N
6-46
Chapter 6
348819
Figure 6-26
Internet
Expressway-E and
Expressway-C
Session
Management
Edion Cluster
Cisco Unied CM
Leaf Cluster 1
Cisco Unied CM
Leaf Cluster N
6-47
Chapter 6
Configure one SIP trunk to each set of SME nodes in each regional data center. For example, if there
are four regional SME data centers, create four SIP trunks in each leaf cluster (see Figure 6-27). This
allows calls from all SME nodes to be received and accepted by the leaf clusters. Enable Run on all
Unified CM nodes on all of these trunks.
Place two or more of these leaf cluster SIP trunks into a route list and route groups for path
redundancy to the SME CoW+ cluster.
Best Effort Early Offer is recommended for all leaf cluster SIP trunks.
In Unified Communications deployments, receipt of Early Offer only might be required by
non-Cisco Unified Communications applications and services. For leaf clusters, there are two
options to address the requirement that Early Offer is always received:
Cisco Unified Border Element provides a SIP Delayed Offer to Early Offer feature for voice
calls, which converts inbound Delayed Offer calls to outbound Early Offer calls, thus allowing
Unified CM and SME to use Best Effort Early Offer trunks. A typical example of this use case
is for service provider IP PSTN connections via Cisco Unified Border Element, which typically
must always receive SIP Early Offer.
For enterprise Unified Communications applications that accept only SIP Early Offer, a
dedicated Early Offer SIP trunk can be used from the Unified CM Leaf cluster to the Unified
Communications application. If a large number of MTPs are required on the Early Offer SIP
trunk, consider using the Cisco Unified Border Element Delayed Offer to Early Offer
conversion feature instead.
Enable the IPVMS service on all leaf cluster nodes. Activate conferencing, music on hold, and
annunciator resources as required. (Deactivating IPVMS-based MTPs is recommended.)
As required, configure and associate Cisco IOS media resources (MTPs, conferencing, and
transcoding) with the leaf cluster.
6-48
Chapter 6
Figure 6-27
Data Center 1
SME Nodes
1st Choice
Trunk
Route
List
2nd Choice
Trunk
Leaf Cluster
Europe
CoW
Route
List
1st Choice
Trunk
Data Center 2
SME Nodes
Best Effort Early Offer
SIP
348662
Leaf Cluster
North America
CoW
2nd Choice
Trunk
Configure one SIP trunk from the SME cluster to each leaf cluster (see Figure 6-28). Enable Run
on all Unified CM nodes on these trunks, and configure trunk destinations to every call processing
node in the leaf clusters.
Best Effort Early Offer is recommended for all SME cluster SIP trunks.
If receipt of Early Offer only is required by non-Cisco Unified Communications applications and
services connected to SME clusters, there are two options to address the requirement:
Cisco Unified Border Element provides a SIP Delayed Offer to Early Offer feature for voice
calls, which converts inbound Delayed Offer calls to outbound Early Offer calls, thus allowing
Unified CM and SME to use Best Effort Early Offer trunks only. A typical example of this use
case is for service provider IP PSTN connections via Cisco Unified Border Element, which
typically must always receive SIP Early Offer.
For enterprise Unified Communications applications that accept only SIP Early Offer, if a
dedicated Early Offer SIP trunk is used from the SME cluster to the Unified Communications
application, media resources will have to be associated with the SME trunks, which if used will
cause unwanted media hair-pinning. The media resources typically used in this case are MTPs
to create an Early Offer or address DTMF mismatches and transcoders to address codec
mismatches. Using media resources in the SME cluster is not generally recommended; as an
6-49
Chapter 6
alternative, consider using the Cisco Unified Border Element Delayed Offer to Early Offer
feature between SME and the Unified Communications application, or use a direct trunk to the
application from the leaf cluster.
Figure 6-28
Disable the IPVMS service on all SME nodes. This disables Unified CM media termination points,
conferencing, music on hold, and annunciator resources.
Do not associate any Cisco IOS media resources with the SME cluster.
Enable Accept Audio Codec Preference in Received Offer on all SME SIP trunks.
Data Center 1
SME Nodes
Leaf Cluster
Europe
CoW
Data Center 2
SME Nodes
Best Effort Early Offer
SIP
348663
Leaf Cluster
North America
CoW
The outbound leaf cluster will originate a SIP trunk call from the same node that the calling device is
registered to (using the Route Local rule). The leaf cluster will randomly select a destination address
from the SIP trunk route list. (For the example in Figure 6-29, the first-choice trunk is selected.)
Outbound calls from the SME cluster will originate from the same node that the inbound call arrived on
(using the Route Local rule). With Run on all Unified CM nodes enabled on all SME trunks, calls will
never be set up between call processing nodes within the SME cluster. The SME cluster will randomly
select a destination address on the SIP trunk pointing toward the destination leaf cluster.
6-50
Chapter 6
For inbound SIP trunk calls to the destination leaf cluster, calls may be extended from the call processing
node on which the inbound call arrived, to the node where the called device is registered.
Media resources, if needed, will be inserted by the leaf clusters (or end Unified Communications
systems). If the device in the calling leaf cluster SIP trunk uses Delayed Offer, the media decision will
be made by this cluster, which will insert media resources (MTPs and/or transcoders) as required. If the
device in the calling leaf cluster SIP trunk send an Early Offer, the media decision will be made by the
destination leaf cluster, which will insert media resources (MTPs and/or transcoders) as required.
Figure 6-29
Recommended Trunk Configuration for Call Routing Through Leaf and SME Clusters
Data Center 1
SME Nodes
1st Choice
Trunk
Leaf Cluster
North America
CoW
Route
List
Leaf Cluster
Europe
CoW
Data Center 2
SME Nodes
Best Effort Early Offer
SIP
Signalling Path
348664
2nd Choice
Trunk
This feature is used to address interoperability issues with third-party products. When Unified CM
places a call hold over a SIP trunk, it sends a mid-call INVITE with audio direction media attribute
a=inactive in the SDP body to disconnect the media connection. On call resumption, Unified CM sends
a Delayed Offer INVITE (without SDP) to the held device to obtain its media characteristics through an
SDP Offer. According to RFC 3261 (section 14.2), the held device should construct the Offer as if it
6-51
Chapter 6
were making a new call; that is, with a list of all supported codecs and a=sendrecv. Some third-party
products respond with only the last used codec and media direction attributes, with the result that the
call always remains in the inactive state and media cannot be resumed. When Send "sendrecv" in mid
call INVITE is enabled, this feature inserts an MTP into the media path between the calling and calling
devices, allowing the media connection to be disconnected between the Unified CM device and the MTP,
while establishing and maintaining media between the MTP and the held device with a=sendrecv. On
call resumption, the MTP is removed from the media path. This feature addresses the mid-call Delayed
Offer INVITE issue for audio direction, but it cannot resolve the issue of a device responding with its
last used codec rather than the full list of all supported codecs. This issue can be problematic in cases
where a codec change is required on re-establishment of media, such as placing a G.729 call on hold and
connecting it to a music on hold source where G711 is preferred.
Require SDP Inactive in Mid-Call Media Exchange
SIP allows mid-call updates to codecs and connection information, such as IP addresses and UDP port
numbers, without disconnecting the media connection. Some third-party devices cannot accept media
changes using this method, and they require the media path to be closed gracefully and reopened to make
media changes. If this feature is enabled, during mid-call codec or connection updates Cisco Unified CM
sends an INVITE a=inactive SDP message to the endpoint to break the media exchange.
Note
For SIP trunks enabled for Early Offer, this parameter will be overridden by the Send send-receive SDP
in mid-call INVITE parameter.
Disable Early Media on 180
By default, Cisco Unified CM signals the registered calling phone to play local ringback if SDP is not
received in a 180 Ringing or 183 Session Progress Response.
If SDP is included in the 180 or 183 Response, instead of playing ringback locally, Cisco Unified CM
connects media, and the calling phone plays whatever the called device is sending in its media stream
(such as ringback or busy signal).
If ringback is not received, the device to which you are connecting might be including SDP in the 180
response, but it is not sending any media before the 200 OK response. In this case, check this check box
to play local ringback on the calling phone and connect the media upon receipt of the 200 OK response
Redirect by Application
Apply a specific calling search space to redirected contacts that are received in the 3xx response.
Apply digit analysis to the redirected contacts to make sure that the call gets routed correctly.
Prevent a DOS attack by limiting the number of redirection (recursive redirection) requests
If the Redirect by Application check box is unchecked, outbound SIP trunk calls can be redirected to a
restricted phone number (such as an international number) because redirection is handled and routed at
the SIP stack level without intervention of Unified CM digit analysis and class of service.
Re-Route Incoming Request to New Trunk Based on
Inbound SIP trunk calls to Unified CM will be accepted only if the source IP address and port number
of the calling device match the destination IP address and port number of a configured SIP trunk. Once
the call has been accepted, it can then optionally be re-routed to another Unified CM SIP trunk based on
information contained within the received SIP message header.
6-52
Chapter 6
By default, calls are never re-routed after being matched to a SIP trunk based on IP address and port
number.
Optionally, incoming Requests can be re-routed to a new trunk based on the received:
Contact header
The call is re-routed to another SIP trunk based on the IP address and port number received in the
contact header. This feature is typically used to re-route calls from a SIP Proxy to a Unified CM SIP
trunk assigned to a specific end user or system.
This feature can be useful if, for example, you wish to send a company switchboard number and
company name instead of the caller's number and name in the SIP messages of calls sent over the SIP
trunk. (See Figure 6-30.)This feature can be applied at the device level (for branch offices using a
centralized SIP trunk) or at the trunk level.
At the device level, use the Caller ID DN and Caller Name fields of the Incoming Requests FROM URI
Setting section on the SIP Profile associated with the device.
At the trunk level, use the Caller ID DN and Caller Name fields of the Outbound Calls - Caller
Information section of the trunk configuration page.
By default the Caller ID DN and Caller Name sent in the From header, Contact header, and
P-Asserted-Identity and Remote-Party-ID headers are modified in outbound SIP trunk calls. If you wish
to keep the original Caller ID in the P-Asserted-Identity and Remote-Party-ID headers, check the
Maintain Original Caller ID DN and Caller Name in Identity Headers check box on the trunk
configuration page. Checking this check box allows the originator of the call to be traced.
Overwriting Caller ID Number and Caller Name on Outbound SIP Trunk Calls
348665
Figure 6-30
6-53
Chapter 6
SIP headers
SIP parameters
Content bodies
SDP
Normalization applies to all calls that traverse a SIP trunk with an associated script, regardless of what
protocol is being used for the other endpoint involved in the call. For example, a SIP trunk normalization
script can operate on a call from a SIP line device to a SIP trunk, from an SCCP device to a SIP trunk,
from MGCP trunk to SIP trunk, from H.323 trunk to SIP trunk, and so forth. (See Figure 6-31.)
6-54
Chapter 6
Figure 6-31
SIP Trunk
Normalization Script
SCCP Line
SIP Trunk
IP PSTN
Normalization Script
SIP Trunk
SIP Trunk
Normalization Script
H.323 Trunk
SIP Trunk
IP PSTN
Normalization Script
MGCP Trunk
Normalization Script
348666
SIP Trunk
SIP headers
SIP parameters
Content bodies
SIP Profiles also support SDP Transparency Profiles, which can be used to pass either all unknown SDP
parameters (default) or selected SDP parameters that are not natively supported by Unified CM, from
one SIP trunk (or SIP endpoint) to another without using Lua transparency scripts.
6-55
Chapter 6
Figure 6-32
SIP Trunk
SIP Trunk
IP PSTN
Transparency Script
SIP Trunk
SIP Trunk
IP PSTN
Transparency Script
SIP Trunk
SIP Transparency
Profile
348667
SIP Trunk
Normalization and transparency scripts use Lua, a powerful, fast and lightweight, embeddable scripting
language to modify SIP messages and SDP body content on SIP trunks. (For more information on Lua,
refer to the documentation available at http://lua-users.org/wiki/LuaOrgGuide.)
For more information on SIP trunk normalization and transparency scripts, refer to the latest version of
the Developer Guide for SIP Transparency and Normalization, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/products_programming_reference_guide
s_list.html
The developer guide describes the scripting environment and APIs used to manipulate and pass through
SIP message information.
For more information on script management, refer to the latest version of the Cisco Unified
Communications Manager Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
Refer-passthrough script This script allows Unified CM to be removed from the call signaling
path when a blind transfer (using an in-dialog REFER) is invoked between two SIP trunks.
ContactHeader script This script removes the audio and video attributes from the contact header
in an inbound Delayed Offer mid-call re-invite.
Diversion-Counter script This script provides the capability to adjust the diversion counter for
various Call Forward scenarios.
VCS-interop script This script provides interoperability for endpoints registered to the Cisco
TelePresence Video Communication Server (VCS).
6-56
Chapter 6
Media interworking (DTMF translation, fax, transcoding, transrating, volume and gain control)
Call admission control (based on total calls, CPU, memory, call arrival spike detection, or maximum
calls per destination)
Security (including RTP to SRTP interworking, SIP malformed packet detection, non-dialog RTP
packet drops, SIP listening port configuration, digest authentication, simultaneous call limits, call
rate limits, toll fraud protection, and a number of signaling and media encryption options)
QoS and bandwidth management (QoS marking using ToS, DSCP, and bandwidth enforcement
using RSVP and codec filtering)
Domain-based routing
Enterprise Phone Proxy VPN-Less IP Phone registration to Unified CM through Cisco Unified
Border Element
The Cisco Unified Border Element is a licensed Cisco IOS application available on a wide range of Cisco
router and gateway platforms. Depending on your choice of hardware platform, the Cisco Unified Border
Element can provide session scalability from 4 to 16,000 concurrent voice calls with in-box or
box-to-box failover options.
6-57
Chapter 6
For more information on the Cisco Unified Border Element, refer to the documentation at
http://www.cisco.com/go/cube
Centralized
IP PSTN
Cisco Unified
Border Element
Cisco Unified CM
Session
Management
Edition
SIP
Campus
Campus
Remote
Office
TDM
PBX
IP PBX
Unified
CME
348668
Distributed trunks connect to the service provider through several geographically distributed logical
connections. (See Figure 6-34.) Each branch of an enterprise may have its own local trunk to the service
provider. With distributed trunks in each branch site, media no longer needs to traverse the enterprise
WAN, but flows to the service provider interface through a local SBC.
6-58
Chapter 6
Figure 6-34
Campus
IP PSTN
Remote Office
IP PSTN
TDM
PBX
IP PBX
Unified
CME
PSTN
IP PSTN
IP PSTN
348669
Each connectivity model has its own advantages and disadvantages. Centralized trunks are generally
easier to deploy in terms of both physical equipment and configuration complexity, but media and
signaling must traverse the enterprise to reach the PSTN, therefore requiring high availability in the
enterprise WAN. Distributed trunks have the advantage of local hand-off of media and better number
portability from local providers. As illustrated in Figure 6-35, a hybrid connectivity model that groups
some of the branches together for connectivity, or that provides trunks from each Unified CM cluster of
a multi-cluster deployment, captures the advantages of both forms of deployment models.
6-59
Chapter 6
Figure 6-35
Centralized
IP PSTN
Cisco Unified
Border Element
Cisco Unified CM
Session
Management
Edition
SIP
Unified
CME
Remote Office
TDM
PBX
IP PSTN
IP PBX
IP PSTN
348670
Campus
6-60
CH A P T E R
Media Resources
Revised: June 15, 2015
A media resource is a software-based or hardware-based entity that performs media processing functions
on the data streams to which it is connected. Media processing functions include mixing multiple
streams to create one output stream (conferencing), passing the stream from one connection to another
(media termination point), converting the data stream from one compression type to another
(transcoding), streaming music to callers on hold (music on hold), echo cancellation, signaling, voice
termination from a TDM circuit (coding/decoding), packetization of a stream, streaming audio
(annunciation), and so forth. The software-based resources are provided by the Cisco Unified
Communications Manager (Unified CM) IP Voice Media Streaming Service (IP VMS). Digital signal
processor (DSP) cards provide both software and hardware based resources.
This chapter explains the overall Cisco Unified CM media resources architecture and Cisco IP Voice
Media Streaming Application service, and it focuses on the following media resources:
Use this chapter to gain an understanding of the function and capabilities of each media resource type
available on Unified CM and to determine which resource are required for your deployment. For
information on conferencing resources, refer to the chapter on Cisco Rich Media Conferencing,
page 11-1.
For proper DSP sizing of Cisco Integrated Service Router (ISR) gateways, you can use the Cisco Unified
Communications Sizing Tool (Unified CST), available to Cisco employees and partners at
http://tools.cisco.com/cucst. If you are not a Cisco partner or employee, you can use the DSP Calculator
at http://www.cisco.com/go/dspcalculator.
7-1
Chapter 7
Media Resources
Described in:
Revision Date
7-2
Chapter 7
Media Resources
Media Resources Architecture
Figure 7-1
Unified CM
Subscriber
IP
Cisco Unified
Border Element
253636
Xcode
IP Phone
Call Signaling
Media
Unified CM communicates with media resources using Skinny Client Control Protocol (SCCP). This
messaging is independent of the protocol that might be in use between Unified CM and the
communicating entities. Figure 7-2 shows an example of the message flow, but it does not show all of
the SCCP or SIP messages exchanged between the entities.
Figure 7-2
Cisco Unified
Border Element
Unified CM
Subscriber
IP Phone
IP
Xcode
253637
7-3
Chapter 7
Media Resources
Conference bridge
Annunciator
When the IP Voice Media Streaming Application is activated, one of each of the above resources is
automatically configured. Conferencing, annunciator, and MTP services can be disabled if required. If
these resources are not needed, Cisco recommends that you disable them by modifying the appropriate
service parameter in the Unified CM configuration. The service parameters have default settings for the
maximum number of connections that each media device can handle. For details on how to modify the
service parameters, refer to the appropriate version of the Cisco Unified Communications Manager
Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
Give careful consideration to situations that require multiple resources and to the load they place on the
IP Voice Media Streaming Application. The media resources can reside on the same server as
Unified CM or on a dedicated server not running the Unified CM call processing service. If your
deployment requires more than the default number of any resource, Cisco recommends that you
configure that resource to run on its own dedicated server. Cisco strongly recommends that you do not
activate the Cisco IP Voice Streaming Media Application on a Cisco Unified CM node that has a high
call processing load because it can adversely affect the performance of Cisco Unified CM. If heavy use
of media resources is expected within a deployment, Cisco recommends deploying dedicated
Unified CM media resource nodes (non-publisher nodes that do not perform call processing within the
cluster) or relying on hardware-based media resources. Software-based media resources on Unified CM
nodes are intended for small deployments or deployments where need for media resources is limited.
Voice Termination
Voice termination applies to a call that has two call legs, one leg on a time-division multiplexing (TDM)
interface and the second leg on a Voice over IP (VoIP) connection. The TDM leg must be terminated by
hardware that performs encoding/decoding and packetization of the stream. This termination function is
performed by a digital signal processor (DSP) resource residing in the same hardware module, blade, or
platform.
All DSP hardware on Cisco TDM gateways is capable of terminating voice streams, and certain
hardware is also capable of performing other media resource functions such as conferencing or
transcoding (see Transcoding, page 7-6 and Transcoding, page 7-6). The DSP hardware has either fixed
DSP resources that cannot be upgraded or changed, or modular DSP resources that can be upgraded.
The number of supported calls per DSP depends on the computational complexity of the codec used for
a call and also on the complexity mode configured on the DSP. Cisco IOS enables you to configure a
complexity mode on the hardware module. Hardware platforms such as the PVDM2, PVDM3, and
PVDM4 DSPs support three complexity modes: medium, high and flex mode. Some of the other
hardware platforms support only medium and high complexity modes.
7-4
Chapter 7
Media Resources
Media Resources Architecture
Flex Mode
Flex mode, available on hardware platforms that use the C5510 chipset and on PVDM3 DSPs, eliminates
the requirement to specify the codec complexity at configuration time. A DSP in flex mode accepts a
call of any supported codec type, as long as it has available processing power.
For C5510-based DSPs, the overhead of each call is tracked dynamically via a calculation of processing
power in millions of instructions per second (MIPS). Cisco IOS performs a MIPS calculation for each
call received and subtracts MIPS credits from its budget whenever a new call is initiated. The number of
MIPS consumed by a call depends on the codec of the call. The DSP will allow a new call as long as it
has remaining MIPS credits greater than or equal to the MIPS required for the incoming call.
Similarly, PVDM3 DSP modules use a credit-based system. Each module is assigned a fixed number of
"credits" that represent a measure of its capacity to process media streams. Each media operation, such
as voice termination, transcoding, and so forth, is assigned a cost in terms of credits. As DSP resources
are allocated for a media processing function, its cost value is subtracted from the available credits. A
DSP module runs out of capacity when the available credits run out and are no longer sufficient for the
requested operation. The credit allocation rules for PVDM3 DSPs are rather complex.
Flex mode has an advantage when calls of multiple codecs must be supported on the same hardware
because flex mode can support more calls than when the DSPs are configured as medium or high
complexity. However, flex mode does allow oversubscription of the resources, which introduces the risk
of call failure if all resources are used. With flex mode it is possible to have fewer DSP resources than
with physical TDM interfaces.
Compared to medium or high complexity mode, flex mode has the advantage of supporting the most
G.711 calls per DSP. For example, a PVDM2-16 DSP can support 8 G.711 calls in medium complexity
mode or 16 G.711 calls in flex mode.
7-5
Chapter 7
Media Resources
Transcoding
Transcoding
A transcoder is a device that converts an input stream from one codec into an output stream that uses a
different codec. Starting with Cisco IOS Release 15.0.1M, a transcoder also supports transrating,
whereby it connects two streams that utilize the same codec but with a different packet size.
Transcoding from G.711 to any other codec is referred to as traditional transcoding. Transcoding
between any two non-G.711 codecs is called universal transcoding and requires Universal Cisco IOS
transcoders. Universal transcoding is supported starting with Cisco IOS Release 12.4.20T. Universal
transcoding has a lower DSP density than traditional transcoding.
In a Unified CM system, the typical use of a transcoder is to convert between a G.711 voice stream and
the low bit-rate compressed voice stream G.729a. The following cases determine when transcoder
resources are needed:
Multiple codecs in use in the system, with all endpoints capable of all codec types
The most common reason for multiple codecs is to use G.711 for LAN calls to maximize the call
quality and to use a low-bandwidth codec to maximize bandwidth efficiency for calls that traverse
a WAN with limited bandwidth. Cisco recommends using G.729a as the low-bandwidth codec
because it is supported on all Cisco Unified IP Phone models as well as most other Cisco Unified
Communications devices, therefore it can eliminate the need for transcoding. Although Unified CM
allows configuration of other low-bandwidth codecs between regions, some phone models do not
support those codecs and therefore would require transcoders. They would require one transcoder
for a call to a gateway and two transcoders if the call is to another IP phone. The use of transcoders
is avoided if all devices support and are configured for both G.711 and G.729 because the devices
will use the appropriate codec on a call-by-call basis.
Multiple codecs in use in the system, and some endpoints support or are configured for G.711 only
This condition exists when G.729a is used in the system but there are devices that do not support
this codec, or a device with G.729a support may be configured to not use it. In this case, a transcoder
is also required. Devices from some third-party vendors may not support G.729.
A transcoder is also capable of performing the same functionality as a media termination point (MTP).
In cases where transcoder functionality and MTP functionality are both needed, a transcoder is allocated
by the system. If MTP functionality is required, Unified CM will allocate either a transcoder or an MTP
from the resource pool, and the choice of resource will be determined by the media resource groups, as
described in the section on Media Resource Groups and Lists, page 7-33.
To finalize the design, it is necessary to know how many transcoders are needed and where they will be
placed. For a multi-site deployment, Cisco recommends placing a transcoder local at each site where it
might be required. If multiple codecs are needed, it is necessary to know how many endpoints do not
support all codecs, where those endpoints are located, what other groups will be accessing those
resources, how many maximum simultaneous calls these device must support, and where those resources
are located in the network.
7-6
Chapter 7
Media Resources
Media Termination Point (MTP)
Video Interoperability
Video interoperability is the audio and video support for point-to-point calls between Cisco
TelePresence System (CTS) endpoints, other Cisco Unified Communications video endpoints, and
third-party video endpoints. Prior to Cisco Unified CM 8.5, video interoperability between the various
families of video endpoints was possible only with the insertion of a video component between
endpoints, such as a video transcoder or a multipoint control unit (MCU).
Cisco Unified CM 8.5 and later releases not only offer native video interoperability between the various
video endpoint family types, point-to-point, but also provide better video interoperability in general with
H.264 codec negotiation in SIP and H.323 protocols and enable the endpoints to negotiate high definition
(HD) resolutions when available. Video interoperability, however, is dependent on the endpoints to
support the interoperation. For further information, refer to Interoperability Between CTS Endpoints and
Other Cisco Endpoints or Devices, available at
http://www.cisco.com/en/US/docs/telepresence/interop/endpoint_interop.html
Re-Packetization of a Stream
An MTP can be used to transcode G.711 a-law audio packets to G.711 mu-law packets and vice versa,
or it can be used to bridge two connections that utilize different packetization periods (different sample
sizes).
7-7
Chapter 7
Media Resources
DTMF Conversion
DTMF tones are used during a call to signal to a far-end device for purposes of navigating a menu
system, entering data, or other manipulation. They are processed differently than DTMF tones sent
during a call setup as part of the call control. There are several methods for sending DTMF over IP, and
two communicating endpoints might not support a common procedure. In these cases, Unified CM may
dynamically insert an MTP in the media path to convert DTMF signals from one endpoint to the other.
Unfortunately, this method does not scale because one MTP resource is required for each such call. The
following sections help determine the optimum amount of MTP resources required, based on the
combination of endpoints, trunks, and gateways in the system.
If Unified CM determines that an MTP needs to be inserted but no MTP resources are available, it uses
the setting of the service parameter Fail call if MTP allocation fails to decide whether or not to allow
the call to proceed.This service parameter is set to a default value of False. With this default
configuration, an incoming call on a SIP Early Offer trunk would result in an outbound Delayed Offer.
Named Telephony Events (NTEs) defined by RFC 2833 are a method of sending DTMF from one
endpoint to another after the call media has been established. The tones are sent as packet data using the
already established RTP stream and are distinguished from the audio by the RTP payload type field. For
example, the audio of a call can be sent on a session with an RTP payload type that identifies it as G.711
data, and the DTMF packets are sent with an RTP payload type that identifies them as NTEs. The
consumer of the stream utilizes the G.711 packets and the NTE packets separately.
Key Press Markup Language (RFC 4730)
The Key Press Markup Language (KPML) is defined in RFC 4730. Unlike NTEs, which is an in-band
method of sending DTMF, KPML uses the signaling channel (out-of-band, or OOB) to send SIP
messages containing the DTMF digits.
KPML procedures use a SIP SUBSCRIBE message to register for DTMF digits. The digits themselves
are delivered in NOTIFY messages containing an XML encoded body.
Unsolicited Notify (UN)
Unsolicited Notify procedures are used primarily by Cisco IOS SIP Gateways to transport DTMF digits
using SIP NOTIFY messages. Unlike KPML, these NOTIFY messages are unsolicited, and there is no
prior registration to receive these messages using a SIP SUBSCRIBE message. But like KPML,
Unsolicited Notify messages are out-of-band.
Also unlike KPML, which has an XML encoded body, the message body in these NOTIFY messages is
a 10-character encoded digit, volume, and duration, describing the DTMF event.
H.245 Signal, H.245 Alphanumeric
H.245 is the media control protocol used in H.323 networks. In addition to its use in negotiating media
characteristics, H.245 also provides a channel for DTMF transport. H.245 utilizes the signaling channel
and, hence, provides an out-of-band (OOB) way to send DTMF digits. The Signal method carries more
information about the DTMF event (such as its actual duration) than does Alphanumeric.
7-8
Chapter 7
Media Resources
Media Termination Point (MTP)
This method sends DTMF digits in-band, that is, in the same stream as RTP packets. However, the DTMF
packets are encoded differently than the media packets and use a different payload type. This method is
not supported by Unified CM but is supported on Cisco IOS Gateways.
Skinny Client Control Protocol (SCCP)
The Skinny Client Control Protocol is used by Unified CM for controlling the various SCCP-based
devices registered to it. SCCP defines out-of-band messages that transport DTMF digits between
Unified CM and the controlled device.
DTMF Relay Between Endpoints in the Same Unified CM Cluster
The following rules apply to endpoints registered to Unified CM servers in the same cluster:
Enabling this option on the SIP trunk assigns an MTP for every outbound call. This option does not
support codec pass-through mode, which imposes a single codec (G.711 or G.729) limitation over the
SIP trunk, thus limiting media to voice calls only. With this option enabled, calls over the trunk uses
MTPs assigned to the trunk rather than using calling device MTPs, which forces the media to follow the
same signaling path.
7-9
Chapter 7
Media Resources
Note
Enabling the Media Termination Point Required option on the SIP trunk increases MTP usage because
an MTP is assigned for every inbound and outbound call rather than on an as-needed basis.
Early Offer support for voice and video calls Mandatory (insert MTP if needed)
Enabling this Unified CM configuration option on the SIP Profile associated with the SIP trunk inserts
an MTP only if the calling device cannot provide Unified CM with the media characteristics required to
create the outbound Early Offer (for example, where an inbound call to Unified CM is received on a
Delayed Offer SIP trunk or a Slow Start H.323 trunk and on calls from older SCCP-based phones such
as Cisco Unified IP Phones 7940 or 7960 registered to Unified CM). Unified CM creates a super-set of
the endpoint and MTP codec capabilities and applies the codec filtering based on the applicable
region-pair settings. The outbound Offer SDP will use the IP address and port number of the MTP and
voice codec supported by the calling phone.
When Unified CM receives an inbound call on an H.323 Slow Start or SIP Delayed Offer trunk, the
media capabilities of the calling device are not available when the call is initiated. In this case,
Unified CM must insert an MTP and use its IP address and UDP port number to advertise all supported
audio codecs (after region-pair filtering) in the Offer SDP of the initial INVITE sent over the outbound
SIP trunk. When the Answer SDP is received on the SIP trunk, if it contains a codec that the calling
endpoint supports, no additional offer-answer transaction is needed. In case of codec mismatch,
Unified CM can either insert a transcoder to address the mismatch or send a Re-INVITE or UPDATE to
trigger media negotiation. Calls from H.323 Slow Start or SIP Delayed Offer trunks support audio only
in the initial call setup, but they can be upgraded mid-call to support video and SRTP if the call media
is renegotiated (for example, after Hold/Resume).
When you configure Early Offer support for voice and video calls Mandatory (insert MTP if
needed) on the SIP Profile of a trunk, calls from older SCCP-based phones, SIP Delayed Offer trunks,
and H.323 Slow Start trunks cause Unified CM to allocate an MTP, if an MTP or transcoder is not
already allocated for that call for another reason. The MTP is used to generate an Offer SDP with a valid
media port number and IP address. The MTP is allocated from the media resources that are associated
with the calling device rather than from the media resources of the outbound SIP trunk. (This prevents
the media path from being anchored to the MTP of the outbound SIP trunk.) If the MTP cannot be
allocated from the media resource group list (MRGL) of the calling device, the MTP allocation is
attempted from the MRGL of the SIP trunk.
Note
If no MTP resources are available, the call will proceed as a Delayed Offer call.
Unified CM does not need to insert an MTP to create an outbound Early Offer call over a SIP trunk if
Unified CM receives the inbound call by any of the following means:
On an MGCP trunk
Early Offer support for voice and video calls Best Effort (No MTP inserted)
If this Unified CM SIP profile configuration option is enabled, the SIP trunk will never use MTPs to
create an Early Offer but will send either an Early Offer or a Delayed Offer, depending on the capabilities
of the calling device.
7-10
Chapter 7
Media Resources
Media Termination Point (MTP)
Best Effort Early Offer SIP trunks send outbound calls as Early Offer (INVITE with SDP content) in the
following situations:
An inbound call to Unified CM or SME is received over a SIP trunk using Early Offer.
An inbound call to Unified CM or SME is received over an H.323 trunk using Fast Start.
A call is initiated from a newer model SCCP-based Cisco Unified IP Phone registered to
Unified CM.
Best Effort Early Offer trunks send outbound calls as Delayed Offer (INVITE without SDP content) in
the following situations:
An inbound call to Unified CM or SME is received over a Delayed Offer SIP trunk.
An inbound call to Unified CM or SME is received over an H.323 Slow Start trunk.
A call is initiated from an older model SCCP-based IP phone registered to Unified CM.
Calls over a Best Effort Early Offer SIP trunk support voice, video, and encrypted media.
In general, Cisco recommends Early Offer support for voice and video calls Best Effort (No MTP
inserted) for all Unified CM and Unified CM Session Management Edition SIP trunks.
For more information on this option, refer to the section on Best Effort Early Offer [Early Offer support
for voice and video calls Best Effort (no MTP inserted)], page 6-23.
Is the far-end SIP device defined by this SIP trunk capable of accepting an inbound call without a
SIP Early Offer?
If not, then on the SIP Profile associated with this trunk, select Early Offer support for voice and
video calls (insert MTP if needed). For outbound SIP trunk calls, an MTP will be inserted only if
the calling device cannot provide Unified CM with the media characteristics required to create the
Early Offer, or if DTMF conversion is needed.
If yes, then select Early Offer support for voice and video calls Best Effort (No MTP inserted),
and use Step 2. to determine whether an MTP is inserted dynamically for DTMF conversion. Note
that DTMF conversion can be performed by the MTP regardless of the codec in use.
2.
Select a Trunk DTMF Signaling Method, which controls the behavior of DTMF selection on that
trunk. Available MTPs will be allocated based on the requirements for matching DTMF methods for
all calls.
a. DTMF Signaling Method: No Preference
In this mode, Unified CM attempts to minimize the usage of MTP by selecting the DTMF
signaling method supported by the endpoint.
If both devices support RFC 2833, then no MTP is required.
If both devices support any out-of-band DTMF mechanism, then Unified CM will use KPML
over the SIP trunk. The only case where MTP is required is when one of the endpoints supports
out-of-band only and the other supports RFC 2833 only.
7-11
Chapter 7
Media Resources
If both devices support RFC 2833 and any out-of-band DTMF mechanism, then Unified CM
negotiates both RFC 2833 and KPML but relies on RFC 2833 to receive the digits.
b. DTMF Signaling Method: RFC 2833
By placing a restriction on the DTMF signaling method across the trunk, Unified CM is forced
to allocate an MTP if any one or both the endpoints do not support RFC 2833. In this
configuration, the only time an MTP will not be allocated is when both endpoints support
RFC 2833.
c. DTMF Signaling Method: OOB and RFC 2833
In this mode, the SIP trunk signals both KPML and RFC 2833 DTMF methods across the trunk,
and it is the most intensive MTP usage mode. The only cases where MTP resources will not be
required is when both endpoints support RFC 2833 and any OOB DTMF method (KPML or
SCCP).
Note
Cisco Unified IP Phones play DTMF to the end user when DTMF is received via SCCP, but they do not
play tones received by RFC 2833. However, there is no requirement to send DTMF to another end user.
It is necessary only to consider the endpoints that originate calls combined with endpoints that might
need DTMF, such as PSTN gateways, application servers, and so forth.
7-12
Chapter 7
Media Resources
Media Termination Point (MTP)
DTMF Conversion
An H.323 trunk supports the signaling of DTMF by means of H.245 out-of-band methods. H.323
intercluster trunks also support DTMF by means of NTE. There are no DTMF configuration options for
H.323 trunks; Unified CM dynamically chooses the DTMF transport method.
The following scenarios can occur when two endpoints on different clusters are connected with an H.323
trunk:
When both endpoints are SIP, then NTE is used. No MTP is required for DTMF.
When one endpoint is SIP and supports both KPML and NTE, but the other endpoint is not SIP, then
DTMF is sent as KPML from the SIP endpoint to Unified CM, and H.245 is used on the trunk. No
MTP is required for DTMF.
If one endpoint is SIP and supports only NTE but the other is not SIP, then H.245 is used on the
trunk. An available MTP is allocated for the call. The MTP will be allocated on the Unified CM
cluster where the SIP endpoint is located.
For example: A Cisco Unified IP Phone 7970 using SIP to communicate with a Cisco Unified IP
Phone 7970 running SCCP, will use NTE when connected via a SIP trunk but will use OOB methods
when communicating over an H.323 trunk (with the trunk using the H.245 method).
When a call is inbound from one H.323 trunk and is routed to another H.323 trunk, NTE will be used for
DTMF when both endpoints are SIP. H.245 will be used if either endpoint is not SIP. An MTP will be
allocated if one side is a SIP endpoint that supports only NTE and the other side is non-SIP.
7-13
Chapter 7
Media Resources
Assume the example system has CTI route points with first-party control (the CTI port terminates
the media), which integrate to a system that uses DTMF to navigate an IVR menu. If all phones in
the system are running SCCP, then no MTP is required. In this case Unified CM controls the CTI
port and receives DTMF from the IP phones via SCCP. Unified CM provides DTMF conversion.
However, if there are phones running a SIP stack (that support only NTE and not KPML), an MTP
is required. NTEs are part of the media stream; therefore Unified CM does not receive them. An
MTP is invoked into the media stream and has one call leg that uses SCCP, and the second call leg
uses NTEs. The MTP is under SCCP control by Unified CM and performs the NTE-to-SCCP
conversion. Note that the newer phones that do support KPML will not need an MTP.
7-14
Chapter 7
Media Resources
Media Termination Point (MTP)
MTP Resources
The following types of devices are available for use as an MTP:
Software MTP (Cisco IP Voice Media Streaming Application)
A software MTP is a device that is implemented by enabling the Cisco IP Voice Media Streaming
Application on a Unified CM server. When the installed application is configured as an MTP application,
it registers with a Unified CM node and informs Unified CM of how many MTP resources it supports.
A software MTP device supports only G.711 streams or passthrough mode in a codec. The IP Voice
Media Streaming Application is a resource that may also be used for several functions, and the design
guidance must consider all functions together (see Cisco IP Voice Media Streaming Application,
page 7-4).
Software MTP (Based on Cisco IOS)
The capability to provide a software-based MTP on the router is available beginning with Cisco IOS
Release 12.3(11)T for the Cisco 3800 Series Routers; Release 15.0(1)M for the Cisco 2900 Series
and 3900 Series Routers; Release IOS-XE for ASR1002, 1004, and 1006 Routers; Release
IOS-XE 3.2 for ASR1001 Routers; and Release 12.3(8)T4 for other router models.
This MTP allows configuration of any of the following codecs, but only one may be configured at a
given time: G.711 mu-law and a-law, G.729a, G.729, G.729ab, G.729b, and passthrough. Some of
these are not pertinent to a Unified CM implementation.
Router configurations permit up to 1,000 individual streams, which support 500 transcoded
sessions. This number of G.711 streams generates 10 Mbytes of traffic. The Cisco ISR G2s and ASR
routers can support significantly higher numbers than this.
Hardware MTP
DSP resources located in the voice modules or in the on-board Cisco Packet Voice/Fax Digital Signal
Processor (PVDM2, PVDM3, or PVDM4) slots for Cisco Integrated Services Routers (ISRs) can also
be used as MTP resource.
For more information on supported sessions with each PVDM module, refer the section on Capacity
Planning for Media Resources, page 7-30.
Note
When Cisco IOS MTP resources are invoked by Unified CM for a call flow, a software session rather
than a hardware DSP session is consumed unless the media legs of the call flow require transrating. Thus,
for flows invoking an MTP, a DSP session is used only when transrating (conversion between media legs
with the same codec but different packetization times) is required.
7-15
Chapter 7
Media Resources
Annunciator
An annunciator is a software function of the Cisco IP Voice Media Streaming Application that provides
the ability to stream spoken messages or various call progress tones from the system to a user. It uses
SCCP messages to establish RTP streams, and it can send multiple one-way RTP streams to devices such
as Cisco IP phones or gateways. For most SIP devices, the call progress tones are downloaded (pushed)
to the device at registration so that they can be invoked as needed by SIP signaling messages from
Unified CM. Some SIP devices such as intercluster SIP trunks may still use an annunciator for
call-progress tones. An annunciator may be used for verbal messages for almost any device regardless
of whether it is using SIP or SCCP.
In some installations, it might be a requirement to establish a two-way media connection with an
annunciator. To enable this capability, set the Cisco Unified CM service parameter Duplex Streaming
Enabled to True. This may be required for firewall transversal or possibly for SIP early-offer scenarios.
Tones and announcements are predefined by the system. The announcements support localization and
may also be customized by replacing the appropriate .wav file. The annunciator is capable of supporting
G.711 a-law and mu-law, G.729, and Cisco L16 Wideband codecs without any transcoding resources.
The following features require an annunciator resource:
7-16
Chapter 7
Media Resources
Annunciator
System messages
During the following call failure conditions, the system plays a streaming message to the end user:
A dialed number that the system cannot recognize
A call that is not routed due to a service disruption
A number that is busy and not configured for preemption or call waiting
Conferencing
During a conference call, the system plays a barge-in tone to announce that a participant has joined
or left the bridge.
An annunciator is automatically created in the system when the Cisco IP Voice Media Streaming
Application is activated on a server. If the Media Streaming Application is deactivated, then the
annunciator is also deactivated. A single annunciator instance can service the entire Unified CM cluster
if it meets the performance requirements (see Annunciator Performance, page 7-17); otherwise, you
must configure additional annunciators for the cluster. Additional annunciators can be added by
activating the Cisco IP Voice Media Streaming Application on other servers within the cluster.
The annunciator registers with a single Unified CM at a time, as defined by its device pool and
CM Group. It will automatically fail over to a secondary Unified CM if a secondary is configured for the
device pool. Any announcement that is playing at the time of an outage will not be maintained.
An annunciator is considered a media device, and it can be included in media resource groups (MRGs)
to control which annunciator is selected for use by phones and gateways.
Annunciator Performance
By default, the annunciator is configured to support 48 simultaneous streams, which is the maximum
recommended for an annunciator running on the same server (co-resident) with the Unified CM service.
If the server has only 10 Mbps connectivity, lower the setting to 24 simultaneous streams.
For more information on supported annunciator sessions with each server platform, refer to the section
on Media Resources, page 25-28, in the chapter on Collaboration Solution Sizing Guidance, page 25-1.
7-17
Chapter 7
Media Resources
Music on Hold
The Music on Hold (MoH) feature requires that each MoH server must be part of a Unified CM cluster
and participate in the data replication schema. Specifically, the MoH server must share the following
information with the Unified CM cluster through the database replication process:
Audio sources - The number and identity of all configured MoH audio sources
Multicast or unicast - The transport nature (multicast or unicast) configured for each of these sources
Multicast address - The multicast base IP address of those sources configured to stream as multicast
To configure a MoH server, enable the Cisco IP Voice Media Streaming Application Service on one or
more Unified CM nodes. An MoH server can be deployed along with Unified CM on the same server or
in standalone mode.
7-18
Chapter 7
Media Resources
Music on Hold
the endpoint device to join the multicast group address of the configured multicast MoH audio stream.
A multicast MoH server continuously streams each of the configured multicast MoH audio sources,
regardless of whether any callers are on hold.
Multicast MoH is available only for IPv4. Multicast for IPv6 is not currently supported by the MoH
server.
For a detailed look at MoH call flows, see the section on MoH Call Flows, page 7-23.
User Hold Audio Source and Media Resource Group List (MRGL)
MRGL B
MRG B
MoH B
Audio-source 1
Audio-source 2
Audio-source 3
Audio-source 4
MRGL A
MRG A
MoH A
Audio-source 1
Audio-source 2
Audio-source 3
Audio-source 4
Hold
IP
IP
Phone B
97984
Phone A
7-19
Chapter 7
Media Resources
Music on Hold
Because the configured MRGL determines the server from which a unicast-only device will receive the
MoH stream, you must configure unicast-only devices with an MRGL that points to a unicast MoH
resource or media resource group (MRG). Likewise, a device capable of multicast should be configured
with an MRGL that points to a multicast MRG containing a MoH server configured for multicast.
Figure 7-4 shows these two types of call flows. If phone A is in a call with phone B and phone A (holder)
pushes the Hold softkey, then a music stream is sent from the MoH server to phone B (holdee). The
music stream can be sent to holdees within the IP network or holdees on the PSTN, as is the case if
phone A places phone C on hold. In the case of phone C, the MoH stream is sent to the voice gateway
interface and converted to the appropriate format for the PSTN phone. When phone A presses the
Resume softkey, the holdee (phone B or C) disconnects from the music stream and reconnects to
phone A.
Figure 7-4
PSTN
Phone C
Central Site
M
M
M
IP
IP
Phone A
Phone B
97763
Hold
Call transfer
Call Park
Conference setup
Application-based hold
7-20
Chapter 7
Media Resources
Music on Hold
Figure 7-5 illustrates an example of network hold during a call transfer. The call flow involves the
following steps:
1.
2.
Phone A answers the call and then transfers it to phone B. During the transfer process, phone C is
put on network hold.
3.
Phone C receives an MoH stream from the MoH server via the gateway. After phone A completes
the transfer action, phone C disconnects from the music stream and gets redirected to phone B.
This process is the same for other network hold operations such as call park and conference setup.
Figure 7-5
PSTN
Phone C
Central Site
1
Unified CM Cluster with
dedicated MoH Server
M
M
M
IP
Phone A
IP
Phone B
97764
Transfer
7-21
Chapter 7
Media Resources
Music on Hold
MoH Sources
A Unified CM MoH server can generate a MoH stream from two types of sources:
An audio file from a Unified CM MoH server or from Cisco IOS router flash
A live fixed audio source from a Cisco IOS router or third-party device supporting multicast
You can configure a maximum of 501 MoH audio sources per Unified CM cluster, of which one (the
51st) can be a fixed live source or audio file multicast streamed from a Cisco IOS router. An MoH server
registers with the Unified CM cluster to provide IPv4 or dual-mode IPv4/IPv6 media address support.
Audio File
Audio files (.wav format) can be uploaded to Unified CM, which then automatically generates MoH
audio source files for the MoH codecs. Unified CM supports G711 (a-law and mu-law), G.729 Annex A,
and Cisco L16 Wideband codecs for MoH streams.
Note
Before configuring a MoH audio source, you must upload the .wav formatted audio source file to every
MoH server within the cluster using the upload file function in the Unified CM Administration interface.
Cisco recommends that you first upload the audio source file onto each MoH server in the cluster, then
upload it onto the publisher (even if not an MoH server), and finally assign an MoH Audio Stream
Number and configure the MoH audio source in the Unified CM Administration interface on the
publisher. This ensures that each MoH server has the MoH audio file available when it is assigned to an
MoH Audio Stream Number.
Fixed Source
If recorded or live audio is needed, multicast MoH can be generated from a fixed live source connected
to the analog interface of a Cisco IOS router or third-party device that supports multicast.
Note
Cisco Unified CM no longer supports a USB sound card for fixed live audio source connection to an
MoH server due to lack of USB port support for MoH when Unified CM nodes are virtualized.
This mechanism enables you to use radios, CD players, or any other compatible sound source to stream
multicast MoH. The stream from the fixed audio source is transcoded in real-time by the Cisco IOS
router.
Note
Prior to using a fixed audio source to transmit music on hold, you should consider the legalities and the
ramifications of re-broadcasting copyrighted audio materials. Consult your legal department for
potential issues.
For more information on live MoH from a Cisco IOS router, refer the section on MoH from a Live Feed
in the latest version of the Cisco Unified SCCP and SIP SRST System Administrator Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps2169/products_installation_and_configuratio
n_guides_list.html
7-22
Chapter 7
Media Resources
Music on Hold
MoH Selection
To determine which User and Network Audio Source configuration setting to apply in a particular case,
Unified CM interprets these settings for the holder device in the following priority order:
1.
Directory or line setting (Devices with no line definition, such as gateways, do not have this level.)
2.
Device setting
3.
4.
Unified CM also interprets the MRGL configuration settings of the holdee device in the following
priority order:
1.
Device setting
2.
3.
Note that system default MoH resources are resources that are not assigned to any MRG and they are
always unicast.
7-23
Chapter 7
Media Resources
Music on Hold
Figure 7-6
MoH Server
Phone B
Multicast Address:
239.192.240.1
Subscriber
Phone A
IP
IP
Multicast RTP
Stream
SCCP: CloseReceiveChannel
SCCP: StopMediaTransmission
SCCP: CloseReceiveChannel
SCCP: StartMulticastMediaReception
239.192.240.1
SCCP: StopMediaTransmission
IGMP: V2
MembershipReport
Phone B on HOLD
SCCP: StoptMulticastMediaReception
IGMP: V2
Leave Group
(224.0.0.2)
SCCP: OpenReceiveChannel
SCCP: StartMediaTramsmission
Phone A IP Address
RTP Stream: Active call
Multicast RTP
Stream
SCCP: OpenReceiveChannel
SCCP: StartMediaTramsmission
Phone B IP Address
141871
Multicast RTP
Stream
Meanwhile, the MoH server has been sourcing RTP audio to this multicast group address and, upon
joining the multicast group, phone B begins receiving the MoH stream. Once phone A presses the
Resume softkey, Unified CM instructs phone B to Stop Multicast Media Reception. Phone B then sends
an IGMP V2 Leave Group message to 224.0.0.2 to indicate that the multicast stream is no longer needed.
This effectively ends the MoH session. Next, Unified CM sends a series of Open Receive Channel
messages to phones A and B, just as would be sent at the beginning of a phone call between the two
phones. Soon afterwards, Unified CM instructs both phones to Start Media Transmission to each others
IP addresses. The phones are once again connected by means of an RTP two-way audio stream.
Note
The call flow diagrams in Figure 7-6 and Figure 7-7 assume that an initial call exists between phones
A and B, with a two-way RTP audio stream. These diagrams are representative of call flows and
therefore include only the pertinent traffic required for proper MoH operation. Thus, keep-alives,
acknowledgments, and other miscellaneous traffic have been eliminated to better illustrate the
interaction. The initial event in each diagram is the Hold softkey action performed by phone A.
7-24
Chapter 7
Media Resources
Music on Hold
MoH Server
Phone B
Subscriber
Phone A
IP
IP
SCCP: OpenReceiveChannel
SCCP: StartMediaTramsmission
Phone B IP Address
RTP Stream:Active one-way
Phone B on HOLD
SCCP: SoftKeyEventMessage RESUME
SCCP: StartMediaTramsmission
Phone A IP Address
SCCP: OpenReceiveChannel
SCCP: StartMediaTramsmission
Phone B IP Address
141872
Next, Unified CM tells phone B (the holdee) to Open Receive Channel. (This is quite different from the
multicast case, where Unified CM tells the holdee to Start Multicast Media Reception.) Then
Unified CM tells the MoH server to Start Media Transmission to the IP address of phone B. (This too is
quite different behavior from the multicast MoH call flow, where the phone is prompted to join a
multicast group address.) At this point, the MoH server is sending a one-way unicast RTP music stream
to phone B. When phone A presses the Resume softkey, Unified CM instructs the MoH server to Stop
Media Transmission and instructs phone B to Close Receive Channel, effectively ending the MoH
session. As with the multicast scenario, Unified CM sends a series of Open Receive Channel messages
and Start Media Transmissions messages to phones A and B with each others IP addresses. The phones
are once again connected by means of an RTP two-way audio stream.
7-25
Chapter 7
Media Resources
Music on Hold
7-26
Chapter 7
Media Resources
Music on Hold
Figure 7-8
MoH Server
M
Phone B
Multicast Address:
239.23.1.1
Subscriber
Phone A
IP
IP
IGMP: V2
Membership
Report
Multicast RTP
Stream
Phone B on HOLD
(RESUME pressed)
SIP: INVITE (SDP: 0.0.0.0,
RTP port#, a=inactive)
Multicast RTP
Stream
IGMP: V2
Leave Group
(224.0.0.2)
Next, phone B in Figure 7-8 issues an IGMP V2 Membership Report message indicating that it is joining
this multicast group. In addition, phone B sends a SIP 200 OK Response back to Unified CM indicating
an SDP media attribute of recvonly in response to the previous SIP INVITE. Meanwhile, the MoH server
has been sourcing RTP audio to this MoH multicast group address and, upon joining the multicast group,
phone B begins receiving the one-way MoH stream.
7-27
Chapter 7
Media Resources
Music on Hold
When the user at phone A presses the Resume softkey, phone A sends a SIP INVITE with an SDP
connection information indication of phone A's IP address and media attribute indications of phone A's
receiving RTP port and sendrecv. Unified CM then instructs phone B to disconnect from the multicast
MoH stream by means of a SIP INVITE with an SDP connection information indication of 0.0.0.0 and
a media attribute indication of inactive. A SIP 200 OK Response is sent back from phone B to
Unified CM, indicating an SDP media attribute of inactive.
Next Unified CM sends a SIP INVITE to phone B, and phone B responds with a SIP 200 OK Response
with an SDP connection information indication of phone B's IP address and media attribute indications
of phone B's receiving RTP port and sendrecv. Unified CM responds by sending a SIP ACK to phone B
with an SDP connection information indication of phone A's IP address and a media attribute of
phone A's receiving RTP port number. Likewise, Unified CM forwards a SIP 200 OK Response to
phone A's original resuming SIP INVITE, with an SDP connection information indication of phone B's
IP address and a media attribute of phone B's receiving RTP port number. Phone B then sends an
IGMP V2 Leave Group message to 224.0.0.2 to indicate that the multicast stream is no longer needed.
Finally, the RTP two-way audio stream between phones A and B is reestablished.
Note
The call flow diagrams in Figure 7-8 and Figure 7-9 assume that an initial call exists between phones A
and B, with a two-way RTP audio stream. These diagrams are representative of call flows and therefore
include only the pertinent traffic required for proper MoH operation. Thus, keep-alives, some
acknowledgements, progression indications, and other miscellaneous traffic have been eliminated to
better illustrate the interaction. The initial event in each diagram is the Hold softkey action performed
by phone A.
7-28
Chapter 7
Media Resources
Music on Hold
Figure 7-9
MoH Server
Phone B
Subscriber
Phone A
IP
IP
SCCP: StartMediaTransmission
Phone B IP Address, RTP port#
RTP Stream: Active one-way
Phone B on HOLD
(RESUME pressed)
RTP Stream: Active one-way
SIP: INVITE (SDP: 0.0.0.0,
a=inactive)
Unified CM then sends a SIP INVITE to phone B, and phone B responds back with a SIP 200 OK
Response indicating SDP connection information with phone B's IP address and media attribute
indications of phone B's receiving RTP port number and sendrecv. Unified CM then sends a SCCP
StartMediaTransmission message to the MoH server, with phone B's address and receiving RTP port
7-29
Chapter 7
Media Resources
number. This is followed by a SIP ACK from Unified CM to phone B indicating SDP connection
information of the Unified CM IP address and a media attribute of sendonly. Meanwhile, the MoH server
begins sourcing RTP audio to phone B, and phone B begins receiving the one-way MoH stream.
When the user at phone A presses the Resume softkey, phone A sends a SIP INVITE with an SDP
connection information indication of phone A's IP address and media attribute indications of phone A's
receiving RTP port and sendrecv. Unified CM then instructs phone B to disconnect from the multicast
MoH stream by means of a SIP INVITE with an SDP connection information indication of 0.0.0.0 and
a media attribute indication of inactive. A SIP 200 OK Response is sent back from phone B to
Unified CM, indicating an SDP media attribute of inactive. Then Unified CM sends an SCCP
StopMediaTransmission message to the MoH server, causing the MoH server to stop forwarding the
MoH stream to phone B.
Next Unified CM sends a SIP INVITE to phone B, and phone B responds with a SIP 200 OK Response
with an SDP connection information indication of phone B's IP address and media attribute indications
of phone B's receiving RTP port and sendrecv. Unified CM responds by sending a SIP ACK to phone B,
with an SDP connection information indication of phone A's IP address and a media attribute of
phone A's receiving RTP port number. Likewise, Unified CM forwards a SIP 200 OK Response to
phone A's original resuming SIP INVITE with an SDP connection information indication of phone B's
IP address and a media attribute of phone B's receiving RTP port. Finally, the RTP two-way audio stream
between phones A and B is reestablished.
7-30
Chapter 7
Media Resources
Capacity Planning for Music on Hold
For capacity information on PVDM3 modules, refer to the High-Density Packet Voice Video Digital
Signal Processor Module for Cisco Unified Communications Solutions data sheet, available at
http://www.cisco.com/en/US/prod/collateral/modules/ps3115/data_sheet_c78-553971.html
For capacity information on PVDM4 modules, refer to the Cisco Fourth-Generation Packet Voice Digital
Signal Processor Module for Cisco Unified Communications Solutions Data Sheet, available at
http://www.cisco.com/c/en/us/products/routers/4000-series-integrated-services-routers-isr/datashe
et-listing.html
Co-resident deployment
The term co-resident refers to two or more services or applications running on the same server. In a
co-resident deployment, the MoH feature runs on any server (either publisher or subscriber) in the
cluster that is also running the Unified CM software.
Standalone deployment
A standalone deployment, places the MoH feature on a dedicated media resource server node within
the Unified CM cluster. This server acts as neither a publisher or a subscriber. That is, the Cisco IP
Voice Media Streaming Application service is the only service enabled on the server. The only
function of this dedicated server is to send MoH streams to devices within the network.
7-31
Chapter 7
Media Resources
The following two MoH Server Configuration parameters affect MoH server capacity:
Maximum half-duplex
streams
1,000
2)
= 976
- (12
Therefore, in this example, the Maximum Half Duplex Streams parameter would be configured with
a value of no more than 976. Each of the multicast MoH audio sources will have an automatic
multicast RTP stream created for each enabled MoH codec.
Failure to configure these parameters properly could lead to under-utilization of MoH server resources
or failure of the server to handle the network load. For details on how to configure the service
parameters, refer to the Cisco Unified Communications Manager Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
Note
The maximum limit of 1,000 sessions per MoH server applies to unicast, multicast, or simultaneous
unicast and multicast RTP streams. The limit represents the recommended maximum number of MoH
streams a platform can support, irrespective of the transport mechanism.
7-32
Chapter 7
Media Resources
High Availability for Media Resources
Resource Provisioning
When provisioning for co-resident or standalone MoH server configurations, network administrators
should consider the type of transport mechanism used for the MoH audio streams. If using unicast MoH,
each device on hold requires a separate MoH stream. However, if using multicast MoH and only a single
audio source, then only a single MoH stream is required for each configured MoH codec type, no matter
how many devices of that type are on hold.
For example, given a cluster with 30,000 phones and a 2% hold rate (only 2% of all endpoint devices are
on hold at any given time), 600 MoH streams or sessions would be required. Given a unicast-only MoH
environment, one co-resident (or standalone) MoH server running on a Cisco Unified Computing System
(UCS) using the 10K OVA template would be required to handle this load.
By comparison, a multicast-only MoH environment with 36 unique MoH audio streams, for example,
would require one co-resident MoH server. These 36 unique multicast streams could be provisioned in
any one of the following ways:
In the preceding examples, the 2% hold rate is based on 30,000 phones and does not take into account
gateways or other endpoint devices in the network that are also capable of being placed on hold. You
should consider these other devices when calculating a hold rate because they could potentially be placed
on hold just as the phones can.
The preceding calculations also do not provide for MoH server redundancy. If an MoH server fails or if
more than 2% of the users go on hold at the same time, there are no other MoH resources in this scenario
to handle the overflow or additional load. Your MoH resource calculations should include enough extra
capacity to provide for redundancy. Additional MoH servers can be provisioned for redundancy or high
availability as explained in the section on High Availability for Media Resources, page 7-33.
The system defines a default media resource group that is not visible in the user interface. All
resources are members of this default MRG when they are created. When using MRGs to control
access to resources, it is necessary to move the resources out of the default MRG by explicitly
7-33
Chapter 7
Media Resources
configuring them in some other MRG. If the desired effect is for resources to be available only as a
last resort for all calls, then the resources may remain in the default group. Also, if no control over
resources is necessary, they may remain in the default group.
Consumers of media resources use resources first from any media resource group (MRG) or media
resource group list (MRGL) that their configuration specifies. If the required resource is not
available, the default MRG is searched for the resource. For simple deployments, the default MRG
alone may be used.
Use media resource groups (MRGs) and media resource group lists (MRGLs) to provide sharing of
resources across multiple Unified CMs. If you do not use MRGs and MRGLs, the resources are
available to a single Unified CM only.
MRGLs will use MRGs in the order that they are listed in the configuration. If one MRG does not
have the needed resource, the next MRG is searched. If all MRGs are searched and no resource is
found, the search terminates.
Within an MRG, resources are allocated based on their order in their configuration even though
Unified CM Administration displays the devices in an MRG in alphabetical order. If you want media
resources to be allocated in a specific order, Cisco recommends that you create a separate MRG for
each individual resource and use MRGLs to specify the order of allocation.
When there are multiple devices providing the same type of resource within an MRG, the algorithm
for allocating that resource load-balances across all those devices. Cisco Unified CM uses a
throttling mechanism to load balance across MTP and transcoder resources using the MTP and
Transcoder Resource Throttling Percentage service parameter, which defines a percentage of the
configured number of MTP or transcoder resources. When the number of active MTP or transcoder
resources is equal to or greater than the percentage that is configured for this parameter, Cisco
Unified CM stops sending calls to this resource and hunts through the MRGL (including the default
MRG) one time to find a resource that uses matching codecs on both sides of the call. If Cisco
Unified CM cannot find an available resource with matching codecs, it returns to the top of the
MRGL to repeat the search, which then includes those resources that are in a throttled state and that
match a smaller subset of capabilities for the call. Cisco Unified CM extends the call to the resource
that is the best match for the call when such a resource is available. The call fails when Cisco
Unified CM cannot allocate a resource for the call.
Unified CM server-based software MTPs are pass-through enabled by default. Cisco IOS Enhanced
MTP devices can be configured to support codec pass-through or non-codec pass-through modes. If
a codec pass-through MTP is required and if, after the first iteration through the MRGL (including
the default MRG), a codec pass-through MTP is not found, then there will be a second iteration that
will ignore codec pass-through capabilities.
An MRG may contain multiple types of resources, and the appropriate resource will be allocated
from the group based on the feature needed. MTPs and transcoders are a special case because a
transcoder may also be used as an MTP. For example, when both MTPs and transcoders exist in the
same MRG and an MTP is required, the allocation is done based on the order in which the resources
appear in the MRG. If transcoder devices appear earlier than MTPs in the MRG, transcoder
resources will be allocated for the MTP requirement until the transcoder resources are exhausted and
then the system will start allocating MTPs. For this reason, it is important to consider the order of
resources when creating MRGs and MRGLs.
MRGs can also be used to group resources of similar types. As explained in the example above,
because a transcoder is a more expensive resource, Cisco recommends grouping transcoders and
MTPs into separate MRGs and invoking the right resource by adding MRGs to the MRGL in
appropriate order.
You can also use MRGs and MRGLs to separate resources based on geographical location, thereby
conserving WAN bandwidth whenever possible.
7-34
Chapter 7
Media Resources
High Availability for Media Resources
Ensure that the media resources themselves have configurations that prevent further invocation of
other media resources. For example, if an MTP is inserted into a call and the codec configured on
that MTP does not match the one needed by Unified CM for the call, then a transcoder may also be
invoked. A frequent mistake is to configure an MTP for G.729 or G.729b when Unified CM needs
G.729a.
7-35
Chapter 7
Media Resources
Deployment Models
This section examines where and when the MTP and transcoding resources are used within the following
three enterprise IP Telephony deployment models:
Single-Site Deployments
In a single-site deployment, there is no need for transcoding because there are no low-speed links to
justify the use of a low bit-rate (LBR) codec. Some MTP resources might be required in the presence of
a significant number of devices that are not compliant with H.323v2, such as older versions of Microsoft
NetMeeting or certain video devices. MTP resources may be required for DTMF conversion if SIP
endpoints are present (see Named Telephony Events (RFC 2833), page 7-8.)
In a single-site deployment, if Unified CM receives an inbound call from an SCCP-based Cisco Unified
IP Phone 7940 or 7960, the media capabilities of the calling device are not available when the call is
initiated, and most of the SIP PSTN service providers require an early offer. In this case, Unified CM
must insert an MTP and use its IP address and UDP port number to advertise all supported audio codecs
(after region-pair filtering) in the Offer SDP of the initial INVITE sent over the outbound SIP trunk.
7-36
Chapter 7
Media Resources
Design Considerations for Media Resources
Figure 7-10
VM/UM Server
G.711 only
Unified CM
Cluster
M
Router/GW
Router/GW
IP WAN
IP
IP
IP
Xcode
DSP Farm
Xcode
Unified CM uses media resource groups (MRGs) to enable sharing of MTP and transcoding resources
among the Unified CM servers within a cluster. In addition, when using an LBR codec (for example,
G.729a) for calls that traverse different regions, the transcoding resources are used only if one (or both)
of the endpoints is unable to use the LBR codec.
In Figure 7-10, Unified CM knows that a transcoder is required and allocates one based on the MRGL
and/or MRG of the device that is using the higher-bandwidth codec. In this case it is the VM/UM server
that determines which transcoder device is used. This behavior of Unified CM is based on the
assumption that the transcoder resources are actually located close to the higher-bandwidth device. If
this system was designed so that the transcoder for the VM/UM server was located at the remote site,
then G.711 would be sent across the WAN, which would defeat the intended design. As a result, if there
are multiple sites with G.711-only devices, then each of these sites would need transcoder resources
when an LBR is run on the WAN.
The placement of other resources is also important. For example, if a conference occurs with three
phones at a remote site and the conference resource is located in the central (call processing) site, then
three media streams are carried over the WAN. If the conference resource were local, then the calls
would not traverse the WAN. It is necessary to consider this factor when designing the bandwidth and
call admission control for your WAN.
7-37
Chapter 7
Media Resources
A distributed call processing deployment might need transcoding and MTP services in the following
situations:
With current versions of Cisco applications, it is possible and recommended to avoid the use of
transcoding resources. There might be specific instances where G.711 on a specific device cannot
be avoided.
Some endpoints (for example, video endpoints) do not support the H.323v2 features.
Figure 7-11
Region A
G.711 only
IVR
Region B
Unified CM A
Unified CM B
Router/GW
IP
Router/GW
IP
IP WAN
IP
IP
Xcode
Xcode
Transcoding
A to A = G.711
A to B = G.729
= DSP Farm
Regi on BW M atrix
B to B = G.711
B to A = G.729
77306
Region BW M atrix
Unified CM uses media resource groups (MRGs) to enable sharing of MTP and transcoding resources
among the Unified CM servers within a cluster. In addition, for calls across intercluster trunks, MTP and
transcoding resources are used only when needed, thus eliminating the need to configure the MTP
service for applications that do not support LBR codecs.
The following characteristics apply to distributed call processing deployments:
Only the intercluster calls that require transcoding will use the MTP service. For example, if both
endpoints of a call are capable of using a G.729 codec, no transcoding resources will be used.
Sharing MTP resources among servers within a cluster provides more efficient resource utilization.
7-38
Chapter 7
Media Resources
Design Considerations for Media Resources
If such paths cannot be avoided, voice quality can generally be improved by using a higher bandwidth,
low-compression codec, such as the G.711 or G.722 codecs, which are recommended wherever such
paths are anticipated. Use of lower bandwidth, higher compression codecs in such scenarios is not
recommended.
Codec Selection
If you need multiple codecs for MoH deployment, configure them in the IP Voice Media Streaming
Application service parameter Supported MoH Codecs under the Clusterwide Unified CM Service
Parameters Configuration. From the Supported MoH Codecs list under the Clusterwide Parameters,
select all the desired codec types that should be allowed for MoH streams. By default, only G.711
mu-law is selected. To select another codec type, click on it in the scrollable list. For multiple selections,
hold down the CTRL key and use the mouse to select multiple codecs from the scrollable list. The actual
codec used for a MoH event is determined by the Region settings of the MoH server and the device being
put on hold (IP phone, gateway, and so forth). Therefore, assign the proper Region setting to your MoH
servers and configure the desired Region Relationships to control the codec selection of MoH
interactions.
Note
If you are using the G.729 codec for MoH audio streams, be aware that this codec is optimized for speech
and it provides only marginal audio fidelity for music.
Multicast Addressing
Proper IP addressing is important for configuring multicast MoH. Addresses for IP multicast range from
224.0.1.0 to 239.255.255.255. The Internet Assigned Numbers Authority (IANA), however, assigns
addresses in the range 224.0.1.0 to 238.255.255.255 for public multicast applications. Cisco strongly
discourages using public multicast addresses for music on hold. Instead, Cisco recommends that you
configure multicast MoH audio sources to use IP addresses in the range 239.1.1.1 to 239.255.255.255,
which is reserved for administratively controlled applications on private networks.
Furthermore, you should configure multicast audio sources to increment on the IP address and not the
port number, for the following reasons:
7-39
Chapter 7
Media Resources
When configuring multiple multicast MoH servers, assign a different base multicast IP address and/or
range to each MoH server. If multiple MoH servers are transmitting to the same multicast IP address,
then when an endpoint joins the multicast group address, it will receive multiple MoH streams (from
different MoH servers).
Deploy separate MoH servers, with one server configured as a unicast MoH server and the second
server configured as a multicast MoH server.
Deploy a single MoH server with two media resource groups (MRGs), each containing the same
MoH server, with one MRG configured to use multicast for audio streams and the second MRG
configured to use unicast.
In either case, you must configure at least two MRGs and at least two media resource group lists
(MRGLs). Configure one unicast MRG and one unicast MRGL for those endpoints requiring unicast
MoH. Likewise, configure one multicast MRG and one multicast MRGL for those endpoints requiring
multicast MoH.
When deploying separate MoH servers, configure one server without multicast enabled (unicast-only)
and configure a second MoH server with multicast enabled. Assign the unicast-only MoH media
resource and the multicast-enabled MoH media resource to the unicast and multicast MRGs,
respectively. Ensure that the Use Multicast for MoH Audio box is checked for the multicast MRG but
not for the unicast MRG. Also assign these unicast and multicast MRGs to their respective MRGLs. In
this case, an MoH stream is unicast or multicast based on whether the MRG is configured to use
multicast and then on the server from which it is served.
When deploying a single MoH server for both unicast and multicast MoH, configure the server for
multicast. Assign this same MoH media resource to both the unicast MRG and the multicast MRG, and
check the Use Multicast for MoH Audio box for the multicast MRG. In this case, an MoH stream is
unicast or multicast based solely on whether the MRG is configured to use multicast.
Note
When configuring the unicast MRG, do not be confused by the fact that the MoH media resource you
are adding to this MRG has [Multicast] appended to the end of the resource name even though you are
adding it to the unicast MRG. This label is simply an indication that the resource is capable of being
multicast, but the Use Multicast for MoH Audio box determines whether the resource will use unicast
or multicast.
7-40
Chapter 7
Media Resources
Design Considerations for Media Resources
In addition, you must configure individual devices or device pools to use the appropriate MRGL. You
can place all unicast devices in a device pool or pools and configure those device pools to use the unicast
MRGL. Likewise, you can place all multicast devices in a device pool or pools and configure those
device pools to use the multicast MRGL. Optionally, you can configure individual devices to use the
appropriate unicast or multicast MRGL. Lastly, configure a User Hold Audio Source and Network Hold
Audio Source for each individual device or (in the case of phone devices) individual lines or directory
numbers to assign the appropriate audio source to stream.
When choosing a method for deploying both multicast and unicast MoH in the same cluster, an important
factor to consider is the number of servers required. When using a single MoH server for both unicast
and multicast, fewer MoH servers are required throughout the cluster. Deploying separate multicast and
unicast MoH servers will obviously require more servers within the cluster.
7-41
Chapter 7
Media Resources
control depends on the type of MoH stream. If multicast MoH is streamed, then call admission control
will not consider the 24 kbps being consumed (therefore, QoS on the downstream WAN interfaces
should be provisioned accordingly). However, if unicast MoH is streamed, call admission control will
subtract 24 kbps from the available WAN bandwidth (step 2).
The preceding example might seem to imply that unicast MoH should be streamed across the WAN.
However, this is merely an example used to illustrate locations-based call admission control with MoH
and is not intended as a recommendation or endorsement of this configuration. As stated earlier,
multicast MoH is the recommended transport mechanism for sending MoH audio streams across the
WAN.
Note
Figure 7-12
Phone B
Central Site
Branch Office
PSTN
M
M
Gateway/SRST
Router
Hold
IP WAN
IP
IP
Phone C
IP
CAC Bandwidth =
24Kbps/1 call
96853
IP
7-42
Chapter 7
Media Resources
Design Considerations for Media Resources
7-43
Chapter 7
Media Resources
The following example illustrates how MoH streaming works for up to 500 locations:
For branches 1 to 500, the centralized PSTN gateway or set of gateways is/are configured with
MRGL pointing to MoH server node MoH_1 (with base multicast address of 239.1.1.1), with all
phones at each branch pointing to one of the 500 audio sources configured on the cluster. Thus,
phones in Branch 1 point to audio source 1, which on MoH_1 server is 239.1.1.1 to 4 (depending on
codec and assuming audio sources are configured in order); phones in Branch 2 point to audio
source 2, which on MoH_1 server is 239.1.1.5 to 8; phones in Branch 3 point to audio source 3,
which on MoH_1 server is 239.1.1.9 to 12; and so on up to phones in Branch 500 pointing to audio
source 501, which on MoH_1 server is 239.1.8.197 to 200.
Non-Fallback Mode
When the WAN is up and the phones are controlled by Unified CM, this configuration can eliminate
the need to forward MoH across the WAN to remote branch sites by providing locally sourced MoH.
Fallback Mode
When SRST is active and the branch devices have lost connectivity to the central-site Unified CM,
the branch router can continue to provide multicast MoH.
When using the live feed option in either scenario, the SRST router provides redundancy by monitoring
the live feed input, and it will revert to streaming MoH from a file in flash if the live feed is disconnected.
You can use only a single multicast address and port number per SRST router to provide multicast MoH;
therefore, the SRST router does not support streaming from both the live feed and the flash file at the
same time. In addition, the SRST router can stream only a single audio file from flash.
Note
An SRST license is required regardless of whether the SRST functionality will actually be used. The
license is required because the configuration for streaming MoH from branch router flash is done under
the SRST configuration mode and, even if SRST functionality will not be used, at least one
max-ephones and one max-dn must be configured.
Non-Fallback Mode
During non-fallback mode (when the WAN is up and SRST is not active), the branch SRST or E-SRST
router can provide multicast MoH to all local Cisco Unified Communications devices. To accomplish
this, you must configure a Unified CM MoH server with an audio source that has the same multicast IP
address and port number as configured on the branch router. The audio source multicast IP address and
port number used on the branch router can correspond to the multicast address and port number of either
an audio source file or the fixed audio source of the centralized Unified CM MoH server. In this scenario,
because the multicast MoH audio stream is always coming from the SRST or E-SRST router, it is not
necessary for the central-site MoH server audio source to traverse the WAN.
To prevent the central-site audio stream(s) from traversing the WAN, use one of the following methods:
7-44
Chapter 7
Media Resources
Design Considerations for Media Resources
Figure 7-13 illustrates streaming multicast MoH from a branch router when it is not in fallback mode.
After phone A places phone C on hold, phone C receives multicast MoH from the local SRST router. In
this example, the MoH server is streaming a multicast audio source to 239.192.240.1 (on RTP port
16384); however, this stream has been limited to a maximum hop of one (1) to ensure that it will not
travel off the local MoH server's subnet and across the WAN. At the same time, the branch office SRST
router/gateway is multicasting an audio stream from either flash or a live feed. This stream is also using
239.192.240.1 as its multicast address and 16384 as the RTP port number. When phone A presses the
Hold softkey, phone C receives the MoH audio stream sourced by the SRST router.
Figure 7-13
Central Site
Branch Office
Max Hop (TTL)=1
M
M
SRST or E-SRST
Router
IP WAN
Phone A
IP
Phone B
Phone C
Hold
IP
IP
348694
MCast Address:
239.192.240.1
RTP Port: 16384
When using this method for delivering multicast MoH, configure all devices within the Unified CM
cluster to use the same user hold and network hold audio source and configure all branch routers with
the same multicast group address and port number. Because the user or network hold audio source of the
holder is used to determine the audio source, if you configure more than one user or network hold audio
source within the cluster, there is no way to guarantee that a remote holdee will always receive the local
MoH stream. For example, suppose a central-site phone is configured with an audio source that uses
group address 239.192.254.1 as its user and network hold audio source. If this phone places a remote
device on hold, the remote device will attempt to join 239.192.254.1 even if the local router flash MoH
stream is sending to multicast group address 239.192.240.1. If instead all devices in the network are
configured to use the user/network hold audio source with multicast group address 239.192.240.1 and
all branch routers are configured to multicast from flash on 239.192.240.1, then every remote device will
receive the MoH from its local router.
7-45
Chapter 7
Media Resources
Fallback Mode
During fallback mode (when the WAN is down and SRST is active), the branch SRST router can stream
multicast MoH to all analog and digital ports within the chassis, thereby providing MoH to analog
phones and PSTN callers.
The branch router's configuration for fallback mode multicast MoH is the same as the normal operation
configuration. However, which multicast address you configure on the router depends on the intended
operation. If you want the branch router to provide multicast MoH to devices only in fallback mode (for
example, if MoH received by remote devices is to be sourced from the central-site MoH server during
non-fallback mode), then the multicast address and port number configured on the SRST router should
not overlap with any of the central-site MoH server audio sources. Otherwise, remote devices might
continue to receive MoH from the local router flash, depending on the configured user/network hold
audio sources.
Note that, once the branch SRST/gateway router is configured to provide multicast MoH, the router will
continue to multicast the MoH stream even when not in fallback mode.
It is also possible to configure the fallback mode to use Cisco Unified Communications Manager Express
(Unified CME) in SRST mode, referred to as Enhanced SRST (E-SRST). Fallback mode behavior is still
the same, but the configuration commands are slightly different. SRST commands are entered under the
Cisco IOS call-manager-fallback construct, while the commands for E-SRST mode are entered under
telephony-service.
There are four methods of providing multicast MoH via SRST:
For more details on configuration of Cisco Unified SRST and E-SRST, refer to the following
documentation:
For more information on using Cisco Unified SRST as a multicast MoH resource, refer to the section
on Integrating Cisco Unified Communications Manager and Cisco Unified SRST to Use Cisco
Unified SRST as a Multicast MoH Resource in the latest version of the Cisco Unified SCCP and SIP
SRST System Administrator Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps2169/products_installation_and_configur
ation_guides_list.html
7-46
Chapter 7
Media Resources
Design Considerations for Media Resources
codec for all MoH audio streams traversing them. Because the G.729 codec is optimized for voice and
not music applications, you should use G.729 only across the WAN/MAN links, where the bandwidth
savings far outweighs the lower quality afforded by G.729 for MoH transport.
Unlike with centralized multisite deployments, in situations where G.711 might be required for MoH
audio streams traveling across a WAN, MoH audio streams cannot be forced to G.711 in a distributed
multisite deployment. Even when MoH servers are placed in a separate Unified CM region and the G.711
codec is configured between this region and the intercluster or SIP trunk's region, the codec of the
original voice call is maintained when a call between the two clusters is placed on hold by either phone.
Because these intercluster calls are typically encoded using G.729 for bandwidth savings, a MoH stream
from either cluster will also be encoded using G.729.
Another option is to provision multicast MoH for intercluster calls across an intercluster trunk (ICT) or
SIP trunk. This allows endpoints in one Unified CM cluster to hear multicast MoH streamed from
another Unified CM cluster, while making more efficient use of intercluster bandwidth. A properly
designed IP Multicast environment is required to take advantage of this feature. For more information
on IP Multicast, refer to the documentation available at
http://www.cisco.com/en/US/products/ps6552/products_ios_technology_home.html
Proper multicast address management is another important design consideration in the distributed
intercluster environment. All MoH audio source multicast addresses must be unique across all
Unified CM clusters in the deployment to prevent possible overlap of streaming resources throughout
the distributed network.
7-47
Chapter 7
Media Resources
7-48
CH A P T E R
Collaboration Endpoints
Revised: July 30, 2015
A variety of endpoints can be used in a Cisco Collaboration deployment. These endpoints range from
gateways that support ordinary analog phones in an IP environment to an extensive set of native IP
phones offering a range of capabilities.
When deploying endpoints, you need to consider several factors, including authentication, upgrades,
signaling protocol, Quality of Service (QoS), and so forth. The collaboration system must be designed
appropriately to accommodate these factors.
This chapter summarizes various types of collaboration endpoints and covers design and deployment
considerations including high availability and capacity planning. The collaboration endpoints covered
in this chapter can be categorized into the following major types:
The sections listed above provide information about each endpoint type, including deployment
considerations. That information is followed by a discussion related to high availability, capacity
planning, and design considerations for effectively deploying endpoints.
Use this chapter to understand the range of available endpoint types and the high-level design
considerations that go along with their deployment.
8-1
Chapter 8
Collaboration Endpoints
Described in:
Revision Date
TM
Cisco IP Phone 7800 Series, 8800 Series, and the Desk Phones, page 8-8
dual-screen Cisco TelePresence MX800
Video Endpoints, page 8-15
The following topics were removed from this chapter because they do not apply to Cisco Collaboration June 15, 2015
System Release 11.x:
Cisco Unified IP Phone 6900 Series and Cisco Unified IP Conference Station 7937
8-2
Chapter 8
Collaboration Endpoints
Collaboration Endpoints Architecture
Figure 8-1
Enterprise
nterprise Collaboration
Colla
Call Control
Enterprise Collaboration
Services
Cisco Unified CM
Express
Cisco
sco Busines
Business
Edition
Cisco Unity
Connection
Cisco Unified CM
Enterprise Collaboration
Media Resources
Enterprise
Collaboration Edge
Cisco IM
and Presence
Cisco Collaboration
Cloud Serv
Services
r ices
Cisco
TelePresence
Server / MCU
WebEx
b
Cisco Expressway
Mobile & Remote
Access
IOS Media
Resources
(Conferencing, MTPs, etc.)
Spark
Internet
rnet
Enterprise IP
P
(LAN, WAN, MAN)
M
Mobile
Data
Network
N
PSTN
PS
STN
Mobile
Voice
Network
Desk
k Phones
Ph
Enterprise
Wireless
(802.11 WLAN)
Mobile Endpoints
Analog Endpoints
Virtualized
Experience Clients
Software-Based
oftware Base
Endpoints
Wireless
s Endpoints
COLLABORATION
ENDPOINTS
348820
Video Endpoints
8-3
Chapter 8
Collaboration Endpoints
While there are multiple options when deploying collaboration call control for voice and video services,
each call control platform provides endpoint registration, call setup and routing services, and access to
provisioned media resources. The high-level call control interactions between endpoints and the
enterprise Cisco Unified Communications Manager is described in the following sub-section.
When connected to the access switch, if the endpoint is not plugged in to a power source, it attempts
to obtain power from the switch (Power over Ethernet). Wireless and mobile endpoints are not
connected to the enterprise network via Ethernet and therefore always derive power from a battery
or power outlet.
2.
Once power is obtained, if device security is enabled, the endpoint presents its credentials to the
security server or network authentication infrastructure.
3.
If it is allowed to use the network, the endpoint obtains its network parameters such as IP address,
Domain Name Service (DNS) servers, gateway address, and so forth, either through static
provisioning in the endpoint or through Dynamic Host Control Protocol (DHCP).
4.
The endpoint also obtains a TFTP server address either through static provisioning or through
DHCP options.
5.
The endpoint then uses the TFTP server address to obtain its configuration files that, among other
parameters, details the call processing server(s) or router(s) that the endpoint may associate and
register with, the directory numbers that the endpoint must support, and so forth.
6.
The endpoint registers with the call processing platform and is available for use.
To confirm which endpoints support registration to Cisco Unified CM, refer to the endpoint data sheets
listed in various other sections of this chapter.
8-4
Chapter 8
Collaboration Endpoints
Analog Endpoints
Enable Quality of Service (QoS) and call admission control on the network to ensure optimal quality
of voice and video so that enterprise communications are as clear and precise as possible.
Configure only the G.711 codec for phones that will be connected to a terminal teletype (TTY)
device or a Telephone Device for the Deaf (TDD). Although low bit-rate codecs such as G.729 are
acceptable for audio transmissions, they do not work well for TTY/TDD devices if they have an
error rate higher than 1% Total Character Error Rate (TCER).
Voice Activity Detection (VAD) does not appear to have an effect on the quality of the TTY/TDD
connection, so it may be disabled or enabled. However, Cisco recommends leaving VAD (also
known as silence suppression) disabled on Unified CM call control and using the no vad command
on H.323 and Cisco IOS SIP dial-peers.
Configure the appropriate regions and device pools in Unified CM to ensure that the TTY/TDD
devices always use G.711 codecs.
Connect the TTY/TDD to the Cisco Unified Communications network in either of the following
ways:
Direct connection (Recommended method)
Plug a TTY/TDD with an RJ-11 analog line option directly into a Cisco FXS port. Any Cisco
voice gateway with an FXS port will work. Cisco recommends this method of connection.
Acoustic coupling
Place the IP phone handset into a coupling device on the TTY/TDD. Acoustic coupling is less
reliable than an RJ-11 connection because the coupling device is generally more susceptible to
transmission errors caused by ambient room noise and other factors.
If stutter dial tone for audible message waiting indication (AMWI) is required, use an analog phone
in conjunction with an FXS port on the Cisco VG Analog Gateways or Analog Telephony Adaptor
(ATA). In addition, most Cisco IP Phones support stutter dial tone.
When you deploy immersive Cisco TelePresence rooms, ensure that ample room is provided to
accommodate and provide for unimpeded movement of wheel chairs and other assistive
conveyances.
Analog Endpoints
An analog gateway typically is used to connect analog devices such as fax machines, modems,
telecommunications device for the deaf (TDD)/teletypewriter (TTY), and analog phones, to the VoIP
network so that the analog signal can be packetized and transmitted over the IP network. Analog
gateways also provide physical connectivity to the PSTN and other traditional telephony equipment such
as PBXs and key systems. Analog gateways include Cisco IOS router-based analog interface or service
modules as well as fixed-port standalone gateways. Generally analog gateways rely on Cisco
8-5
Chapter 8
Collaboration Endpoints
Analog Endpoints
Unified CM, Cisco Business Edition, Unified CM Express, and even Survivable Remote Site Telephony
(SRST) for call control, supplementary services, and in some cases interface registration and
configuration. Call control protocols supported across Cisco analog gateways include SIP, H.323, SCCP,
and Media Gateway Control Protocol (MGCP).
8-6
Chapter 8
Collaboration Endpoints
Analog Endpoints
start line typically requires specialized equipment on the PSTN side, which typically makes these lines
more expensive than loop start lines. However, with ground start lines, because a loss of current on the
line is immediately detected on the far side of the analog connection, the gateway or PBX gets immediate
indication regarding connects and disconnects, thus providing better control over the connection. In
addition, a ground start trunk reduces the possibility of "glare," or the collision of simultaneous
incoming and outgoing calls on the line.
E&M interfaces support different signaling methods, including wink start and immediate start. Wink
start is the most common form of E&M signaling, and it relies on a "wink" sequence (on-hook, off-hook,
on-hook) indication from the far end in response to an initial off-hook indication at the call origination
side before digits can be sent over the interface. In contrast, immediate start signaling relies on a brief
pause rather than a response from the far end after the initial off-hook indication before digits are sent.
The analog interface type used in a particular deployment will ultimately be dictated by the interface
supported by the PSTN provider or by the equipment deployed in the case of internal analog connections.
In all cases, you should use the supported method of signaling for the analog connection type that
provides the most visibility and control of the line. For example, with FXS or FXO, ground start is
preferred over loop start because of the end-to-end line current which, when broken, can be detected
immediately. Likewise, with E&M, wink start is preferred over immediate start because of the positive
indication from the far end that digits can be sent.
For additional information on Cisco analog telephony signaling, refer to the documentation available at
http://www.cisco.com/c/en/us/tech/voice/analog-signaling-e-m-did-fxs-fxo/index.html
Paging Systems
In some IP telephony deployments, the enterprise IP PBX is integrated with a paging system that allows
users to call an extension on the system that forwards the audio broadcast to overhead loudspeakers.
These overhead paging systems are useful in workshops, parking lots, and open plant areas where a
called party is not near a telephone handset. Integration to these paging systems is done using an analog
interface module port.
Cisco analog gateways and interface modules support all traditional analog port types used for paging
system integration, including FXO, FXS, and E&M. When integrating with overhead paging systems,
ensure that the appropriate analog interface module port type, signaling, and configuration are used as
required by the paging system being integrated. The port type, signaling, and configuration will
ultimately be dictated by the paging system.
An example of an E&M interface integration to an overhead paging system is available at
http://www.cisco.com/c/en/us/support/docs/voice/analog-signaling-e-m-did-fxs-fxo/27627-e-mpag
ing.html
Quality of Service
When configuring network-level quality of service (QoS), Cisco analog gateways such as the standalone
Cisco VG Series and the Cisco IOS-based analog interface modules can be trusted and their packet
markings honored. By default they mark their voice media and signaling packets with appropriate
Layer 3 values (voice media as DSCP 46 or PHB EF; call signaling as DSCP 24 or PHB CS3), which
match Cisco QoS recommendations for appropriate voice media and signaling marking, so as to ensure
end-to-end voice quality on a converged network.
8-7
Chapter 8
Collaboration Endpoints
Desk Phones
Desk Phones
The Cisco IP Phone portfolio includes the following family of desk phones:
Note
When two Expansion Modules are used with a single phone, the second module must be the same model
as the first one.
For more information about the Cisco Unified IP Phone 7900 Series, refer to the data sheets and
documentation at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phone-7900-series/inde
x.html
Note
Starting with Cisco IP Phone 7800 Series firmware version 10.3.1, the 7800 Series supports Cisco
Expressway as an alternative to VPN access. Expressway provides enterprise firewall traversal for
7800 Series voice calls.
8-8
Chapter 8
Collaboration Endpoints
Desk Phones
For more information about the Cisco IP Phone 7800 Series, refer to the data sheets and product
documentation available at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phone-7800-series/inde
x.html
Note
Starting with Cisco IP Phone 8800 Series firmware version 10.3.1, the 8800 Series supports Cisco
Expressway as an alternative to VPN access. Expressway provides enterprise firewall traversal for
8800 Series voice and video calls.
For more information about the Cisco IP Phone 8800 Series, refer to the data sheets and documentation
at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phone-8800-series/inde
x.html
8-9
Chapter 8
Collaboration Endpoints
Desk Phones
For more information about the Cisco Unified IP Phone 9900 Series, refer to the data sheets and product
documentation at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phones-9900-series/ind
ex.html
The Cisco Unified IP 8900 and 9900 Series devices may also be equipped with up to three (dependent
on phone model) Cisco Unified IP Color Key Expansion Modules for administrative assistants and others
user who need to answer and/or determine the status of a number of lines beyond the current line
capability of their phone. These modules extend the capability of the Cisco Unified IP Phone 8900 and
9900 Series desk phones by adding an additional LCDs and buttons.
Some Cisco Unified IP Phone 8900 and 9900 Series models provide video capabilities either through a
built-in camera for the Cisco Unified IP Phone 8900 Series or through the Cisco Unified Video Camera
add-on accessory for the Cisco Unified IP Phone 9900 Series.
For more information about the Cisco Unified IP Phone 9900 and 8900 Series accessories, refer to the
data sheets and product documentation at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phone-9900-8900-serie
s-accessories/index.html
Deployment Considerations for the Cisco Unified IP Phone 8900 and 9900 Series
The Cisco Unified IP Phone 8900 and 9900 Series provide call capabilities that generate JTAPI events
that must be handled by applications that monitor the phone through CTI. These call features allow the
user to cancel an in-progress transfer or conference, or to perform a join or direct transfer of calls across
the same or different lines. If the monitoring applications have not been upgraded to versions that
properly handle these events, unexpected application behavior could result, including applications that
no longer have their view of the phone or call state in synchronization with the phone itself. Therefore,
by default, all applications are restricted from monitoring or controlling these phones.
For applications that have been upgraded to properly handle these new events, or for applications that
have verified that they are not impacted by these events, the administrator may enable the role of
Standard CTI Allow Control of Phones supporting Connected Xfer and conf in the application or
end-user configuration associated with the application. Only after this role has been enabled can the
application monitor or control these phones.
Note
The Cisco Unified SIP Phone 3900 Series does not support features such as CTI (for Jabber phone
control), speed dials, or Built-in Bridge for Silent Monitoring and Recording. The Cisco IP Phone
7800 Series or 8800 Series are recommended for environments requiring the full set of enterprise grade
IP telephony features.
8-10
Chapter 8
Collaboration Endpoints
Desk Phones
For more information about the Cisco Unified SIP Phone 3900 Series, refer to the data sheets and
documentation at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-sip-phone-3900-series/ind
ex.html
Cisco DX Series
The DX Series of endpoints delivers integrated Unified Communications, high-definition (HD) video,
and collaboration applications and services. The Cisco DX Series endpoints provide wideband audio and
HD video for enterprise-class communications with integrated 7 to 23 inch (model dependent)
multi-touch LCD display and front-facing camera. These devices run secure Android operating system
and provide access to a variety of integrated collaboration and communication applications including
calendaring, corporate directory searches, email, Jabber IM and presence, visual voicemail, and WebEx
conferencing as well as AnyConnect VPN for secure network attachment. In addition, as an open
Android platform, these devices are capable of accessing the Google Play store for access to many
third-party applications that enable additional features and functionality. These endpoints also provide
a variety of external interfaces for attaching accessories, including: HDMI for connecting external
devices such as a laptop or external display (model dependent); USB for keyboard, mouse, or wired
headset attachments; and Bluetooth for connecting a wireless headset, keyboard, and/or mouse, or for
leveraging Intelligent Proximity for Mobile Voice.
The DX Series endpoints support SIP signaling protocol for registering and communicating with Cisco
call processing platforms. Cisco Unified Communications Manager is required to deploy and support the
DX Series.
Note
Starting with Cisco DX Series firmware version 10.2.4, the DX Series supports Cisco Expressway as an
alternative to VPN access. Expressway provides enterprise firewall traversal for DX Series voice and
video, as well as the built-in Jabber IM application.
For more information about the Cisco DX Series, refer to the data sheets and documentation at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/desktop-collaboration-experiencedx600-series/index.html
Firmware Upgrades
Most commonly, and by default, IP phones upgrade their images using HTTP, which uses Port 6970,
from TFTP services integrated into one or more of the call processing platforms. When HTTP is not
available, IP phones use TFTP, which is a UDP-based protocol from the same TFTP services. With this
arrangement, all the phones obtain their images directly from these TFTP services. This method works
well for a relatively small number of phones or if all of the phones are located in a single campus region
that has a LAN environment with essentially unlimited bandwidth.
For larger deployments that use centralized call processing, upgrading phones in branch offices that are
connected to the central data center by low-speed WAN links, can require a large amount of data traffic
over the WAN. The same set of files will have to traverse the WAN multiple times, once for each phone.
8-11
Chapter 8
Collaboration Endpoints
Desk Phones
Transferring this amount of data is not only wasteful of the WAN bandwidth but can also take a long
time as each data transfer competes with the others for bandwidth. Moreover, due to the nature of TFTP
protocol, some phones might be forced to abort their upgrades and fall back to the existing version of
the code.
Note
During the upgrade, the Cisco IP Phones 7800, 8800, 8900, 9900, and DX Series stay in service, unlike
the 7900 Series phones. The 7800, 8800, 8900, 9900, and DX Series phones download and store the new
firmware in their memory while still maintaining their active status, and they reboot with the new
firmware only after a successful download.
Two methods are available to alleviate problems created by the need to upgrade phones over the WAN.
One method is to use a local TFTP server just for the upgrades. The administrator can place a TFTP
server in branch offices (particularly in branches that have a larger number of phones, or whose WAN
link is not speedy or robust), and can configure the phones in those offices to use that particular TFTP
server just for new firmware. With this change, phones will retrieve new firmware locally. This upgrade
method would require the administrator to pre-load the phone firmware on the TFTP server in the branch
and manually configure the TFTP server address in the load server parameter in the affected phone
configurations. Note that the branch router may be used as a TFTP server.
The second method to upgrade phones without using the WAN resources excessively is to use the Peer
File Sharing (PFS) feature. With this feature, typically only one phone of each model in the branch
downloads each new firmware file from the central TFTP server. Once the phone downloads the
firmware file, it distributes that file to other phones in the branch. This method avoids the manual loading
and configuration required for the load server method.
The PFS feature works when the same phone models in the same branch subnet arrange themselves in a
hierarchy (chain) when asked to upgrade. They do this by exchanging messages between themselves and
selecting the "root" phone that will actually perform the download. The root phone sends the firmware
file to the second phone in the chain using a TCP connection; the second phone sends the firmware file
to the third phone in the chain, and so on until all of the phones in the chain are upgraded. Note that the
root phone may be different for different files that make up the complete phone firmware.
8-12
Chapter 8
Collaboration Endpoints
Desk Phones
Note
In addition to using the inline power from the access switch or local wall power, a Cisco Unified IP
Phone can also be supplied power by a Cisco Unified IP Phone power injector. The Cisco Unified IP
Phone power injector connects Cisco Unified IP Phones to Cisco switches that do not support inline
power or to non-Cisco switches. The Cisco Unified IP Phone power injector is compatible with most
Cisco Unified IP Phones. It has two 10/100/1000 Base-T Ethernet ports. One Ethernet port connects to
the switch access port and the other connects to the Cisco Unified IP Phone.
Quality of Service
When configuring network-level quality of service (QoS), Cisco desk phones such as the Cisco Unified
IP Phone 7900, 8900, 9900, and DX Series can be trusted and their packet markings honored. By default
these endpoints mark their voice media and signaling packets with appropriate Layer 3 values (voice
media as DSCP 46 or PHB EF; call signaling as DSCP 24 or PHB CS3), which match Cisco QoS
recommendations for appropriate voice media and signaling marking, to ensure end-to-end voice quality
on a converged network. While many Cisco desk phones support the attachment of a desktop computer,
Cisco desk phones are capable of separating the voice and data traffic, placing voice traffic onto the voice
VLAN and data traffic from the desktop onto the data VLAN. This enables the network to extend trust
to the phone but not to the PC port of the phone. However, for multipurpose devices such as the Cisco
DX Series endpoints, which are capable of generating both voice and data traffic without an attached
desktop computer, both voice and data traffic will traverse the same VLAN. In these cases, whether the
device is attached to the voice or data VLAN, extending trust to these devices might not be advisable.
Instead, re-marking the traffic based on port and protocol will ensure that all traffic is appropriately
marked regardless of the VLAN it traverses.
In deployment where there are concerns about the potential volume of data traffic generated by
multipurpose devices such as the Cisco DX Series and the possibility of adversely impacting real-time
voice and video traffic, these devices should be deployed in the data VLAN or in a separate VLAN. This
will alleviate concerns about impacting call quality of voice and video-only devices. Further, with packet
re-marking based on ports and protocols, priority treatment can still be provided within the VLAN to
real-time traffic generated by these multipurpose devices.
Note
While many Cisco desk phones support Link Layer Discovery Protocol for Media Endpoint Devices
(LLDP-MED), they do so only for VLAN and Power over Ethernet negotiation. Cisco Unified IP Phones
do not honor DSCP and CoS markings provided by LLDP-MED.
8-13
Chapter 8
Collaboration Endpoints
Desk Phones
Intelligent Proximity
Intelligent Proximity refers to features that leverage proximity-based connections between Cisco
hardware endpoints and mobile devices.
Intelligent Proximity for Mobile Voice capabilities available on the Cisco DX Series and select
8800 Series endpoints rely on the use of Bluetooth pairing between the DX or 8800 endpoint and a
cellular or smart phone.
Bluetooth paired mobile devices are able to invoke two features:
Hands-free audio, providing the ability to send audio of a cellular terminated call through the
DX Series, 8845, 8851, 8861, or 8865 IP endpoint speaker or handset. Audio play-out of the cellular
terminated call can be moved back and forth between the DX, 8845, 8851, 8861, or 8865 and the
mobile device. In addition, because the Bluetooth paired mobile device appears on the 8845, 8851,
8865, or DX Series endpoints as another line, cellular calls on the Bluetooth paired mobile device
can be initiated using the DX or 8800 IP endpoint.
Mobile contact and call history sharing, providing the ability to share mobile device contact and call
history sharing with the DX Series, 8845, 8851, 8861, or 8865 endpoints.
Because Intelligent Proximity for Mobile Voice relies on Bluetooth pairing, there is no requirement to
run an application or client on the mobile device. All communication and interaction occurs over the
standard-based Bluetooth interfaces.
8-14
Chapter 8
Collaboration Endpoints
Video Endpoints
The Intelligent Proximity for Mobile Voice feature set on the DX Series endpoints and the 8845, 8851,
8861, and 8865 IP phones is compatible with the Unified Mobility feature set, including single number
reach (SNR), remote destination and desk phone pickup, two-stage enterprise dialing, and mobile
voicemail avoidance. In the case of the 8845, 8851, 8861, and 8865 IP phones, Intelligent Proximity for
Mobile Voice is compatible with Cisco Jabber mobile clients. When a Jabber client running on a mobile
device is paired with the 8845, 8851, 8861, or 8865 IP phone, the audio portion of a Jabber call may be
played out using the 8845, 8851, 8861, or 8865 handset or speaker while the video portion of the call
continues to play on the Jabber mobile client. In the case of DX Series endpoints, Intelligent Proximity
for Mobile Voice functionality is limited exclusively to the cellular line of Bluetooth paired mobile
devices running Jabber.
The Intelligent Proximity for Mobile Voice feature set on the DX and select 8800 Series endpoints
requires firmware version 10.1.1 or later.
For more information about Intelligent Proximity for Mobile Voice, refer to
http://www.cisco.com/go/proximity and the product documentation for the DX and 8800 Series
endpoints.
Video Endpoints
Cisco video endpoints provide IP video telephony features and functions similar to IP voice telephony,
enabling users to make point to point and point to multi-point video calls. Cisco offers the following
video-capable endpoints:
Cisco Jabber software-based desktop clients such as Cisco Jabber for Windows
Cisco Unified IP Phone 9900 Series with the optional USB camera attachment.
Cisco Unified IP Phones 8800 Series (8845 or 8865) with built-in camera
Cisco TelePresence System 500, 1100, C, EX, MX, SX, and TX Series
Cisco video endpoints deliver high-quality video for all user types and environments within any
organization. Cisco video endpoints are classified into families based on the features they support,
hardware screen size, and environment where the endpoint is deployed. This section categorizes the
Cisco video endpoint families into personal, multipurpose, and immersive endpoints groups.
8-15
Chapter 8
Collaboration Endpoints
Video Endpoints
For additional information on Cisco Jabber desktop clients, refer to Software-Based Endpoints,
page 8-24.
For more information about the video capabilities of Cisco Jabber for Windows, refer to the data sheet
and product documentation available at
http://www.cisco.com/c/en/us/products/unified-communications/jabber-windows/index.html
Cisco DX Series
The Cisco DX Series endpoints are capable of transmitting video by means of the built-in front-facing
camera. These endpoints are capable of receiving and displaying video natively on their screens with a
variety of video resolutions and frame rates. The video capabilities of these phones can be enabled and
disabled or tuned as desired from the Cisco call control platform configuration pages.
These devices register and communicate with Unified CM using SIP signaling protocol.
For more information about the Cisco DX Series video capabilities, refer to the data sheets and product
documentation available at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/desktop-collaboration-experiencedx600-series/index.html
8-16
Chapter 8
Collaboration Endpoints
Video Endpoints
8-17
Chapter 8
Collaboration Endpoints
Video Endpoints
SX20 is a codec with one of three camera options, and it supports the multisite features, providing
the ability to add up to three more participants in a Cisco TelePresence call.
SX80 is a codec that includes integrator packages supporting different camera and touch panel
options.
8-18
Chapter 8
Collaboration Endpoints
Video Endpoints
C90 supports five simultaneous video inputs and is ideal for TelePresence and collaboration studios
and boardrooms
C60 supports three simultaneous video inputs and is ideal for TelePresence and collaboration studios
and boardrooms
All of these models support the multisite feature that provides the ability to add up to three more
participants into a Cisco TelePresence call.
For more information about the Cisco TelePresence System Integrator C Series video endpoints, refer to
the data sheets and product documentation available at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/telepresence-integrator-c-series/in
dex.html
8-19
Chapter 8
Collaboration Endpoints
Video Endpoints
Quality of Service
When configuring network-level quality of service (QoS), Cisco video endpoints (including Cisco
Unified IP Phone 8900 and 9900 Series, Cisco DX Series, and Cisco TelePresence System devices)
generally mark traffic at Layer 3 according to Cisco general QoS guidelines related to voice and video
packet marking (voice media as DSCP 46 or PHB EF; desktop video media as DSCP 34 or PHB AF41;
telepresence video media as DSCP 32 or PHB CS4; call signaling as DSCP 24 or PHB CS3), and
therefore these devices can be trusted. In the case of personal desktop video endpoints, including the
Cisco Unified IP Phone 8900 and 9900 Series and Cisco DX Series devices, both voice and video media
packets are marked as DSCP 34 or PHB AF41 to preserve lip synchronization during a video call.
While proper network QoS configuration is essential even when the endpoint marking is trusted, Cisco
recommends ensuring that sufficient bandwidth is provisioned on the network and then using
network-based policing and rate limiting to ensure that all endpoints do not consume more network
bandwidth than they should. Software-based video-capable endpoints do present challenges when they
do not or cannot mark traffic appropriately. In these situations, typical guidance is to re-mark media and
signaling traffic within the network from best-effort to appropriate and recommended values (voice
media as DSCP 46 or PHB EF; desktop video and voice media for video calls as DSCP 34 or PHB AF41;
telepresence video media as DSCP 32 or PHB CS4; call signaling as DSCP 24 or PHB CS3) based on
protocols and/or port numbers.
In the case of software-based Cisco Jabber for Windows, appropriate Layer 3 DSCP QoS marking can
be applied to audio and video streams based on voice and video media source port numbers using
Microsoft Windows group policies.
For more information about Cisco Jabber for Windows QoS with Microsoft Windows group policies,
refer to the Quality of Service configuration information in the latest version of the Deployment and
Installation Guide for Cisco Jabber, available at
http://www.cisco.com/c/en/us/support/unified-communications/jabber-windows/products-installati
on-guides-list.html
Note
While some Cisco video-capable endpoints support Link Layer Discovery Protocol for Media Endpoint
Devices (LLDP-MED), they do so only for VLAN and Power over Ethernet negotiation. Cisco video
endpoints do not honor DSCP and CoS markings provided by LLDP-MED.
8-20
Chapter 8
Collaboration Endpoints
Video Endpoints
For more information on video endpoint network bandwidth consumption and QoS marking and
classification, see the section on WAN Quality of Service (QoS), page 3-36.
Inter-VLAN Routing
When deploying video endpoints on networks with voice and data VLAN separation, it is important to
consider software-based video-capable endpoints as well as hardware-based video endpoints that need
to access resources. Because software-based endpoints running on a desktop computer are primarily
attached to the data VLAN, inter-VLAN routing should be configured and allowed so that voice traffic
from these endpoints on the data VLAN can reach endpoints on the voice VLAN. Likewise, if
hardware-based video endpoints such as the Cisco TelePresence System endpoints need access to
network resources such as directory or management services deployed on the data VLAN, inter-VLAN
routing must be allowed.
8-21
Chapter 8
Collaboration Endpoints
Video Endpoints
Intelligent Proximity
As previously mentioned, Intelligent Proximity refers to features that leverage proximity-based
connections between Cisco hardware endpoints and mobile devices.
Intelligent Proximity for Mobile Voice capabilities available on the Cisco DX Series rely on Bluetooth
pairing between the DX endpoint and a cellular or smart phone, enabling hands-free audio and mobile
contact and call history sharing.
For more information on Intelligent Proximity and Bluetooth pairing, see Intelligent Proximity,
page 8-14.
Video Interoperability
Video interoperability is the audio and video support for point-to-point calls between Cisco
TelePresence System video endpoints, other Cisco Collaboration video endpoints, and third-party video
endpoints. Previously, video interoperability between different families of video endpoints was possible
only with the insertion of a video component between endpoints, such as a video transcoder or a
multipoint control unit (MCU).
Cisco Unified CM not only offers native video interoperability between different video endpoint family
types, but also provides better video interoperability in general with H.264 codec negotiation in SIP and
H.323 protocols and enable the endpoints to negotiate high definition (HD) resolutions when available.
Video interoperability, however, is dependent on the endpoints to support the interoperation.
Video interoperability in Unified CM also enables Cisco TelePresence System video endpoints to
communicate with non-video endpoints, provided that the installed firmware supports such
interoperability. For further information, refer to the Cisco TelePresence Interoperability Database,
available at
https://tp-tools-web01.cisco.com/start/
Additionally, Cisco Unified CM provides support for enhanced interoperability with call agents other
than Unified CM. Through scripting, Unified CM supports the following features:
SIP transparency The ability to pass through known and unknown message components
SIP normalization Transformations on inbound and outbound SIP messages and content bodies
The primary motivation for video interoperability support is to facilitate the interaction of a diverse set
of video endpoints without the need for deploying an expensive hardware-based DSP infrastructure that
would otherwise be required. There are additional benefits that can be derived from the use of advanced
conferencing and transcoding resources (for example, active presence where participants of multi-point
conferences can see the active speaker); however, the desired feature set and video calling needs will
dictate when and where those advanced resources would be required.
The following sections present general considerations and recommendations for the use of video
interoperability:
8-22
Chapter 8
Collaboration Endpoints
Video Endpoints
Two different video endpoint family types (Cisco TelePresence System video endpoints, other Cisco
Collaboration video endpoints such as the Cisco DX80, or third-party endpoints) engaged in a video
call
The following sections offer further information about the scope of the video interoperability support:
In most cases a video endpoint that supports SIP or H.323 without using proprietary signaling would be
able to interoperate with a Cisco Collaboration video endpoint that supports video interoperability. For
specific information on the scope of the interoperability between common sets of deployed devices and
general information about the testing that was conducted to validate these more common examples of
interoperability, refer to the Cisco Collaboration Systems documentation available at
http://www.cisco.com/c/en/us/td/docs/voice_ip_comm/uc_system/unified/communications/system/
ucstart.html
Limitations of Video Interoperability
While video interoperability support attempts to enable any-to-any point-to-point video call
interoperability, it is important to note that not all features of an individual video endpoint can be
supported when interoperating with another endpoint. There are many reasons for this. For example,
incompatibilities between different call control protocols could render a feature unavailable or offer a
different representation of that feature. H.264 video media parameters can be represented differently in
H.323 than in SIP, as another example. H.323 also does not have support for presence, but presence is
quite commonly supported in SIP. Skinny Client Control Protocol (SCCP) does not have any notion of
application sharing, which is commonly available in SIP and H.323 endpoint implementations. For
instance, an SCCP user trying to share his/her PC screen would be hampered because Binary Floor
Control Protocol (BFCP) and H.239 are not available with SCCP.
Quality of Service (QoS) and Call Admission Control Considerations for Video Interoperability,
page 8-24
8-23
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
The following guidelines and restrictions apply with regard to video interoperability in a Unified CM
deployment:
If H.323 or SCCP protocols are used in conjunction with video interoperability, Unified CM will
support only a single H.264 payload and the packetization mode is treated as 0. An example side
effect (but not the only one) of this circumstance is the fact that 1080p resolution is not available
with these protocols because 1080p requires packetization mode 1.
If a SIP endpoint omits the level-asymmetry-allowed parameter in the Session Description Protocol
(SDP), Cisco products will assume that the endpoint can support asymmetric resolution
transmission. Therefore, different receiving and sending video resolutions could be negotiated
during a call.
Unified CM will not negotiate Real-Time Transport Control Protocol (RTCP) feedback if the video
call invokes a media termination point (MTP) or transcoder.
Quality of Service (QoS) and Call Admission Control Considerations for Video Interoperability
There are no changes to the configuration of regions and locations in Unified CM as a result of video
interoperability support. However, regions play a significant role in determining the resolution between
groups of endpoints, and they can be used to maximize or minimize the resolution that these devices use
when interoperating. The Max Video Call Bit Rate field in the regions settings is used to determine the
amount of bandwidth and, thus, the resolution that endpoints are able to negotiate.
For further information about QoS and call admission control with native video interoperability, see the
section on Call Admission Control Design Recommendations for Video Deployments, page 13-79.
Software-Based Endpoints
A software-based endpoint is an application installed on a client desktop computer that registers and
communicates with Cisco call processing platforms for voice and video services. In addition, these
endpoint software client applications may provide collaboration features and services such as
messaging, presence, directory access, and conferencing. Software-based endpoint desktop client
applications include Cisco IP Communicator and Cisco Jabber.
Cisco IP Communicator
Cisco IP Communicator is a Microsoft Windows-based application that provides enterprise IP phone
functionality to desktop computers. This application provides enterprise-class IP voice calling for
remote users, telecommuters, and other mobile users. Cisco IP Communicator supports both SCCP and
SIP signaling protocols for registering and communicating with Cisco call processing platforms. For
more information about Cisco IP Communicator, refer to the data sheets and product documentation at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/ip-communicator/index.html
8-24
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
SIP-based call control for voice and video softphone clients from Unified CM
Deskphone call control and "Click to Call" services from the Unified CM CTI interface
Instant messaging and presence services using XMPP, from either the Cisco IM and Presence
Service or Cisco WebEx Messenger service. Cisco WebEx Meeting Center also offers hosted
collaboration services such as online meetings and events.
Desktop sharing using either video desktop sharing (BFCP) or WebEx desktop sharing
Visual voicemail services from Cisco Unity Connection using Internet Message Access Protocol
(IMAP) or Representational State Transfer (REST)
Contact management relying on Unified CM User Data Service (UDS), Microsoft Active Directory,
or other supported LDAP directories; or in the case of cloud-based integrations, the WebEx
Messenger service
Microsoft Outlook Integration, which provides user availability status and messaging capabilities
directly through the user interface of Microsoft Office applications such as Microsoft Outlook
The ability to communicate and abstract services and APIs, as shown in Figure 8-2, allows the Jabber
Desktop Client to coordinate the management of protocols to these services and APIs, handle event
notifications, and control the low-level connection logic for local system resources. Depending on the
deployment type, some features might not be supported.
8-25
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
Figure 8-2
Voice and
Video Media
IMAP
HTTP
RTP
REST
Call Control
SIP, TFTP,
CTI, HTTP
RTP
Contact
Search
IM and
Presence
REST
LDAP
SSL
AES
HTTPS
XMPP
SSL
AES
Audio, Video,
Web Conferencing
HTTPS
RTP
SIP
TFTP
HTTP
CTI Mgr
Unied CM
Unity
Connecon
LDAP
Directory
IM and
Presence
348741
CCMCIP
The choice between Cisco IM and Presence or WebEx Messenger service for instant messaging and
presence services can depend on a number of factors. WebEx Messenger service deployments use a
cloud-based service that is accessible from the Internet. On-premises deployments based on Cisco IM
and Presence provide the administrator with direct control over their IM and presence platform and also
allow presence federation using SIP/SIMPLE to other presence services.
For information on the full set of features supported by each IM and Presence platform, refer to the
following documentation:
8-26
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
Deskphone Control Mode Using a Cisco IP Phone for audio (and video, if supported)
When a Jabber Desktop Client is in deskphone control mode, it does not register with Unified CM
using SIP, but instead it uses CTI/JTAPI to initiate, monitor, and terminate calls, monitor line state,
and provide call history, while controlling a Cisco Unified IP Phone. The Cisco CallManager Cisco
IP Phone (CCMCIP) or UDS service on Unified CM is used by the Jabber Desktop Client to retrieve
a list of devices associated with each user. This list of devices is used by a client in deskphone mode
to choose which Cisco IP Phone it wishes to control.
Softphone Mode
When operating in softphone mode, the Jabber Desktop Client is a SIP line-side registered device on
Unified CM, utilizing all the call control capabilities and functionality of a Cisco Unified IP Phone,
including configuration of registration, redundancy, regions, locations, dial plan management,
authentication, encryption, user association, and so forth. The Jabber Desktop Client supports a single
line appearance for the user.
The SIP registered device of the Jabber Desktop Client must be factored in as a regular SIP endpoint,
like any other SIP registered endpoint, for purposes of sizing calculations for a Unified CM cluster. The
Jabber Desktop Client in softphone mode uses the CCMCIP or UDS service to discover its device name
for registration with Unified CM.
Deskphone Control Mode
When operating in deskphone control mode, the Jabber Desktop Client uses CTI/JTAPI to provide the
ability to place, monitor, and receive calls using Cisco Unified IP Phones. When audio calls are received
or placed in this mode, the audio path is through the Cisco Unified IP Phone. For video calls, the video
stream can originate and terminate either on the Cisco IP Phone (if it has a camera) or on the computer
using a approved camera. The Jabber Desktop Client uses the CCMCIP or UDS service on Unified CM
to discover the associated devices of the user.
When using deskphone control mode for the Jabber Desktop Client, factor the CTI scaling numbers into
the Unified CM deployment calculations. For additional information about capacity planning, see the
chapter on Collaboration Solution Sizing Guidance, page 25-1.
8-27
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
Cisco Unified CM User database via the User Data Service (UDS)
Contacts can also be stored and retrieved locally using either of the following:
The Jabber Desktop Client uses reverse number lookup to map an incoming telephone number to a
contact, in addition to photo retrieval. The Jabber Desktop Client contact management allows for up to
five search bases to be defined for LDAP queries.
Cisco Unified CM User Data Service (UDS)
UDS provides clients with a contact search service on Cisco Unified Communications Manager. You can
synchronize contact data into the Cisco Unified CM User database from Microsoft Active Directory or
other LDAP directory sources. Clients can then automatically retrieve that contact data directly from
Unified CM using the UDS REST interface.
LDAP Directory
You can configure a corporate LDAP directory to satisfy a number of different requirements, including
the following:
User provisioning You can provision users automatically from the LDAP directory into the Cisco
Unified Communications Manager database using directory integration. Cisco Unified CM
synchronizes with the LDAP directory content so that you avoid having to add, remove, or modify
user information manually each time a change occurs in the LDAP directory.
User authentication You can authenticate users using the LDAP directory credentials. Cisco IM
and Presence synchronizes all the user information from Cisco Unified Communications Manager
to provide authentication for client users.
User lookup You can enable LDAP directory lookups to allow Cisco clients or third-party XMPP
clients to search for contacts in the LDAP directory.
Use the WebEx Administration Tool to implement WebEx Directory Integration. WebEx imports a
comma-separated value (CSV) file of your enterprise directory information into its WebEx Messenger
service. For more information, refer to the documentation at
http://www.webex.com/webexconnect/orgadmin/help/index.htm?toc.htm?17444.htm
Jabber Desktop Client Cache
The Jabber Desktop Client maintains a local cache of contact information derived from previous
directory queries and contacts already listed, as well as the local address book or contact list. If a contact
for a call already exists in the cache, the Jabber Desktop Client does not search the directory. If a contact
does not exist in the cache, the Jabber Desktop Client performs a directory search.
8-28
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
Directory Search
When a contact cannot be found in the local Jabber Desktop Client cache or contact list, a search for
contacts can be made. The WebEx Messenger user can utilize a predictive search whereby the cache,
contact list, and local Outlook contact list are queried as the contact name is being entered. If no matches
are found, the search continues to query the corporate directory (WebEx Messenger database).
For more information about Cisco Jabber for Windows, refer to the data sheets and product
documentation at
http://www.cisco.com/c/en/us/products/unified-communications/jabber-windows/index.html
For more information about the Cisco Jabber for Mac, refer to the data sheets and product documentation
at
http://www.cisco.com/c/en/us/products/unified-communications/jabber-mac/index.html
8-29
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
Figure 8-3
Microso Outlook
IMAP
HTTP
RTP
RTP
SRTP
Call Control
SIP, TFTP,
CTI, HTTP
Contact Search
IM and Presence
Audio, Video,
Web Conferencing
HTTPS
RTP
LDAP
TFTP
HTTP
MS-SIP
HTTP
CTI Mgr
Unity
Connecon
Unied CM
Microso
Acve
Directory
Microso
Front-End
Server
348745
CCMCIP
With a deployment of Cisco UC IntegrationTM for Microsoft Lync, the client utilizes user information
from the Office Communications Server Address Book that gets downloaded to the client. The address
book is generated and delivered to the clients from the Office Communications Server once the user is
enabled for presence and instant messaging. Cisco recommends that administrators populate the user
directory number information with an E.164 value (for example, +18005551212) and enable LDAP
synchronization and authentication on Unified CM for user account consistency. Cisco UC IntegrationTM
for Microsoft Lync connects to both Cisco Unified CM and Microsoft Active Directory and provides for
account credential synchronization rules.
Note
With Cisco UC IntegrationTM for Microsoft Lync, instant messaging and presence services are provided
by Microsoft rather than by Cisco Unified Communications services.
8-30
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
Microsoft Active Directory group policies can be extended using administration templates, and Cisco
UC IntegrationTM for Microsoft Lync provides a template that the administrator can add to provide the
group policy support. After the administrative template is loaded, a Cisco UC IntegrationTM for
Microsoft Lync configuration policy can be created by the administrator for the registry configuration
settings (TFTP servers, CTI servers, CCMCIP servers, voicemail, and LDAP servers).
The Group Policy Management Console can be used to control how and where these group policies are
applied to different organizational units. From a client policy perspective, when you deploy Cisco UC
IntegrationTM for Microsoft Lync, Cisco recommends setting the Microsoft Telephony Mode Policy to
IM and Presence Only and DisableAVConferencing. These client policy changes will allow for only
a single set of call options to be displayed in the Microsoft Lync user experience.
A Cisco UC IntegrationTM for Microsoft Lync deployment also allows for custom presence states to be
defined and deployed in the cisco-presence-states-config.xml file that gets installed. However, Cisco
recommends that administrators relocate this file to an HTTPs location, such as the Microsoft Office
Communications Server, to allow Microsoft Lync to use this custom presence state file based on the
following registry location:
HKLM\Software\Policies\Microsoft\Communicator\CustomStateURL
Quality of Service
Cisco software-based client applications do mark their traffic at Layer 3 in accordance with QoS
marking best practices; however, even when the applications do mark traffic properly, the underlying
operating system or hardware might not honor the markings. Given the general unpredictability and
unreliability of traffic marking coming from desktop computers, as a general rule these traffic markings
should not be trusted. This means that all traffic flows must be re-marked by the network based on
protocol and/or port numbers, with real-time traffic flows being marked based on best practices. This
includes re-marking of voice-only call media with DSC 46 or PHB EF, video call media (including
voice) with DSCP 34 or PHB AF41, and call signaling with DSCP 24 or PHB CS3. These markings
along with a properly configured network infrastructure ensure priority treatment for voice-only call
media and dedicated bandwidth for video call media and call signaling. In addition to re-marking of
software-based endpoint traffic, Cisco recommends using network-based policing and rate limiting to
ensure that the software-based endpoint does not consume too much network bandwidth. This can occur
when the desktop computer generates too much data traffic or when the endpoint application misbehaves
and generates more voice and/or video media and signaling traffic than would be expected for a typical
call. In cases where third-party software is used to fully control desktop computer network traffic
marking, administrators may decide to trust desktop computer marking, in which case re-marking of
packets would not be required. Network-based policing and rate limiting is still recommended to protect
the overall network in case of a misbehaving endpoint.
In the case of software-based Cisco IP Communicator and Cisco Jabber for Windows, appropriate
Layer 3 DSCP QoS marking can be applied to audio and video streams based on voice and video media
source port numbers using Microsoft Windows group policies.
For more information about Cisco Jabber for Windows QoS with Microsoft Windows group policies,
refer to the Quality of Service configuration information in the latest version of the Deployment and
Installation Guide for Cisco Jabber, available at
http://www.cisco.com/c/en/us/support/unified-communications/jabber-windows/products-installati
on-guides-list.html
8-31
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
Inter-VLAN Routing
Because software-based endpoints run on a desktop computer usually deployed on a data VLAN, when
software-based endpoints are deployed on networks with voice and data VLAN separation, inter-VLAN
routing should be configured and allowed so that voice traffic from these endpoints on the data VLAN
can reach endpoints on the voice VLAN.
Dial Plan
Dial plan and number normalization considerations must be taken into account when deploying
software-based endpoints. Jabber desktop clients typically use the directory for searching, resolving, and
adding contacts. The number that is associated with those contacts must be in a form that the client can
recognize, resolve, and dial.
Deployments may vary, depending on the configuration of the directory and Unified CM. In cases where
the directory contains E.164 numbering (for example, +18005551212) for business, mobile, and home
telephone numbers and Unified CM also contains an E.164 dial plan, the need for additional dial rules
is minimized because every lookup, resolution, and dialed event results in an E.164 formatted dial string.
If a Unified CM deployment has implemented a private dial plan (for example, 5551212), then
translation of the E.164 number to a private directory number needs to occur on Unified CM. Outbound
calls can be translated by Unified CM translation patterns that allow the number being dialed (for
8-32
Chapter 8
Collaboration Endpoints
Software-Based Endpoints
example, +18005551212) to be presented to the endpoint as the private number (5551212 in this
example). Inbound calls can be translated by means of directory lookup rules. This allows an incoming
number of 5551212 to be presented for reverse number lookup caller identification as +18005551212.
Private numbering plan deployments may arise, where the dial plan used for your company and the
telephone number information stored in the LDAP directory may require the configuration of translation
patterns and directory lookup rules in Cisco Unified Communications Manager to manage number
format differences. Directory lookup rules define how to reformat the inbound call ID to be used as a
directory lookup key. Translation patterns define how to transform a phone number retrieved from the
LDAP directory for outbound dialing.
Translation patterns are used by Unified CM to manipulate the dialed digits before a call is routed, and
they are handled strictly by Unified CM. Translation patterns are the recommended method for
manipulating dialed numbers. For additional guidelines on translation pattern usage and dial plan
management, see the chapter on Dial Plan, page 14-1.
Application dialing rules can be used as an alternative to translation patterns to manipulate numbers that
are dialed. Application dialing rules can automatically strip numbers from, or add numbers to, phone
numbers that the user dials. Application Dial Rules are configured in Unified CM and are downloaded
to the client from Unified CM. Translation patterns are the recommended method for manipulating
dialed numbers.
Directory lookup rules transform caller identification numbers into numbers that can be looked up in the
directory. A directory lookup rule specifies which numbers to transform based on the initial digits and
the length of the number. Directory lookup rules are configured in Unified CM and are downloaded to
the client from Unified CM.
Before a call is placed through contact information, the client application removes everything from the
phone number to be dialed, except for letters and digits. The application transforms the letters to digits
and applies the dialing rules. The letter-to-digit mapping is locale-specific and corresponds to the letters
found on a standard telephone keypad for that locale. For example, for a US English locale,
1-800-4UCSRND transforms to 18004827763. Users cannot view or modify the client transformed
numbers before the application places the call.
Contact Sources
Cisco Jabber for Windows defaults to using Enhanced Directory Integration (EDI), which uses
preconfigured directory attribute mappings for integration with Microsoft Active Directory.
Cisco Jabber for Mac defaults to using Basic Directory Integration (BDI), which relies on integration to
LDAP v3 compatible directories, including Microsoft Active Directory.
For integration with an LDAP directory that requires custom attribute mapping, these attribute mappings
can be created in a configuration file that can be downloaded to the client from the Unified CM server.
Cisco Jabber desktop clients also support the Unified CM User Data Service (UDS), which allows a
client to search for contacts using the Unified CM user database (which may be synchronized with an
LDAP directory). UDS is used automatically by Jabber desktop clients for contact resolution when they
are located outside of the corporate firewall and connected via Expressway mobile and remote access.
UDS should be leveraged as a contact source only in cases where Jabber desktop clients are connected
via Expressway or for very small deployments. Otherwise, UDS is not recommended as a contact source
due to resource utilization on Unified CM. When clients leverage UDS as a contact source, the
registration capacity of the Unified CM node is reduced by half.
In addition, Jabber for Windows supports Microsoft Outlook local contact, which allows users to search
for contacts that are in the user's Microsoft Outlook client.
8-33
Chapter 8
Collaboration Endpoints
Wireless Endpoints
Wireless Endpoints
Cisco wireless endpoints rely on an 802.11 wireless LAN (WLAN) infrastructure for network
connectivity and to provide IP telephony functionality and features. This type of endpoint is ideal for
mobile users that move around within a single enterprise location or between enterprise locations or
environments where traditional wired phones are undesirable or problematic. Cisco offers the following
voice and video over WLAN (VVoWLAN) IP phones:
Cisco Unified Wireless IP Phones, including the Cisco Unified Wireless IP Phone 7925G,
7925G-EX, and 7926G
Cisco DX Series
All are hardware-based phones with built-in radio antenna. The Cisco Unified Wireless IP Phones as
well as the wirelessly attached Cisco Unified IP Phone 9971 enable 802.11b, 802.11g, or 802.11a
connectivity to the network. The Cisco IP Phone 8861 enables 802.11a, 802.11b, 802.11g, 802.11n, and
802.11ac wireless connectivity while the Cisco DX Series endpoints enable 802.11a, 802.11b, 802.11g,
and 802.11n wireless connectivity. The Cisco Unified Wireless IP Phones register and communicate with
Cisco call processing platforms using SCCP signaling protocol, while the Cisco IP Phone 8861, Unified
IP Phone 9971, and DX Series endpoints use the SIP signaling protocol to register and communicate
with Cisco call processing platforms.
For more information about the Cisco Unified Wireless IP Phones, refer to the data sheets and product
documentation available at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phone-7900-series/inde
x.html
For more information about the Cisco IP Phone 8861, refer to the data sheets and product documentation
available at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phone-8800-series/inde
x.html
For more information about the Cisco Unified IP Phone 9971, refer to the data sheets and product
documentation available at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/unified-ip-phones-9900-series/ind
ex.html
For more information about the Cisco DX Series endpoints, refer to the data sheets and product
documentation available at
http://www.cisco.com/c/en/us/products/collaboration-endpoints/desktop-collaboration-experiencedx600-series/index.html
8-34
Chapter 8
Collaboration Endpoints
Wireless Endpoints
Note
In dual-band WLANs (those with both 2.4 GHz and 5 GHz bands), it is possible to roam between
802.11b/g and 802.11a with the same SSID, provided the client is capable of supporting both bands.
However, with some devices this can cause gaps in the voice or video path. In order to avoid these gaps,
use only one band for voice and video communications.
8-35
Chapter 8
Collaboration Endpoints
Wireless Endpoints
minimized by deploying sufficient numbers of WLAN access points (APs) to handle required call
capacities. AP call capacities are based on the number of simultaneous bidirectional streams that can be
supported in a single channel cell area. The general rule for VVoWLAN call capacities is as follows:
Maximum of 27 simultaneous VoWLAN bidirectional streams per 802.11g/n (2.4 GHz) channel cell
with Bluetooth disabled or per 802.11a/n/ac (5 GHz) channel and 24 Mbps or higher data rates
enabled.
Maximum of 8 simultaneous VVoWLAN bidirectional streams per 802.11 g/n (2.4 GHz) channel
cell with Bluetooth disabled or per 802.11 a/n/ac (5 GHz) channel cell assuming a video resolution
of 720p (high-definition) and video bit rate of up to 1 Mbps.
These call capacity values are highly dependent upon the RF environment, the wireless handset features,
and underlying WLAN system features. Actual capacities for a particular deployment could be less.
Note
A single call between two wireless endpoints associated to the same AP is considered to be two
simultaneous bidirectional streams.
The above capacities are based on voice activity detection (VAD) being disabled and a packetization
sample size of 20 milliseconds (ms). VAD is a mechanism for conserving bandwidth by not sending RTP
packets while no speech is occurring during the call. However, enabling or disabling VAD, also referred
to as Silence Suppression, is sometimes a global configuration depending on the Cisco call control
platforms. Thus, if VAD is enabled for wirelessly attached Cisco Unified IP Phones, then it may be
enabled for all devices in the deployment. Cisco recommends leaving VAD (Silence Suppression)
disabled to provide better overall voice quality.
At a sampling rate of 20 ms, a voice call will generate 50 packets per second (pps) in either direction.
Cisco recommends setting the sample rate to 20 ms for almost all cases. By using a larger sample size
(for example, 30 or 40 ms), you can increase the number of simultaneous calls per AP, but a larger
end-to-end delay will result. In addition, the percentage of acceptable voice packet loss within a wireless
environment decreases dramatically with a larger sample size because more of the conversation is
missing when a packet is lost. For more information about voice sampling size, see the section on
Bandwidth Provisioning, page 3-52.
Bluetooth Support
The Cisco Unified Wireless IP Phones 7925G, 7925G-EX, and 7926G, the Cisco Unified IP Phone 9971,
the Cisco IP Phone 8861, and the Cisco DX Series endpoints are Bluetooth-enabled devices. The
Bluetooth radio or module within these wireless Cisco IP phones enables support for Bluetooth headsets.
In addition, as previously mentioned, the Cisco IP Phone 8851, 8861, and DX Series endpoints support
Intelligent Proximity for Mobile Voice with Bluetooth pairing for hands-free audio and mobile contact
and call history sharing. Because Bluetooth devices use the same 2.4 GHz radio band as 802.11b/g
devices, it is possible that Bluetooth and 802.11b/g-capable devices can interfere with each other, thus
resulting in connectivity issues.
While the Bluetooth and 802.11 WLAN radios coexist natively in the Cisco Unified Wireless IP Phones,
Cisco Unified IP Phone 9971, Cisco IP Phone 8861, and Cisco DX Series endpoints, greatly reducing
and avoiding radio interference between the Bluetooth and 802.11b/g radio, the Bluetooth radio in these
wirelessly attached phones can cause interference for other 802.11b/g and Bluetooth radio devices
deployed in close proximity. Due to the potential for interference and disruption of 802.11b/g WLAN
voice and video devices (which can result in poor voice and video quality, de-registration, and/or call
setup delays), Cisco recommends deploying all WLAN voice and video devices on 802.11a, 802.11n, or
802.11ac, which use the 5 GHz radio band. By deploying wireless phones on the 5 GHz radio band, you
can avoid interference caused by Bluetooth devices.
8-36
Chapter 8
Collaboration Endpoints
Wireless Endpoints
If deploying 802.11 WLAN devices on the 5 GHz radio band is not an option and interference on the
2.4 GHz radio band is causing connectivity, functionality, or voice and video quality issues, consider
prohibiting or restricting the use of Bluetooth headsets and Bluetooth dependent features such as
Intelligent Proximity for Mobile Voice within these deployments.
For more information on Intelligent Proximity for Mobile Voice and Bluetooth pairing on the Cisco
8851, 8861, and DX Series endpoints, see Intelligent Proximity, page 8-14.
Note
Using Bluetooth wireless headsets with the battery-powered Cisco Unified Wireless IP Phones will
increase battery power consumption on your phone and will result in reduced battery life.
Note
The use of Bluetooth headsets and Bluetooth features such as Intelligent Proximity for Mobile Voice can
cause interference and possibly service disruption for adjacent wireless clients and endpoints relying on
the 2.4 GHz band (802.11b/g/n).
Quality of Service
When configuring network-level quality of service (QoS), Cisco wireless endpoints (including Cisco
Unified Wireless IP Phones, the Cisco Unified IP Phone 9971, the Cisco IP Phone 8861, and Cisco
DX Series endpoints) can be trusted and their packet markings honored. By default these endpoints mark
the recommended and appropriate Layer 3 values for voice and video media and call signaling (voice
media as DSCP 46 or PHB EF; voice and video media as DSCP 34 or PHB AF41 for a video call, and
call signaling as DSCP 24 or PHB CS3). Likewise, these devices mark appropriately at Layer 2 (voice
media WMM User Priority (UP) of 6; voice and video media for video call WMM UP 5; call signaling
WMM UP 4). With these packet markings, end-to-end voice quality on the converged network will be
acceptable.
Despite appropriate packet marking at both Layer 2 and Layer 3, multipurpose devices such as the
Cisco DX80 are capable of generating large amounts of non-real-time traffic. As such, concerns are
sometimes raised regarding commingling of these devices on the same WLAN SSID or VLAN. While
Layer 2 QoS marking and 802.11e WMM work to ensure that more bandwidth and more frequent access
to the wireless medium are provided for real-time traffic, in dense or heavily utilized deployments,
separating multi-purpose devices such as DX Series endpoints into a separate SSID may provide some
relief. However, this separate SSID for multipurpose devices should still be configured with a Platinum
QoS profile to ensure that real-time traffic generated by these devices is still given priority treatment
across the wireless infrastructure.
8-37
Chapter 8
Collaboration Endpoints
Mobile Endpoints
Device Mobility
When wireless endpoints move between locations in a multi-site centralized call processing deployment,
the Cisco Unified CM Device Mobility feature may be used to dynamically update the location of the
device based on the IP address the device uses to register to Unified CM. This prevents issues with call
routing, PSTN egress, and codec and media resource selection typically encountered when devices move
between locations. For more information on Device Mobility, see the section on Device Mobility,
page 21-13.
For more information about deploying wireless IP endpoints such as the Cisco Unified Wireless IP
Phone 7925G, refer to the deployment guides at
http://www.cisco.com/c/en/us/support/collaboration-endpoints/unified-ip-phone-7900-series/produ
cts-implementation-design-guides-list.html
For more information about deploying the Cisco IP Phone 8861 wirelessly, refer to the deployment guide
at
http://www.cisco.com/c/en/us/support/collaboration-endpoints/unified-ip-phone-8800-series/produ
cts-implementation-design-guides-list.html
For more information about deploying the Cisco Unified IP Phone 9971 wirelessly, refer to the
deployment guide at
http://www.cisco.com/c/en/us/support/collaboration-endpoints/unified-ip-phones-9900-series/prod
ucts-implementation-design-guides-list.html
For more information about deploying the Cisco DX Series endpoints wirelessly, refer to the deployment
guide at
http://www.cisco.com/c/en/us/support/collaboration-endpoints/desktop-collaboration-experience-d
x600-series/products-implementation-design-guides-list.html
Mobile Endpoints
Cisco mobile endpoint devices and mobile endpoint client applications register and communicate with
Unified CM for voice and video calling services. These devices and clients also enable additional
features and services such as enterprise messaging, presence, and corporate directory integration by
communicating with other back-end systems such as Cisco Unity Connection, Cisco IM and Presence,
and LDAP directories. Cisco offers the following mobile endpoint devices and clients:
Cisco WebEx Meetings, page 8-39, for Android, BlackBerry and Apple iOS devices
Cisco AnyConnect Secure Mobility Client, page 8-39, for Android and Apple iOS devices
8-38
Chapter 8
Collaboration Endpoints
Mobile Endpoints
enables additional features such as corporate directory access, enterprise visual voicemail, XMPP-based
enterprise instant messaging and presence, and secure remote attachment with Cisco Expressway mobile
and remote access.
For more information about Cisco Jabber for Android, refer to the data sheet and product documentation
at
http://www.cisco.com/c/en/us/products/unified-communications/jabber-android/index.html
For more information about Cisco Jabber for iPhone and iPad, refer to the data sheet and product
documentation at
http://www.cisco.com/c/en/us/products/unified-communications/jabber-iphone-ipad/index.html
8-39
Chapter 8
Collaboration Endpoints
Mobile Endpoints
WLAN Design
Because Cisco Jabber mobile clients are often attached to a WLAN, all of the previously mentioned
WLAN deployment considerations apply to mobile clients and devices, including WLAN RF design and
verification by site survey. In particular, Cisco recommends relying on the 5 GHz WLAN band
(802.11a/n/ac) whenever possible for connecting wireless endpoints capable of generating voice and
video traffic. 5 GHz WLANs provide better throughput and less interference for voice and video calls.
If the 2.4 GHz band is used for mobile clients and devices, Bluetooth should be avoided. Likewise, the
WLAN channel cell voice-only and video call capacity numbers covered in the section on Wireless Call
Capacity, page 8-35, should be considered when deploying these clients and devices.
8-40
Chapter 8
Collaboration Endpoints
Mobile Endpoints
Quality of Service
Cisco mobile client applications and devices generally mark Layer 3 QoS packet values in accordance
with Cisco collaboration QoS marking recommendations. This includes marking voice-only call media
traffic with DSCP 46 or PHB EF, video call media (including voice) traffic with DSCP 34 or PHB AF41,
and call signaling traffic with DSCP 24 or PHB CS3. Despite appropriate mobile client and device
application Layer 3 packet marking, Layer 2 802.11 WLAN packet marking (User Priority, or UP)
presents further challenges. Some devices may appropriately mark wireless Layer 2 802.11 User Priority
(UP) values (voice-only call media UP 6, video call media UP 5, and call signaling UP 3). However,
because Cisco mobile clients run on a variety of mobile devices, Layer 2 wireless QoS marking is
inconsistent and therefore cannot be relied upon to provide appropriate treatment to traffic on the
WLAN. In deployments with Cisco Unified Wireless LAN Controllers, enabling wireless SIP call
admission control (CAC) might provide some relief for incorrect or nonexistent Layer 2 WLAN
marking. SIP CAC utilizes media session snooping and ensures that downstream voice and video frames
are prioritized and/or treated correctly. Even assuming appropriate mobile client application Layer 3 or
even Layer 2 packet marking, mobile devices present many of the same challenges as desktop computers
in terms of generating many different types of traffic, including both data and real-time traffic. Given
this, mobile devices generally fall into the untrusted category of collaboration endpoints. For
deployments where mobile client devices are not considered trusted endpoints, packet re-marking based
on traffic type and port numbers is required to ensure that network priority queuing and dedicated
bandwidth are applied to appropriate traffic. In addition to re-marking the mobile device traffic, Cisco
recommends using network-based policing and rate limiting to ensure that the mobile client devices do
not consume too much network bandwidth.
Note
Mobile clients and devices may attach remotely to the enterprise using Cisco AnyConnect client over
the mobile data network or public or private Wi-Fi hot spots. Because these connections traverse the
Internet, there is no end-to-end QoS on the IP path and therefore all traffic is treated as best-effort. Voice
and video quality cannot be guaranteed over these types of connections.
8-41
Chapter 8
Collaboration Endpoints
Intelligent Proximity
As previously mentioned, Intelligent Proximity refers to features that leverage proximity-based
connections between Cisco hardware endpoints and mobile devices.
Intelligent Proximity for Mobile Voice capabilities available on the Cisco 8851, 8861, and DX Series
endpoints rely on Bluetooth pairing between the IP endpoint and a cellular or smart phone, enabling
hands-free audio and mobile contact and call history sharing.
As indicated previously, Intelligent Proximity for Mobile Voice on the 8851, 8861, and DX Series
endpoints and the Unified Mobility feature set are compatible. Further, Intelligent Proximity for Mobile
Voice on the 8851 and 8861 IP Phones is also compatible with Cisco Jabber, enabling audio-playout on
the IP Phone 8851 and 8861 while the video is played on the Jabber client device.
For more information on Intelligent Proximity and Bluetooth pairing, see Intelligent Proximity,
page 8-14.
Contact Sources
Cisco Jabber for Android and iOS defaults to using Basic Directory Integration (BDI), which relies on
integration to LDAP v3 compatible directories, including Microsoft Active Directory.
For integration with an LDAP directory that requires custom attribute mapping, these attribute mappings
can be created in a configuration file that can be downloaded to the client from the Unified CM server.
Cisco Jabber mobile clients also support the Unified CM User Data Service (UDS), which allows a client
to search for contacts using the Unified CM user database (which may be synchronized with an LDAP
directory). UDS is used automatically by Jabber mobile clients for contact resolution when they are
located outside of the corporate firewall and connected via Expressway mobile and remote access. UDS
should be leveraged as a contact source only in cases where Jabber mobile client are connected via
Expressway or for very small deployments. Otherwise, UDS is not recommended as a contact source due
to resource utilization on Unified CM. When clients leverage UDS as a contact source, the registration
capacity of the Unified CM node is reduced by half.
8-42
Chapter 8
Collaboration Endpoints
Third-Party IP Phones
Quality of Service
No additional configuration is required for Cisco Virtualization Experience Media Engine (VXME) if
the network is set up for 8021.q Dual VLAN. If the network is not setup for 802.1q Dual VLAN, QoS
will be best-effort and the thin client should be placed in the data VLAN. For details on traffic marking,
refer to the QoS design guides available at
http://www.cisco.com/c/en/us/solutions/enterprise/design-zone-ipv6/design-guide-listing.html
Call admission control for voice and video follow existing Cisco Unified IP Phone guidelines, and
bandwidth controls for the virtual desktop are provided through the connection broker settings.
Third-Party IP Phones
Some third-party IP phones and devices may be integrated with Cisco call control to provide basic IP
telephony functionality, as described in this section.
Third-Party SIP IP Phones
Third-party phones have specific local features that are independent of the call control signaling
protocol, such as features access buttons (fixed or variable). Basic SIP RFC support allows for certain
desktop features to be the same as on Cisco Unified IP Phones and also allows for interoperability of
certain features. However, these third-party SIP phones do not provide the full feature functionality of
Cisco Unified IP Phones.
Cisco works with key third-party vendors who are part of the Cisco Developer Network and who are
developing solutions that leverage Cisco Unified CM and Unified CME SIP capabilities. For example,
Tenacity Operating provides a software-based endpoint called accessaphone ipTTY, which enables
terminal teletype (TTY) or text-based communications for IP telephony. This software-based endpoint
can register and communicate with Cisco Unified CM as a third-party SIP phone.
For more information on Cisco's line-side SIP interoperability, refer to the Cisco Unified
Communications Manager programming guides at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-programming-reference-guides-list.html
8-43
Chapter 8
Collaboration Endpoints
For more information on the Cisco Developer Network and third-party development partners, refer to
the information available on the Cisco Developer Community at
http://developer.cisco.com
A Cisco Unified CM cluster, even when deployed as part of Cisco Business Edition 7000, supports
a maximum of 40,000 SCCP or SIP endpoints.
When deployed as part of Cisco Business Edition 6000, a Cisco Unified CM cluster supports a
maximum of 2,500 SCCP or SIP endpoints, depending on the server type.
Cisco Expressway-C and Expressway-E cluster pairs support a maximum of 10,000 remote endpoint
proxy registrations.
The above numbers are nominal maximum capacities. The maximum number of endpoints that the call
control platform will actually support depends on all of the other functions that the platform is
performing, the busy hour call attempts (BHCA) of the users, and so forth, and the actual capacity could
be less than the nominal maximum capacity. Unified CM CTI capacity must also be considered when
8-44
Chapter 8
Collaboration Endpoints
Design Considerations for Collaboration Endpoints
sizing the system to ensure that Jabber desktop clients and other deskphone control applications have
sufficient CTI capacity for operation. For more information on CTI sizing, refer to the chapter on
Collaboration Solution Sizing Guidance, page 25-1.
In addition to call control platform capacity, network capacity must be considered with regard to
bandwidth and call capacity. Of particular concern are 802.11 wireless attached devices such as the
Cisco Unified Wireless IP Phone 7925G or an Android smartphone running Cisco Jabber, where network
endpoint capacity is not determined by the number of physical ports but by the amount of bandwidth and
throughput available on the shared wireless network. See Wireless Call Capacity, page 8-35, for voice
and video call capacities per 802.11 channel cell.
For more information on endpoint capacity with Cisco call control, including platform-specific endpoint
capacities per node, see the chapter on Collaboration Solution Sizing Guidance, page 25-1.
Analog gateways are available both as standalone devices and as integrated interface modules on
Cisco IOS multiservice routers, and both types can be used within the same deployment. Select the
analog gateway or gateways that meet analog port density requirements across company locations.
Ensure that appropriate port capacity is provided for all locations in order to accommodate the
required analog devices.
Enable the role of Standard CTI Allow Control of Phones supporting Connected Xfer and conf
for the end-user configuration associated with the device in order to enable CTI monitoring and
control of Cisco Unified IP Phone 8900 and 9900 Series endpoints. Only after this role has been
enabled can CTI applications monitor or control these phones.
To minimize endpoint firmware upgrade times over the WAN to remote branches, consider
deploying a local TFTP server at the remote location and point endpoints located in that branch to
this local TFTP server using the load server parameter. Alternatively, consider the use of the Peer
File Sharing (PFS) feature when all or most of the devices at a particular remote location are the
same phone model.
Cisco Unified IP desk phones can be powered by power over Ethernet (PoE) when plugged into
inline power-capable switches or when deployed with an inline power injector. Consider the use of
inline power to reduce downtime and eliminate the need for an external power supply and wall power
outlet.
When deploying Cisco endpoints in branch locations separated from a centralized call processing
platform by a low-speed or unreliable WAN link, it is important to consider local call processing
redundancy. By using SRST or Enhanced SRST on a Cisco IOS router in each branch location, basic
IP telephony services can be maintained for the desk phones when connectivity to the centralized
call processing platform is lost. However, the set of available user-facing features is much smaller
when a device is registered to SRST than when the phone is registered to Unified CM.
For deployments with network voice and data VLAN separation, ensure that inter-VLAN routing
has been configured and allowed so that Cisco software-based endpoints that run on desktop
computers usually connected to data VLANs can communicate with endpoints on the voice VLAN.
This is also important for endpoints on the voice VLAN that may be dependent on data VLAN-based
resources that provide services such as directory and management.
A WLAN site survey must be conducted to ensure appropriate RF design and to identify and
eliminate sources of interference prior to deploying wireless and mobile endpoints capable of
generating real-time traffic on the wireless network. This is necessary to ensure acceptable voice and
video quality for calls traversing the WLAN.
8-45
Chapter 8
Collaboration Endpoints
Select a WLAN authentication and encryption method that not only adheres to company security
policies but also enables fast rekeying or authentication so that audio and video calls are not
interrupted when wireless endpoints move from one location to another.
Cisco recommends relying on the 5 GHz WLAN band (802.11a/n/ac) whenever possible for
connecting wireless endpoints and mobile client devices capable of generating voice and/or video
traffic. 5 GHz WLANs provide better throughput and less interference for voice and video calls. If
the 2.4 GHz band is used for connecting wireless client devices and endpoints, Bluetooth should be
avoided.
Provide appropriate network and call control capacity to support the number of endpoints deployed.
First, consider the endpoint registration and configuration capacities per call control platform
(maximum of 40,000 endpoints per Unified CM cluster, even when deployed as part of Cisco
Business Edition 7000; 2,500 endpoints per cluster when deployed as part of Cisco Business
Edition 6000; or 10,000 remote endpoint registrations over Cisco Expressway. Next, consider call
capacities per wireless channel cell for wireless attached endpoints, and the maximum of 27
bidirectional voice-only streams or maximum of 8 simultaneous voice and video streams or calls per
WLAN channel cell.
Ensure that the end-to-end network infrastructure has been configured with appropriate QoS
policies, including marking and re-marking as appropriate, trust boundaries, queuing with both
priority and dedicated bandwidth queues, rate limiting, and policing, so that collaboration endpoints
deliver high-quality voice and video to end users.
8-46
CH A P T E R
Call Processing
Revised: July 30, 2015
The handling and processing of voice and video calls is a critical function provided by IP telephony
systems. This functionality is handled by some type of call processing entity or agent. Given the critical
nature of call processing operations, it is important to design unified communications deployments to
ensure that call processing systems are scalable enough to handle the required number of users and
devices and are resilient enough to handle various network and application outages or failures.
This chapter provides guidance for designing scalable and resilient call processing systems with Cisco
call processing products. These products include Cisco Unified Communications Manager (Unified CM)
and Cisco Unified Communications Manager Express (Unified CME). The discussions focus
predominately on the following factors:
9-1
Chapter 9
Call Processing
Described in
Revision Date
http://www.cisco.com/go/ucsrnd
9-2
Chapter 9
Call Processing
Call Processing Architecture
Cisco Unified Communications Manager (Unified CM) and Cisco TelePresence Video Communication
Server (VCS) are available as standard Cisco Collaboration products or through Cisco Business
Edition 6000 and Cisco Business Edition 7000, which are packaged collaboration solutions that include
call processing services and other services such as messaging, conferencing, and contact center.
The Cisco Business Edition solutions simplify the quoting/ordering process and accelerate deployments
by providing pre-configured hardware, pre-installed licensed hypervisor, and pre-loaded and/or
pre-installed Cisco Collaboration applications. Cisco Business Edition 6000S is targeted for deployment
with up to 150 users and 300 devices. Cisco Business Edition 6000M and Cisco Business Edition 6000H
are targeted for deployments with up to 1,000 users. Cisco Business Edition 7000 is targeted for
deployments with more than 1,000 users. The design and sizing of the Cisco Collaboration applications
have been simplified with Cisco Business Edition 6000. With Cisco Business Edition 7000, however,
normal Unified CM design and sizing guidelines apply.
Cisco Unified CM runs only as a virtual application; it cannot be deployed directly on a Cisco MCS
or UCS server, for example.
Cisco Unified CME runs within the Cisco IOS software and does not support virtualization.
For more information on the considerations for designing and deploying virtualization of Cisco Unified
Communications applications, refer to the information available at
http://www.cisco.com/go/uc-virtualized
9-3
Chapter 9
Call Processing
Specifications-based hardware
Specifications-based hardware (sometimes simply referred as "specs-based") provides more flexible
hardware configurations. For example, it allows you to select a platform based on a Cisco UCS TRC
and to change the CPU model, number of cores, and RAID configuration, and/or to use an iSCSI or
NAS storage. If desired, it also allows you to use a server vendor other than Cisco. Any
specifications-based hardware server, whether it is Cisco or not, must be listed in the following
VMware Compatibility Guide:
http://www.vmware.com/resources/compatibility/search.php
While specification-based hardware provides more flexible hardware configurations, some
requirements must still be met. For example, there are requirements around the CPU model and
minimum CPU speed, and vCenter is required in order to collect logs and statistics. With
specifications-based hardware, it is important to understand that the hardware configuration has not
been explicitly validated by Cisco with Cisco Collaboration applications. Therefore hardware
compatibility cannot be guaranteed, and performance of the Cisco Collaboration applications cannot
be predicted or assured. To obtain guidance on the performance of Cisco Collaboration applications
with specifications-based hardware, use the TRCs or Cisco Business Edition 6000 and 7000
hardware platforms as references. For more information, refer to the documentation at
http://www.cisco.com/go/uc-virtualized.
Cisco Unified CME runs on Cisco Integrated Services Routers (ISR) such as the Cisco 2900, 3900, or
4000 Series ISRs. Cisco Unified CME does not run as a virtual application.
9-4
Chapter 9
Call Processing
Call Processing Architecture
Determining the appropriate call processing type and platform for a particular deployment will depend
on the scale, performance, and redundancy required. In general, Unified CM provides a very wide range
of capacity options and higher availability, while Cisco Unified CME provides lower levels of capacity
and redundancy. For specifics regarding redundancy and scalability, see the sections on High Availability
for Call Processing, page 9-12, and Capacity Planning for Call Processing, page 9-21.
Publisher
M
TFTP Servers
(Subscriber)
MTP, Conferencing,
MoH, and Annunciator
Servers (Subscriber)
191953
Unified CM Servers
(Subscriber)
Publisher
The publisher is a required server node in all clusters, and as shown in Figure 9-1, there can be only one
publisher per cluster. This server node is the first to be installed and provides the database services to all
other subscribers in the cluster. The publisher node is the only server node that has full read and write
access to the configuration database.
9-5
Chapter 9
Call Processing
On larger systems with more than 1250 users, Cisco recommends a dedicated publisher to prevent
administrative operations from affecting the telephony services. A dedicated publisher does not provide
call processing or TFTP services running on the node. Instead, other subscriber nodes within the cluster
provide these services.
The choice of the VM configuration for the publisher should be based on the desired scale and
performance of the cluster. Cisco recommends that the publisher have the same server node performance
capability as the call processing subscribers.
Subscriber
When the software is installed initially, only the database and network services are enabled. All
subscriber nodes subscribe to the publisher to obtain a copy of the database information. However, in
order to reduce initialization time for the Unified CM cluster, all subscriber nodes in the cluster attempt
to use their local copy of the database when initializing. This reduces the overall initialization time for
a Unified CM cluster. All subscriber nodes rely on change notification from the publisher or other
subscriber nodes in order to keep their local copy of the database updated.
As shown in Figure 9-1, multiple subscriber nodes can be members of the same cluster. Subscriber nodes
include Unified CM call processing subscriber nodes, TFTP subscriber nodes, and media resource
subscriber nodes that provide functions such as conferencing and music on hold (MoH).
TFTP Subscriber
A TFTP subscriber or server node performs two main functions as part of the Unified CM cluster:
The serving of files for services, including configuration files for devices such as phones and
gateways, binary files for the upgrade of phones as well as some gateways, and various security files
Generation of configuration and security files, which are usually signed and in some cases encrypted
before being available for download
The Cisco TFTP service that provides this functionality can be enabled on any server node in the cluster.
However, in a cluster with more than 1250 users, other services might be impacted by configuration
changes that can cause the TFTP service to regenerate configuration files. Therefore, Cisco recommends
that you dedicate a specific subscriber node to the TFTP service, as shown in Figure 9-1, for a cluster
with more than 1250 users or any features that cause frequent configuration changes.
Cisco recommends that you use the same VM configuration for the TFTP subscribers as used for the call
processing subscribers.
9-6
Chapter 9
Call Processing
Call Processing Architecture
Music on Hold (MoH) Provides multicast or unicast music to devices that are placed on hold or
temporary hold, transferred, or added to a conference. (See Music on Hold, page 7-18.)
Conference bridges Provide software-based conferencing for instant and permanent conferences.
(See Transcoding, page 7-6.)
Media termination point (MTP) services Provide features for H.323 clients, H.323 trunks, and
Session Initiation Protocol (SIP) endpoints and trunks. (See Media Termination Point (MTP),
page 7-7.)
Because of the additional processing and network requirements for media resource services, it is
essential to follow all guidelines for running media resources within a cluster. Generally, Cisco
recommends non-dedicated media resource subscribers for multicast MoH and annunciator services, but
dedicated media resource subscribers as shown in Figure 9-1 are recommended for unicast MoH as well
as large-scale software-based conferencing and MTPs unless those services are within the design
guidelines detailed in the chapter on Media Resources, page 7-1.
The CTI Manager service acts as a broker between the Cisco CallManager service and TAPI or JTAPI
integrated applications. This service is required in a cluster for any applications that utilize CTI. The
CTI Manager service provides authentication of the CTI application and enables the application to
monitor and/or control endpoint lines. CTI Manager can be enabled only on call processing subscribers,
thus allowing for a maximum of eight nodes running the CTI Manager service in a cluster.
For more details on CTI Manager, see Computer Telephony Integration (CTI), page 9-27.
Unified CM Applications
Various types of application services can be enabled on Unified CM, such as Cisco Unified CM
Assistant, Extension Mobility, and Web Dialer. For detailed design guidance on these applications, see
the chapter on Cisco Unified CM Applications, page 18-1. The Cisco IM and Presence service can also
be added (see the chapter on Collaboration Instant Messaging and Presence, page 20-1).
9-7
Chapter 9
Call Processing
Intracluster Communications
There are two primary kinds of intracluster communications, or communications within a Unified CM
cluster (see Figure 9-2 and Figure 9-3.) The first is a mechanism for distributing the database that
contains all the device configuration information (see Database replication in Figure 9-2). The
configuration database is stored on a publisher node, and a copy is replicated to the subscriber nodes of
the cluster. Most of the database changes are made on the publisher and are then communicated to the
subscriber databases, thus ensuring that the configuration is consistent across the members of the cluster
and facilitating spatial redundancy of the database.
Database modifications for user-facing call processing features are made on the subscriber nodes to
which an end-user device is registered. The subscriber nodes then replicate these database modifications
to all the other nodes in the cluster, thus providing redundancy for the user-facing features. (See "Call
processing user-facing feature replication" in Figure 9-2.) These features include:
Privacy Enable/Disable
Device Mobility
Certificate Authority Proxy Function (CAPF) status for end users and applications users
9-8
Chapter 9
Call Processing
Call Processing Architecture
Publisher
M
Unified
CM
Unified
CM
TFTP
Unified
CM
Publisher
TFTP
Unified
CM
Unified
CM
Unified
CM
Unified
CM
Unified
CM
Subscribers
Subscribers
Database replication
191955
Figure 9-2
The second type of intracluster communication, called Intra-Cluster Communication Signaling (ICCS),
involves the propagation and replication of run-time data such as registration of devices, locations
bandwidth, and shared media resources (see Figure 9-3). This information is shared across all members
of a cluster running the Cisco CallManager Service (call processing subscribers), and it ensures the
optimum routing of calls between members of the cluster and associated gateways.
Intra-Cluster Communication Signaling (ICCS)
Publisher
M
Unified CM
TFTP
M
Unified CM
Unified CM
Unified CM
Subscribers
Intra-Cluster Communication Signaling (ICCS)
191954
Figure 9-3
9-9
Chapter 9
Call Processing
Intracluster Security
Each server node in a Unified CM cluster runs an internal dynamic firewall. The application ports on
Unified CM are protected by source IP filtering. The dynamic firewall opens these application ports only
to authenticated or trusted server nodes. (See Figure 9-4.)
Figure 9-4
Intracluster Security
Management Channel
UDP Port 8500
FIREWALL
Subscriber 1
M
(Subscriber 1,
Authen. Ports)
ACCEPT
(Subscriber 2,
Authen. Ports)
ACCEPT
Subscriber 2
M
Unauthorized Source
(Any,
Authen. Ports)
DENY
191956
Publisher
This security mechanism is applicable only between server nodes in a single Unified CM cluster.
Unified CM subscribers are authenticated in a cluster before they can access the publisher's database.
The intra-cluster communication and database replication take place only between authenticated server
nodes. During the installation process, a subscriber node is authenticated to the publisher using a
pre-shared key authentication mechanism. The authentication process involves the following steps:
1.
2.
3.
Install the subscriber node using the same security password used during publisher server
installation.
4.
After the subscriber is installed, the server node attempts to establish connection to the publisher on
a management channel using UDP 8500. The subscriber sends all the credentials to the publisher,
such as hostname, IP address, and so forth. The credentials are authenticated using the security
password used during the installation process.
5.
The publisher verifies the subscriber's credentials using its own security password.
6.
The publisher adds the subscriber as a trusted source to its dynamic firewall table if the information
is valid. The subscriber is allowed access to the database.
7.
The subscriber gets a list of other subscriber nodes from the publisher. All the subscribers establish
a management channel with each other, thus creating a mesh topology.
9-10
Chapter 9
Call Processing
Call Processing Architecture
Cisco recommends using the same VM configuration for all nodes in a cluster. Mixing Unified CM
VM configurations is allowed, but for more details refer to the section on Mixing Unified CM VM
Configurations, page 9-8.
Under normal circumstances, place all members of the cluster within the same LAN or MAN.
If the cluster spans an IP WAN, follow the guidelines for clustering over an IP WAN as specified in
the section on Clustering Over the IP WAN, page 10-41.
A Unified CM cluster may contain as many as 20 server nodes, of which a maximum of eight call
processing subscribers (nodes running the Cisco CallManager Service) are allowed. The other server
nodes within the cluster may be configured as a dedicated database publisher, dedicated TFTP
subscriber, or media resource subscriber.
When deploying a two-node cluster, Cisco recommends that you do not exceed 1250 users in the
cluster. Above 1250 users, a dedicated publisher and separate server nodes for primary and backup
call processing subscribers is recommended.
Business Edition 6000 provides a single instance of Unified CM (a Unified CM publisher that also
handles call processing). Additional Business Edition 6000 server(s) may be deployed to provide
subscriber redundancy either in an active/standby or load balancing fashion for Unified CM as well
as some other co-resident applications. However, adding new nodes and new hardware platforms
does not increase capacity. For example, the user and device capacities do not increase.
Each Unified CM node instance can be a publisher node, call processing subscriber node, TFTP
subscriber node, or media resource subscriber node. Only a single publisher node per cluster is
supported.
With virtualization, Unified CM no longer supports the Cisco Messaging Interface (CMI) service
for Simplified Message Desk Interface (SMDI) integrations, fixed MoH audio source integration for
live MoH audio feeds using the audio cards (MOH-USB-AUDIO=), or flash drives to these servers.
The following alternate options are available:
For MoH live audio source feed, consider using Cisco IOS-based gateway multicast MoH for
9-11
Chapter 9
Call Processing
9-12
Chapter 9
Call Processing
High Availability for Call Processing
call processing, this means distributing cluster server nodes among multiple buildings or locations
within the LAN or MAN deployment whenever possible. And at the very least, it means physically
distributing network connections between different physical network switches in the same location.
Furthermore, even though Cisco Unified CME is a standalone call processing entity, providing physical
distribution and therefore redundancy for this call processing entity still makes sense when deploying
multiple call processing entities. Whenever possible in those scenarios, install each instance of
Unified CME in a different physical location within the network, or at the very least physically attach
them to different network switches.
Two to one (2:1) For every two primary call processing subscribers, there is one shared secondary
or backup call processing subscriber.
One to one (1:1) For every primary call processing subscriber, there is a secondary or backup call
processing subscriber.
These redundancy schemes are facilitated by the built-in registration failover mechanism within the
Unified CM cluster architecture, which enables endpoints to re-register to a backup call processing
subscriber node when the endpoint's primary call processing subscriber node fails. The registration
failover mechanism can achieve failover rates for Skinny Client Control Protocol (SCCP) IP phones of
approximately 125 registrations per second. The registration failover rate for Session Initiation Protocol
(SIP) phones is approximately 40 registrations per second.
The call processing redundancy scheme you select determines not only the fault tolerance of the
deployment, but also the fault tolerance of any upgrade.
With 1:1 redundancy, multiple primary call processing subscriber failures can occur without impacting
call processing capabilities. With 2:1 redundancy, on the other hand, only one of the primary call
processing subscribers out of the two primary call processing subscribers that share a backup call
processing subscriber can fail without impacting call processing. However, if the total number of
endpoints registered across both primary subscribers and the traffic to those two primary subscribers are
within the capacity limits of the backup subscriber, then the backup subscriber is able to handle the
failure of both primary subscribers.
Note
Do not deploy 2:1 redundancy if the total capacity utilization across the two primary subscribers would
exceed the capacity of the backup subscriber. For example, if the call processing capacity or endpoints
capacity utilization exceeds 50% on both primary subscribers, the backup subscriber would not be able
to handle call processing services properly if both primary subscribers fail. In these scenarios, for
example, some endpoints might not be able to register, some new calls might not be established, and
some services and features might not operate properly because the backup subscriber system capacity
has been exceeded.
9-13
Chapter 9
Call Processing
Likewise, with the 1:1 redundancy scheme, upgrades to the cluster can be performed with only a single
set of endpoint registration failover periods impacting the call processing services. Whereas with the 2:1
redundancy scheme, upgrades to the cluster can require multiple registration failover periods.
A Unified CM cluster can be upgraded with minimal impact to the services. Two different versions
(releases) of Unified CM may be on the same server node, one in the active partition and the other in the
inactive partition. All services and devices use the Unified CM version in the active partition for all
Unified CM functionality. During the upgrade process, the cluster operations continue using its current
release of Unified CM in the active partition, while the upgrade version gets installed in the inactive
partition. Once the upgrade process is complete, the server nodes can be rebooted to switch the inactive
partition to the active partition, thus running the new version of Unified CM.
With the 1:1 redundancy scheme, the following steps enable you to upgrade the cluster while minimizing
downtime:
Step 1
Install the new version of Unified CM in the inactive partition, first on the publisher and then on all
subscribers (call processing, TFTP, and media resource subscribers). Do not reboot.
Step 2
Step 3
Reboot the TFTP subscriber node(s) one at a time and switch to the new version.
Step 4
Reboot any dedicated media resource subscriber nodes one at a time and switch to the new version.
Step 5
Reboot the backup call processing subscribers one at a time and switch to the new version.
Step 6
Reboot the primary call processing subscribers one at a time and switch to the new version. Device
registrations will fail-over to the previously upgraded and rebooted backup call processing subscribers.
After each primary call processing subscriber is rebooted, devices will begin to re-register to the primary
call processing subscriber.
With this upgrade method, there is no period (except for the registration failover period) when devices
are registered to subscriber nodes that are running different versions of the Unified CM software. All
these steps can be automated using Cisco Prime Collaboration.
While the 2:1 redundancy scheme allows for fewer server nodes in a cluster, registration failover occurs
more frequently during upgrades, increasing the overall duration of the upgrade as well as the amount
of time call processing services for a particular endpoint will be unavailable. Because there is only a
single backup call processing subscriber per pair of primary call processing subscribers, it might be
possible to reboot to the new version on only one of the primary call processing subscribers in a pair at
a time in order to prevent oversubscribing the single backup call processing subscriber. As a result, there
may be a period of time after the first primary call processing subscriber in each pair is switched to the
new version, in which endpoint registrations will have to be moved from the backup subscriber to the
newly upgraded primary subscriber before the endpoint registrations on the second primary subscriber
can be moved to the backup subscriber to allow a reboot to the new version. During this time, not only
will endpoints on the second primary call processing subscriber be unavailable while they re-register to
the backup subscriber, but until they re-register to a node running the new version, they will also be
unable to reach endpoints on other subscriber nodes that have already been upgraded.
Note
Before you do an upgrade, Cisco recommends that you back up the Unified CM and Call Detail Record
(CDR) database to an external network directory using the Disaster Recovery Framework. This practice
will prevent any loss of data if the upgrade fails.
9-14
Chapter 9
Call Processing
High Availability for Call Processing
Note
Because an upgrade of a Unified CM cluster results in a period of time in which some or most devices
lose registration and call processing services temporarily, you should plan upgrades in advance and
implement them during a scheduled maintenance window. While downtime and loss of services to
devices can be minimized by selecting the 1:1 redundancy scheme, there will still be some period of time
in which call processing services are not available to some or all users.
For more information on upgrading Unified CM, refer to the install and upgrade guides available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_installation_guides_list.html
Unified CM Redundancy with Survivable Remote Site Telephony (SRST)
Cisco IOS SRST provides highly available call processing services for endpoints in locations remote
from the Unified CM cluster. Unified CM clustering redundancy schemes certainly provide a high level
of redundancy for call processing and other application services within a LAN or MAN environment.
However, for remote locations separated from the central Unified CM cluster by a WAN or other
low-speed links, SRST can be used as a redundancy method to provide basic call processing services to
these remote locations in the event of loss of network connectivity between the remote and central sites.
Cisco recommends deploying SRST-capable Cisco IOS routers at each remote site where call processing
services are considered critical and need to be maintained in the event that connectivity to the
Unified CM cluster is lost. Endpoints at these remote locations must be configured with an appropriate
SRST reference within Unified CM so that the endpoint knows what address to use to connect to the
SRST router for call processing services when connectivity to Unified CM subscribers is unavailable.
Cisco Unified Enhanced SRST (E-SRST) on a Cisco IOS router can also be used at a remote site to
provide backup call processing functionality in the event that connectivity to the central Unified CM
cluster is lost. E-SRST provides more telephony features for the IP phones than are available with the
regular SRST feature on a router. However, the endpoint capacities for Unified E-SRST are typically less
than for basic SRST. Both SRST and E-SRST are supported with Cisco Unified SRST Manager, which
synchronizes configurations from Unified CM with SRST and E-SRST, thus reducing manual
configuration required in the branch SRST or E-SRST router and enabling users to have a similar calling
experience in both SRST and normal modes.
9-15
Chapter 9
Call Processing
The following figures illustrate typical cluster configurations to provide call processing redundancy with
Unified CM.
Basic Redundancy Schemes
Backup
M
M
1 to 2500
Backup
2501 to
5000
Backup
1 to 2500
2501 to
5000
Cost-efficient redundancy
Simplified configurations
87424
Figure 9-5
Figure 9-5 illustrates the two basic redundancy schemes available. In each case the backup server node
must be capable of handling the capacity of at least a single primary call processing server node failure.
In the 2:1 redundancy scheme, the backup might have to be capable of handling the failure of a single
call processing server node or potentially both primary call processing server nodes, depending on the
requirements of a particular deployment. For information on capacity sizing and choosing the VM
configurations, see the section on Capacity Planning for Call Processing, page 9-21.
Note
2:1 redundancy is not supported with the 10,000-User VM configuration due to potential overload on the
backup subscriber.
9-16
Chapter 9
Call Processing
High Availability for Call Processing
Figure 9-6
Publisher
Publisher
Publisher and
backup subscriber
Primary
Backup
Primary
Primary
Backup
Backup
Primary
5
Publisher
Publisher
Backup
Primary
Backup
Primary
Backup
Primary
Backup
Primary
Backup
Primary
Backup
Primary
Backup
Primary
Figure 9-7
114952
Publisher and
backup subscriber
Primary
Publisher
Primary
Publisher
Backup
Backup
Primary
Primary
5
Publisher
Publisher
M
M
Primary
Primary
Primary
Primary
Primary
Primary
Backup
Backup
Backup
Primary
114953
Backup
9-17
Chapter 9
Call Processing
In Figure 9-6, the five options shown all indicate 1:1 redundancy. In Figure 9-7, the five options shown
all indicate 2:1 redundancy. In both cases, Option 1 is used for clusters supporting less than 1250 users.
Options 2 through 5 illustrate increasingly scalable clusters for each redundancy scheme. The exact scale
depends on the hardware platforms chosen or required.
These illustrations show only publisher and call processing subscribers. They do not account for other
subscriber nodes such as TFTP and media resources.
Note
It is possible to define up to three call processing subscribers per Unified CM group. Adding a tertiary
subscriber for additional backup extends the above redundancy schemes to 2:1:1 or 1:1:1 redundancy.
However, with the exception of using tertiary subscriber nodes in deployments with clustering over the
WAN (see Remote Failover Deployment Model, page 10-52), tertiary subscriber redundancy is not
recommended for endpoint devices located in remote sites because failover to SRST will be further
delayed if the endpoint must check for connectivity to a tertiary subscriber. The tertiary subscribers also
count against the maximum number of call processing subscribers in a cluster (8 call processing
subscriber nodes).
Although not shown in the Figure 9-6 or Figure 9-7, it is also possible to deploy a single-node cluster.
The single-node cluster should not exceed 1000 endpoint configuration and registrations. Note that in a
single-node configuration, there is no backup call processing subscriber and therefore no cluster
redundancy mechanism. Survivable Remote Site Telephony (SRST) can be used as a redundancy
mechanism in these types of deployments to provide minimal call processing services during periods
when Unified CM is not available.
Load Balancing
In Unified CM clusters with the 1:1 redundancy scheme, device registration and call processing services
can be load-balanced across the primary and backup call processing subscriber.
Normally a backup server node has no devices registered to it unless its primary is unavailable. This
makes it easier to troubleshoot a deployment because there is a maximum of four primary call processing
subscriber nodes that will be handling the call processing load at a given time. Further, this potentially
simplifies configuration by reducing the number of Unified CM redundancy groups and device pools.
In a load-balanced deployment, up to half of the device registration and call processing load can be
moved from the primary to the secondary subscriber by using the Unified CM redundancy groups and
device pool settings. In this way each primary and backup call processing subscriber pair provides device
registration and call processing services to as many as half of the total devices serviced by this pair of
call processing subscribers. This is referred to as 50/50 load balancing. The 50/50 load balancing model
provides the following benefits:
Load sharing The registration and call processing load is distributed on multiple server nodes,
which can provide faster response time.
Faster failover and failback Because all devices (such as IP phones, CTI ports, gateways, trunks,
voicemail ports, and so forth) are distributed across all active subscribers, only some of the devices
fail-over to the secondary subscriber if the primary subscriber fails. In this way, you can reduce by
50% the impact of any server node becoming unavailable.
To plan for 50/50 load balancing, calculate the capacity of a cluster without load balancing, and then
distribute the load across the primary and backup subscribers based on devices and call volume. To allow
for failure of the primary or the backup server node, do not let the total load on the primary and
secondary subscribers exceed that of a single subscriber node.
9-18
Chapter 9
Call Processing
High Availability for Call Processing
Note
During upgrades of a Unified CM cluster with 50/50 load balancing, upgrades to the backup call
processing subscriber will result in devices registered to that subscriber (half of the total devices serviced
by the primary and backup subscriber pair) failing over to the primary call processing subscriber.
TFTP Redundancy
Cisco recommends deploying more than one dedicated TFTP subscriber node for a large Unified CM
cluster, thus providing redundancy for TFTP services. While two TFTP subscribers are typically
sufficient, more than two TFTP server nodes can be deployed in a cluster.
In addition to providing one or more redundant TFTP subscribers, you must configure endpoints to take
advantage of these redundant TFTP nodes. When configuring the TFTP options using DHCP or
statically, define a TFTP subscriber node IP address array containing the IP addresses of both TFTP
subscriber nodes within the cluster. In this way, by creating two DHCP scopes with two different IP
address arrays (or by manually configuring endpoints with two different TFTP subscriber node IP
addresses), you can assign half of the endpoint devices to use TFTP subscriber A as the primary and
TFTP subscriber B as the backup, and the other half to use TFTP subscriber B as the primary and TFTP
subscriber A as the backup. In addition to providing redundancy during a failure of one TFTP subscriber,
this method of distributing endpoints across multiple TFTP subscribers provides load balancing so that
one TFTP subscriber is not handling all the TFTP service load.
Note
When adding a specific binary or firmware load for a phone or gateway, you must add the file(s) to each
TFTP subscriber node in the cluster.
Two CTI ports would have a Unified CM redundancy group of server node A as the primary call
processing subscriber and server node B as the backup subscriber. The other two ports would have
a Unified CM redundancy group of server node B as the primary subscriber and server node A as
the backup subscriber.
The IVR application would be configured to use the CTI Manager on subscriber A as the primary
and subscriber B as the backup.
9-19
Chapter 9
Call Processing
The above example allows for redundancy in case of failure of the CTI Manager on subscriber A and
also allows for the IVR call load to be spread across two server nodes. This approach also minimizes the
impact of a Unified CM subscriber node failure.
For more details on CTI and CTI Manager, see Computer Telephony Integration (CTI), page 9-27.
When deploying multiple TFTP or media resource subscriber nodes instances for redundancy of
those services, always distribute redundant subscriber nodes across more than one server to ensure
that a failure of a single server does not eliminate those services. This ensures that, given the failure
of a blade containing a TFTP or media resource subscriber, endpoints will still be able to access
TFTP and media resource services on a subscriber node residing on another server. Endpoints can
also be distributed among redundant TFTP and media resource subscriber node instances to balance
system load in non-failure scenarios.
When deploying CTI applications, always make sure that call processing subscriber node instances
running the CTI Manager service are distributed across more than one server to ensure that a failure
of a single server does not eliminate CTI services. Further, CTI applications should be configured
to use the CTI Manager service running on the subscriber node instance on one server as the primary
CTI Manager and the CTI Manager service running on the subscriber node on another server as the
backup CTI Manager.
Blade Server 2
Media Resource Subscriber 1
TFTP Subscriber 2
Subscriber-Backup 1
Subscriber-Primary 2
Blade Server 1
Publisher
TFTP Subscriber 1
Subscriber-Primary 1
Subscriber-Backup 2
UCS B200 M1
!
UCS B200 M1
!
2
!
Reset
Console
Reset
Console
UCS B200 M1
!
UCS B200 M1
!
4
!
Blade Server 3
Media Resource Subscriber 2
Subscriber-Primary 3
Subscriber-Backup 4
Reset
Console
UCS B200 M1
!
Reset
Console
UCS B200 M1
!
6
!
Console
Reset
Console
Reset
Blade Server 4
Subscriber-Backup 3
Subscriber-Primary 4
UCS B-Series Blade Server
348695
Figure 9-8
9-20
Chapter 9
Call Processing
Capacity Planning for Call Processing
When using blade servers with a chassis (for example, B-Series blade servers with a Cisco UCS 5100
Blade chassis), in addition to distributing subscriber node instances across multiple blades, you may
distribute subscriber node instances across multiple blade chassis for additional redundancy and
scalability.
For more information about redundancy and provisioning of host resources for virtual machines, refer
to the documentation at http://www.cisco.com/go/uc-virtualized.
Note
More than two physical servers may be clustered to provide additional redundancy and/or geographic
distribution as with a clustering over the WAN deployment. However, with Cisco Business Edition 6000,
the additional server(s) only provides redundancy and not a capacity increase. For example, with
BE6000S, the total number of users across the cluster may not exceed 150; with BE6000M and
BE6000H, the total number of users across the cluster may not exceed 1,000. A deployment exceeding
this limit is considered to be a standard Unified CM cluster, and as such the deployment must follow
high availability design guidance for standard Unified CM. (See Unified CM High Availability,
page 9-13.) With Cisco Business Edition 7000, the capacity is not limited to 1,000 users; rather, the
standard application capacity planning and design rules apply.
9-21
Chapter 9
Call Processing
Within a cluster, a maximum of 8 call processing subscriber nodes can be enabled with the Cisco
CallManager Service. Other server nodes may be used for more dedicated functions such as
publisher, TFTP subscribers, and media resources subscribers.
Each Unified CM node can support registration for a maximum of 10,000 secured or unsecured
SCCP or SIP endpoints. Each cluster can support configuration and registration for a maximum of
40,000 secured or unsecured SCCP or SIP endpoints.
There are several VM configuration options for Cisco Unified CM available in the OVA, depending
on the required capacity. The names of the VM configurations correspond to the maximum number
of users per node, assuming that each user has one phone. When the ratio of number phones per user
is different than one, the VM configuration names actually correspond to the maximum number of
9-22
Chapter 9
Call Processing
Capacity Planning for Call Processing
endpoints per node. Depending on different variables such as BHCA and feature set used, the actual
number of users or endpoints could be lower. To validate the sizing of a deployment, use the Cisco
Unified Communications Sizing Tool, available at
http://tools.cisco.com/cucst
With Business Edition 6000, the Unified CM VM configuration names refer to the maximum
number of users per node. The number of phones could be higher. For more details, see the section
on Cisco Business Edition Capacity Planning, page 9-24.
Some VM configurations require more powerful hardware platforms. For instance, the Unified CM
2.5k, 7.5k, and 10k user VM configurations are not supported on the Business Edition 6000
platforms, small TRCs, or specifications-based servers that have a restricted UC performance CPU.
They require more powerful platforms such as the Business Edition 7000, medium or larger TRCs,
or specifications-based servers that have a full UC performance CPU. For more details, refer to
www.cisco.com/go/uc-virtualized.
The default trace setting for the CallManager service is 1,500 files of 10 MB for Signaling
Distribution Layer (SDL) traces. Unless specific troubleshooting under high call rates requires
increasing the maximum file setting, the default settings are sufficient for collecting sufficient traces
in most circumstances.
For more information about Unified CM capacity planning considerations, including sizing limits as
well as a complete discussion of system sizing, capacity planning, and deployment considerations, see
the chapter on Collaboration Solution Sizing Guidance, page 25-1.
Megacluster
The term megacluster defines and identifies certain Unified CM deployments that allow for further
increases in scalability. A megacluster provides more device capacity through the support of additional
Unified CM subscriber nodes, with a maximum of eight Unified CM subscriber pairs (1:1 redundancy)
per megacluster, thus allowing for a maximum of 80,000 devices.
A megacluster can also be deployed where customers simply require non-locally redundant call
processing functionality, rather than using Survivable Remote Site Telephony (SRST), to scale beyond
the maximum eight sites allowed in a standard cluster deployment and up to 16 Unified CM subscriber
nodes per megacluster. For example, consider a large hospital that has twelve locations and each location
has only 1,000 devices. This total of 12,000 devices could be accommodated within a standard cluster,
which has a maximum device capacity of 40,000 devices. However, in this case it is the need for
additional Unified CM subscribers, rather than additional device capacity, that requires a megacluster
deployment. In this example, a Unified CM subscriber node could be deployed in each location, and each
Unified CM subscriber could serve as the primary subscriber for the local endpoints and as a backup
subscriber for endpoints from another location.
When considering a megacluster deployment, the primary areas impacting capacity are as follows:
The megacluster may contain a total of 21 server nodes consisting of 16 subscriber nodes, 2 TFTP
server nodes, 2 music on hold (MoH) server nodes, and 1 publisher node.
All other capacities relating to a standard cluster also apply to a megacluster. Note that support for a
megacluster deployment is granted only following the successful review of a detailed design, including
the submission of the results from the Cisco Unified Communications Sizing Tool. For more information
about the Cisco Unified Communications Sizing Tool and the sizing of Unified CM standard clusters and
megaclusters, see the chapter on Collaboration Solution Sizing Guidance, page 25-1.
9-23
Chapter 9
Call Processing
Due to the many potential complexities surrounding megacluster deployments, customers who wish to
pursue such a deployment must engage either their Cisco Account Team, Cisco Advanced Services, or
their certified Cisco Unified Communications Partner.
Note
Unless otherwise specified, all information contained within this SRND that relates to call processing
deployments (including capacity, high availability, and general design considerations) applies only to a
standard cluster.
Cisco Business Edition 6000 and Cisco Business Edition 7000 Co-residency Policy Requirements
http://www.cisco.com/c/en/us/support/unified-communications/business-edition-6000/product
s-device-support-tables-list.html
9-24
Chapter 9
Call Processing
Design Considerations for Call Processing
Cisco Unified CM runs only as a virtualized application on the VMware Hypervisor. It does not run
directly on a hardware platform without the VMware Hypervisor.
You can enable a maximum of 8 call processing subscriber nodes (nodes running the Cisco
CallManager Service) within a Cisco Unified CM cluster. Additional server nodes may be dedicated
and used for publisher, TFTP, and media resources services. An approved megacluster deployment
supports a maximum of 16 call processing subscriber nodes.
Each Unified CM cluster can support configuration and registration for a maximum of 40,000
secured or unsecured endpoints. For additional information about Unified CM capacity planning,
including per-platform sizing limits, see the chapter on Collaboration Solution Sizing Guidance,
page 25-1.
When deploying a two-node cluster, Cisco recommends that you do not exceed 1,250 users in the
cluster. Above 1,250 users, Cisco recommends a dedicated publisher and separate nodes for primary
and backup call processing subscribers.
Cisco recommends using the same VM configuration for all nodes in a cluster. Mixing Unified CM
VM configurations is allowed, but for more details refer to the section on Mixing Unified CM VM
Configurations, page 9-8.
2:1 redundancy is not supported when using the 10,000-user VM configuration option due to
potential overload on the backup subscriber
Use multiple physical ports in the hardware platform for the virtual machine network traffic, and use
a minimum of two upstream switches to provide network connectivity redundancy. If using the
VMware vSphere virtual switch, use VMware NIC teaming.
Whenever possible, distribute the hardware platforms across multiple physical switches within the
network and across multiple physical locations within the same network to minimize the impact of
a switch failure or the loss of a particular network location.
Deploy SRST or E-SRST on Cisco IOS routers at remote locations to provide fallback call
processing services in the event that these locations lose connectivity to the Unified CM cluster.
Cisco recommends that you leave voice activity detection (VAD) disabled within the Unified CM
cluster. VAD is disabled by default in the Unified CM service parameters, and you should disable it
on H.323 and SIP dial peers configured on Cisco IOS gateways by using the no vad command.
Ensure that the Unified CM nodes are distributed across different physical servers so that backup or
redundant subscriber nodes are on different physical servers than the primary subscriber nodes.
Both UCS B-Series Blade Servers and mid-end or high-end C-Series Rack-Mount Servers support
all Unified CM Open Virtualization Archive (OVA) template sizes (including, for example, the VM
configurations that support 10,000 devices). However, some smaller servers support only smaller
VM configurations. For information on proper VM configuration selection as well as the use of the
Cisco Unified Communications Sizing Tool, see the chapter on Collaboration Solution Sizing
Guidance, page 25-1.
9-25
Chapter 9
Call Processing
Access to the USB and serial ports on the hardware platform is not supported with Unified CM
virtual machines. Therefore, attaching fixed live audio sources for MoH, making a serial SMDI
connection to a legacy voicemail system, or attaching a USB flash drive for writing log files are also
not supported. The following alternative options are available:
For MoH live audio source feed, consider using Cisco IOS-based gateway multicast MoH for
Cisco Business Edition 6000S supports a maximum of 150 users and 300 endpoints with
Unified CM. Unified CM is typically deployed as a single Unified CM publisher node that also
handles call processing. To provide high availability for Unified CM with Cisco Business
Edition 6000S, deploy SRST or add a Unified CM subscriber running on a separate Cisco Business
Edition 6000S platform, for example.
Cisco Business Edition 6000M and Cisco Business Edition 6000H support a maximum of 1,000
users. Unified CM is deployed as a single Unified CM publisher node that also handles call
processing. To provide Unified CM redundancy, additional hardware server(s) hosting Unified CM
subscriber node(s) can be deployed.
Note
More than two servers may be clustered for a BE6000 deployment to provide additional
redundancy and/or geographic distribution; however, the capacity limits are not increased.
For example, the total number of users across the cluster may not exceed 150 with BE6000S,
or 1,000 with BE6000M or BE6000H.
BE6000M supports a maximum of 1,200 endpoints, while BE6000H supports a maximum of 2,500
endpoints. However, actual endpoint capacity depends on total system BHCA, which cannot exceed
a maximum of 5,000. For additional information about Cisco Business Edition capacity, including
sizing examples and per-platform sizing limits, see the chapter on Collaboration Solution Sizing
Guidance, page 25-1.
If multiple Business Edition 6000 servers are required in the same deployment, distribute them
across multiple physical switches.
Use an uninterruptible power supply (UPS) to provide maximum availability, especially if the server
has only one power supply.
When deploying Business Edition 6000 with two servers for high availability, a Unified CM node
should run on each server to provide high availability in case one of the servers fails. Furthermore,
Cisco recommends configuring the Unified CM cluster with the subscriber node as the primary call
processing server and the publisher node as the backup call processing server.
With Cisco Business Edition 7000, Unified CM has the same rules, capacities, and design
considerations as a regular (not part of Cisco Business Edition) Unified CM deployment.
Applications that are not part of the Business Edition 6000 solution and that are running on separate
hardware can be integrated to a Business Edition 6000 deployment, but you must ensure that those
applications do not exceed the Business Edition 6000 capacity limits. For example, the overall
BHCA and the number of contact center agents should not exceed the Business Edition 6000
capacity limit of Unified CM. For more information on the Business Edition 6000 capacity limits,
refer to the latest Cisco Business Edition 6000 Installation Guide. Also ensure that those
applications support the Cisco Collaboration VM configuration provided by Business Edition 6000.
9-26
Chapter 9
Call Processing
Computer Telephony Integration (CTI)
For example, Cisco Unified Contact Center Enterprise requires the Unified CM 7.5k-user or larger
VM configuration, so it cannot be integrated with a Unified CM deployment that is running on
Business Edition 6000.
Cisco Unified CME
Unified CME supports a maximum of 450 endpoints. However, depending on the Cisco IOS router
model, endpoint capacity could be significantly lower. For additional information about
Unified CME platforms and capacities, refer to the Cisco Unified Communications Manager
Express compatibility information available at
http://www.cisco.com/en/US/products/sw/voicesw/ps4625/products_device_support_tables_list.ht
ml.
When possible, dual-attach the Unified CME router to the network using multiple IP interfaces to
provide maximum network availability. Likewise, if multiple instances of Unified CME are
required in the same deployment, distribute them across multiple physical switches or locations.
When possible, deploy the Unified CME router with dual power supplies and/or an uninterruptible
power supply (UPS) in order to provide maximum availability of the platform.
A CTI application that monitors the state of a Cisco IP device is called a monitoring application.
A busy-lamp-field application that displays on-hook/off-hook status or uses that information to
indicate a user's availability in the form of Presence are both examples of third-party CTI
monitoring applications.
Call control applications
Any application that uses Cisco CTI to remotely control a Cisco IP device using out-of-band
signaling is a call control application. Cisco Jabber, when configured to remotely control a
Cisco IP device, is a good example of a call control application.
9-27
Chapter 9
Call Processing
These are any CTI applications that monitor and control a Cisco IP device. Cisco Unified
Contact Center Enterprise is a good example of a combined monitor and control application
because it monitors the status of agents and controls agent phones through the agent desktop.
Note
While the distinction between a monitor, call control, and monitor + control application is called out
here, this granularity is not exposed to the application developer. All CTI applications using Cisco CTI
are enabled for both monitoring and control.
The following devices can be monitored or controlled through CTI:
CTI Port
CTI Remote Device provides the ability for a CTI application to have monitoring and limited call control
capabilities over phones that do not support CTI, such as traditional PSTN phones, mobile phones,
third-party phones, or phones attached to a third-party PBX.
CTI Architecture
Cisco CTI consists of the following components (see Figure 9-9), which interact to enable applications
to take advantage of the telephony feature set available in Cisco Unified CM:
JTAPI and TAPI Two standard interfaces supported by Cisco CTI. Developers can choose to write
applications using their preferred method library.
Unified JTAPI and Unified TSP Client Converts external messages to internal Quick Buffer
Encoding (QBE) messages used by Cisco Unified CM.
Provider A logical representation of a connection between the application and CTI Manager, used
to facilitate communication. The provider sends device and call events to the application while
accepting control instructions that allow the application to control the device remotely.
Publisher and subscriber Cisco Unified Communications Manager (Unified CM) server nodes.
CCM The Cisco CallManager Service (ccm.exe), the telephony processing engine.
CTI Manager (CTIM) A service that runs on one or more Unified CM subscribers operating in
primary/secondary mode and that authenticates and authorizes telephony applications to control
and/or monitor Cisco IP devices.
9-28
Chapter 9
Call Processing
Computer Telephony Integration (CTI)
Figure 9-9
Unified CM cluster
Unified CM
CTI
Manager
Unified CM
QBE
CTI
Manager
Unified CM
CCM
CTI Application
SDL
CTI
Manager
SDL
CCM
271549
Cisco CTI
(JTAPI/TAPI)
CCM
SDL
Cisco CallManager
Service must be running
because standalone
CTI Managers are not
currently supported.
Once an application is authenticated and authorized, the CTIM acts as the broker between the telephony
application and the Cisco CallManager Service. (This service is the call control agent and should not be
confused with the overall product name Cisco Unified Communications Manager.) The CTIM responds
to requests from telephony applications and converts them to Signaling Distribution Layer (SDL)
messages used internally in the Unified CM system. Messages from the Cisco CallManager Service are
also received by the CTIM and directed to the appropriate telephony application for processing.
The CTIM may be activated on any of the Unified CM subscriber nodes in a cluster that have the Cisco
CallManager Service active. This allows up to eight CTIMs to be active within a Unified CM cluster.
Standalone CTIMs are currently not supported.
9-29
Chapter 9
Call Processing
Site 2
Site 1
J/TAPI
80ms RTT
Subscriber
running
CTI Manager
CTI
Application
SIP/SCCP
Subscriber
running
Unified CM
IP
CTI
controlled
device
252943
Figure 9-10
TAPI and JTAPI applications over the WAN (CTI application over the WAN; see Figure 9-11)
In this scenario, the CTI application is on one side of the WAN (Site 1), and its associated
CTI Manager is on the other side (Site 2). In this scenario, it is up to the CTI application developer
or provider to ascertain whether or not their application can accommodate the RTT as implemented.
In some cases failover and failback times might be higher than if the application is co-located with
its CTI Manager. In those cases, the application developer or provider should provide guidance as
to the behavior of their application under these conditions.
JTAPI Over the WAN
Site 2
Site 1
SIP/SCCP
J/TAPI
CTI
Application
Note
Maximum RTT is
application
dependent and
based on testing.
Subscriber
running
CTI Manager
and Unified CM
IP
CTI
controlled
device
253856
Figure 9-11
Support for TAPI and JTAPI over the WAN is application dependent. Both customers and
application developers or providers should ensure that their applications are compatible with any
such deployment involving clustering over the WAN.
9-30
Chapter 9
Call Processing
Computer Telephony Integration (CTI)
CTI Manager
CTI Manager must be enabled on at least one and possibly all call processing subscribers within the
Unified CM cluster. The client-side interfaces (TAPI TSP or JTAPI client) allow for two IP addresses
each, which then point to Unified CM server nodes running the CTIM service. For CTI application
redundancy, Cisco recommends having the CTIM service activated on at least two Unified CM server
nodes in a cluster, as shown in Figure 9-12.
Unified CM cluster
Application
Server
(JTAPI/CTI)
CTIM1
IP
M
IP
CTIM2
Redundant CTI Managers, but no load
balancing with one application server.
Unified CM cluster
CTIM1
CTIM2
IP
IP
CTIM3
CTIM4
252945
Application
Servers
(JTAPI/CTI)
9-31
Chapter 9
Call Processing
Figure 9-13 shows an example of this type of configuration for Cisco Unified Contact Center Enterprise
(Unified CCE). This type of configuration has the following characteristics:
Figure 9-13
Peripheral
Gateway
A
Peripheral
Gateway
B
JTAPI
Plug-in
JTAPI
Plug-in
CTI Manager
CTI Manager
Unified CM Server
Unified CM Server
271551
9-32
Chapter 9
Call Processing
Computer Telephony Integration (CTI)
Figure 9-14 shows an example of this type of configuration for Cisco Unified Contact Center Express
(Unified CCX). This type of configuration has the following characteristics:
Unified CCX has two IP addresses configured, one for each CTI Manager.
If connection to the primary CTI Manager is lost, Unified CCX fails-over to its secondary
CTI Manager.
Figure 9-14
JTAPI Plug-in
Secondary
CTI Manager
Unified CM Server
Unified CM Server
271552
Primary
CTI Manager
Implementation
For guidance and support on writing applications, application developers should consult the Cisco
Developer Connection, located at
http://developer.cisco.com/web/cdc/community
9-33
Chapter 9
Call Processing
Unified CM and Unified CME Interoperability via SIP in a Multisite Deployment with Distributed
Call Processing, page 9-36
Cisco Unified CM and Cisco Unified Communications Manager Express (Unified CME) could also be
integrated using H.323, but this section does not cover this integration in detail. For more information
on the H.323 integration, refer to the Cisco Collaboration 9.x SRND, available at
http://www.cisco.com/go/ucsrnd
9-34
Chapter 9
Call Processing
Integration of Multiple Call Processing Agents
Unified CME hairpins the call legs from Unified CM for the VoIP calls across SIP or H.323 trunks when
needed. For more information on allowing auto-detection on a non-H.450 supported Unified CM
network and for enabling or disabling supplementary services for H450.2, H450.3, or SIP, refer to the
Unified CME product documentation available at
http://www.cisco.com/en/US/products/sw/voicesw/ps4625/tsd_products_support_series_home.html
When connected to Unified CM via SIP trunks, Unified CME does not auto-detect Unified CM calls. By
default, Unified CME always tries to redirect calls using either a SIP Refer message for call transfer or
a SIP 302 Moved Temporarily message for call forward; if that fails, Unified CME will then try to
hairpin the call.
Music on Hold
While Unified CM can be enabled to stream MoH in both G.711 and G.729 formats, Unified CME
streams MoH only in G.711 format. Therefore, when Unified CME controls the MoH audio on a call
placed on hold, it requires a transcoder to transcode between a G.711 MoH stream and a G.729 call leg.
9-35
Chapter 9
Call Processing
Multisite Deployment with Unified CM and Unified CME Using SIP Trunks
Unified CM Site
Voicemail
Unified CM
Express
MTP/SDP
V
Unified CM
SIP
IP
SCCP
IP
SCCP
Directory
Server
IP
SIP
SCCP
SIP
IP
PSTN
SCCP
IP
IP
SIP
SIP
WAN
IP
IP
SIP
SCCP
Unified CM
Express
IP
SIP
IP
SCCP
IP
SCCP
SIP
141867
SCCP
Best Practices
Follow these guidelines and best practices when using the deployment model illustrated in Figure 9-15:
Configure a SIP Trunk Security Profile with Accept Replaces Header selected.
Configure a SIP trunk on Unified CM using the SIP Trunk Security Profile created, and also specify
a ReRouting CSS. The ReRouting CSS is used to determine where a SIP user (transferor) can refer
another user (transferee) to a third user (transfer target) and which features a SIP user can invoke
using the SIP 302 Redirection Response and INVITE with Replaces.
For SIP trunks there is no need to enable the use of media termination points (MTPs) when using
SCCP endpoints on Unified CME. However, SIP endpoints on Unified CME require the use of
media termination points on Unified CM to be able to handle delayed offer/answer exchanges with
the SIP protocol (that is, the reception of INVITEs with no Session Description Protocol).
9-36
Chapter 9
Call Processing
Integration of Multiple Call Processing Agents
Route calls to Unified CME via a SIP trunk using the Unified CM dial plan configuration (route
patterns, route lists, and route groups).
Use Unified CM device pools and regions to configure a G.711 codec within the site and the G.729
codec for remote Unified CME sites.
Configure the allow-connections sip to sip command under voice services voip on Unified CME
to allow SIP-to-SIP call connections.
For SIP endpoints, configure the mode cme command under voice register global, and configure
dtmf-relay rtp-nte under the voice register pool commands for each SIP phone on Unified CME.
For SCCP endpoints, configure the transfer-system full-consult command and the
transfer-pattern .T command under telephony-service on Unified CME.
Configure the SIP WAN interface voip dial-peers to forward or redirect calls, destined for
Unified CM, with session protocol sipv2 and dtmf-relay [sip-notify | rtp-nte] on Unified CME.
Design Considerations
This section first covers some characteristics and design considerations for Unified CM and
Unified CME interoperability via SIP in some main areas such as supplementary services for call
transfer and forward, presence service for busy lamp field (BLF) notification for speed-dial buttons and
directory call lists, and out-of-dialog (OOD-Refer) for integration with partner applications and
third-party phone control for click-to-dial between the Unified CM phones and Unified CME phones.
The section also covers some general design considerations for Unified CM and Unified CME
interoperability via SIP.
Supplementary Services
SIP Refer or SIP 302 Moved Temporarily messages can be used for supplementary services such as call
transfer or call forward on Unified CME or Unified CM to instruct the transferee (referee) or phone
being forwarded (forwardee) to initiate a new call to the transfer-to (refer-to) target or forward-to target.
No hairpinning is needed for call transfer or call forward scenarios when the SIP Refer or SIP 302 Moved
Temporarily message is supported.
However, supplementary-service must be disabled if there are certain extensions that have no DID
mapping or if Unified CM or Unified CME does not have a dial plan to route the call to the DID in the
SIP 302 Moved Temporarily message. When supplementary-service is disabled, Unified CME hairpins
the calls or sends a re-invite SIP message to Unified CM to replace the media path to the new called party
ID. Both signaling and media are hairpinned, even when multiple Unified CMEs are involved for further
call forwards. The supplementary-service can also be disabled for transferred calls. In this case, the SIP
Refer message will not be sent to Unified CM, but the transferee (referee) party and transfer-to party
(refer-to target) are hairpinned.
Note
9-37
Chapter 9
Call Processing
The following examples illustrate the call flows when supplementary services are disabled:
Unified CM phone B calls Unified CME phone A, which is set to call-forward (all, busy, or no
answer) to phone C (either a Unified CM phone, a Unified CME phone on the same or different
Unified CME, or a PSTN phone).
Unified CME does not send the SIP 302 Moved Temporarily message to Unified CM, but hairpins
the call between Unified CM phone B and phone C.
Unified CM phone B calls Unified CME phone A, which transfer the call to phone C (either a
Unified CM phone, a Unified CME phone, or a PSTN phone).
Unified CME does not send the SIP Refer message to Unified CM, but hairpins the call between
Unified CM phone B and phone C.
General Design Considerations for Unified CM and Unified CME Interoperability via SIP
Disable supplementary-service if SIP 302 Moved Temporarily or SIP Refer messages are not
supported by Unified CM, otherwise Unified CM cannot route the call to the transfer-to or
forward-to target.
In a SIP-to-SIP call scenario, a Refer message is sent by default from the transferor to the transferee,
the transferee sets up a new call to the transfer-to target, and the transferor hears ringback tone by
default while waiting for the transfer at connect. If supplementary-service is disabled on
Unified CME, Unified CME will provide in-band ringback tone right after the call between the
transferee and transfer-to target is connected.
Presence service is supported on Unified CM and Unified CME via SIP trunk only.
The OOD-Refer feature allows third-party applications to connect two endpoints on Unified CM or
Unified CME through the use of the SIP REFER method. Consider the following factors when using
OOD-Refer:
Both Unified CM and Unified CME must be configured to enable the OOD-Refer feature.
Call Hold, Transfer, and Conference are not supported during an OOD-Refer transaction, but
Note
Control signaling in TLS is supported, but SRTP is not supported over the SIP trunk.
SRTP over a SIP trunk is a gateway feature in Cisco IOS for Unified CM. SRTP support is not
available with Unified CM and Unified CME interworking via SIP trunks.
When multiple PSTN connections exist (one for Unified CM and one for Unified CME), fully attended
transfer between a Unified CM endpoint and a Unified CME endpoint to a PSTN endpoint will fail. The
recommendation is to use blind transfer when using multiple PSTN connections, and it is configured
under telephony-service as transfer-system full-blind.
9-38
CH A P T E R
10
This chapter describes the deployment models for Cisco Unified Communications Systems.
Earlier versions of this chapter based the deployment models discussion exclusively on the call
processing deployment models for Cisco Unified Communications Manager (Unified CM). The current
version of this chapter offers design guidance for the entire Cisco Unified Communications and
Collaboration System, which includes much more than just the call processing service.
For design guidance with earlier releases of Cisco Unified Communications, refer to the Cisco Unified
Communications Solution Reference Network Design (SRND) documentation available at
http://www.cisco.com/go/ucsrnd
Described in
Revision Date
Removed information about Cisco Unified Border VPN-less Enterprise Access, page 10-34
Element VPN-less IP phone access
10-1
Chapter 10
flexibility and scale to meet the demands of a rapidly changing and expanding Unified Communications
environment that will become more URI-centric as users with multiple Unified Communications devices
wish to be identified by a single user name irrespective of the form of communication.
Collaboration as a Service
As Unified Communications has become more commonplace, the deployment options for Cisco Unified
Communications and Collaboration have increased to address the demand for "collaboration as a
service" as well as traditional on-premises Unified Communications and Collaboration deployments.
The main focus of this chapter is to provide the reader with design guidance for on-premises
Collaboration deployments, but it also includes a description of systems such as Cisco Hosted
Collaboration Solution (HCS) and Cisco WebEx, which can be deployed as managed cloud-based
Collaboration services. The choice of using an on-premises, cloud-based, or hybrid solution may be
determined by many factors. For example, cloud-based solutions require less on-site expertise but might
lack the deployment flexibility that many enterprises need. Hybrid designs can also be deployed, where
some Unified Communications functions such as call control are provided on-premises and others are
provided as a cloud-based service.
Cisco offers the following "collaboration as a service" products that can be used to augment or replace
functionality provided in an on-premises Unified Communications deployment:
WebEx Meetings, WebEx Telepresence and WebEx Messenger can be deployed as WebEx cloud services
or as on-premises services using Cisco Unified CM, Cisco WebEx Meetings Server, Cisco Expressway,
and Cisco IM and Presence.
For more information on Cisco Webex, refer to the documentation at
http://www.cisco.com/en/US/products/ps10352/index.html
Cisco Hosted Collaboration Solution provides the following range of applications and services:
Cisco Unified Communications Manager (call control for voice and video)
For more information on Cisco Hosted Collaboration Solution, refer to the documentation available at
http://www.cisco.com/en/US/products/ps11363/index.html
10-2
Chapter 10
Note
Unless otherwise specified, all information contained within this SRND that relates to call processing
deployments (including capacity, high availability, and general design considerations) applies only to a
standard cluster with up to eight call processing subscriber nodes.
There are an infinite number of variations on these three basic deployment models, such as deployments
with centralized or distributed PSTN access and services, but the basic design guidance provided in this
chapter still applies to the majority of them.
10-3
Chapter 10
Failover redundancy
For services that are considered essential, redundant elements should be deployed so that no single
point of failure is present in the design. The redundancy between the two (or more) elements is
automated. For example, the clustering technology used in Cisco Unified Communications Manager
(Unified CM) allows for up to three servers to provide backup for each other. This type of
redundancy may cross technological boundaries. For example, a phone may have as its first three
preferred call control agents, three separate Unified CM servers belonging to the same call
processing cluster. As a fourth choice, the phone can also be configured to rely on a Cisco IOS router
for call processing services.
Redundant links
In some instances, it is advantageous to deploy redundant IP links, such as IP WAN links, to guard
against the failure of a single WAN link.
Geographical diversity
Some products support the distribution of redundant service nodes across WAN links so that, if an
entire site is off-line (such as would be the case during an extended power outage exceeding the
capabilities of provisioned UPS and generator backup systems), another site in a different location
can ensure business continuance.
In this context, size generally refers to the number of users, which translates into a quantity of IP
telephones, voice mail boxes, presence watchers, and so forth. Size also can be considered in terms of
processing capacity for sites where few (or no) users are present, such as data centers.
Network Connectivity
The site's connectivity into the rest of the system has three main components driving the design:
Latency
Reliability
10-4
Chapter 10
These components are often considered adequate in the Local Area Network (LAN): QoS is achievable
with all LAN equipment, bandwidth is typically in the Gigabit range, latency is minimal (in the order of
a few milliseconds), and excellent reliability is the norm.
The Metropolitan Area Network (MAN) often approaches the LAN in all three dimensions: bandwidth
is still typically in the multiple Megabit range, latency is typically in the low tens of milliseconds, and
excellent reliability is common. Packet treatment policies are generally available from MAN providers,
so that end-to-end QoS is achievable.
The Wide Area Network (WAN) generally requires extra attention to these components: the bandwidth
is at a cost premium, the latencies may depend not only on effective serialization speeds but also on
actual transmission delays related to physical distance, and the reliability can be impacted by a multitude
of factors. The QoS performance can also require extra operational costs and configuration effort.
Bandwidth has great influence on the types of Unified Communications services available at a site, and
on the way these services are provided. For example, if a site serving 20 users is connected with
1.5 Mbps of bandwidth to the rest of the system, the site's voice, presence, instant messaging, email, and
video services can readily be hosted at a remote datacenter site. If that same site is hosting 1000 users,
some of the services would best be hosted locally to avoid saturating the comparatively limited
bandwidth with signaling and media flows. Another alternative is to consider increasing the bandwidth
to allow services to be delivered across the WAN from a remote datacenter site.
The influence of latency on design varies, based on the type of Unified Communications service
considered for remote deployment. If a voice service is hosted across a WAN where the one-way latency
is 200 ms, for example, users might experience issues such as delay-to-dialtone or increased media
cut-through delays. For other services such as presence, there might be no problem with a 200 ms
latency.
Reliability of the site's connectivity into the rest of the network is a fundamental consideration in
determining the appropriate deployment model for any technology. When reliability is high, most
Unified Communications components allow for the deployment of services hosted from a remote site;
when reliability is inconsistent, some Unified Communications components might not perform reliably
when hosted remotely; if the reliability is poor, co-location of the Unified Communications services at
the site might be required.
High Availability Requirements
The high availability of services is always a design goal. Pragmatic design decisions are required when
balancing the need for reliability and the cost of achieving it. The following elements all affect a design's
ability to deliver high availability:
Bandwidth reliability, directly affecting the deployment model for any Unified Communications
service
Power availability
Power loss is a very disruptive event in any system, not only because it prevents the consumption of
services while the power is out, but also because of the ripple effects caused by power restoration.
A site with highly available power (for example, a site whose power grid connection is stable,
backed-up by uninterruptible power supplies (UPSs) and by generator power) can typically be
chosen to host any Unified Communications service. If a site has inconsistent power availability, it
would not be judicious to use it as a hosting site.
10-5
Chapter 10
Centralized Services
For applications where enterprise branch sites are geographically dispersed and interconnected over a
Wide Area Network, the Cisco Unified Communications services can be deployed at a central location
while serving endpoints over the WAN connections. For example, the call processing service can be
deployed in a centralized manner, requiring only IP connectivity with the remote sites to deliver
telephony services. Likewise, voice messaging services, such as those provided by the Cisco Unity
Connection platform, can also be provisioned centrally to deliver services to endpoints remotely
connected across an IP WAN.
Centrally provisioned Unified Communications services can be impacted by WAN connectivity
interruptions; for each service, the available local survivability options should be planned. As an
example, the call processing service as offered by Cisco Unified CM can be configured with local
survivability functionality such as Survivable Remote Site Telephony (SRST) or Enhanced SRST.
Likewise, a centralized voice messaging service such as that of Cisco Unity Connection can be
provisioned to allow remote sites operating under SRST to access local voicemail services using Unity
Connection Survivable Remote Site Voicemail (SRSV).
The centralization of services need not be uniform across all Unified Communications services. For
example, a system can be deployed where multiple sites rely on a centralized call processing service, but
can also be provisioned with a de-centralized (distributed) voice messaging service such as Cisco Unity
Express. Likewise, a Unified Communications system could be deployed where call processing is
provisioned locally at each site through Cisco Unified Communications Manager Express, with a
centralized voice messaging service such as Cisco Unity Connection.
In many cases, the main criteria driving the design for each service are the availability and quality of the
IP network between sites. The centralization of Unified Communications services offers advantages of
economy of scale in both capital and operational expenses associated with the hosting and operation of
equipment in situations where the IP connectivity between sites offers the following characteristics:
Enough bandwidth for the anticipated traffic load, including peak hour access loads such as those
generated by access to voicemail, access to centralized PSTN connectivity, and inter-site on-net
communications including voice and video
High availability, where the WAN service provider adheres to a Service Level Agreement to
maintain and restore connectivity promptly
Low latency, where local events at the remote site will not suffer if the round-trip time to the main
central site imparts some delays to the system's response times
10-6
Chapter 10
Also, when a given service is deployed centrally to serve endpoints at multiple sites, there are often
advantages of feature transparency afforded by the use of the same processing resources for users at
multiple sites. For example, when two sites are served by the same centralized Cisco Unified
Communications Manager cluster, the users can share line appearances between the two sites. This
benefit would not be available if each site were served by different (distributed) call processing systems.
These advantages of feature transparency and economies of scale should be evaluated against the relative
cost of establishing and operating a WAN network configured to accommodate the demands of Unified
Communications traffic.
Distributed Services
Unified Communications services can also be deployed independently over multiple sites, in a
distributed fashion. For example, two sites (or more) can be provisioned with independent call
processing Cisco Unified CME nodes, with no reliance on the WAN for availability of service to their
co-located endpoints. Likewise, sites can be provisioned with independent voice messaging systems
such as Cisco Unity Express.
The main advantage of distributing Unified Communications services lies in the independence of the
deployment approach from the relative availability and cost of WAN connectivity. For example, if a
company operates a site in a remote location where WAN connectivity is not available, is very expensive,
or is not reliable, then provisioning an independent call processing node such as Cisco Unified
Communications Manager Express within the remote site will avoid any call processing interruptions if
the WAN goes down.
Inter-Networking of Services
If two sites are provisioned with independent services, they can still be interconnected to achieve some
degree of inter-site feature transparency. For example, a distributed call processing service provisioned
through Cisco Unified Communications Manager Express can be inter-networked through SIP or H.323
trunks to permit IP calls between the sites. Likewise, separate instances of Cisco Unity Connection or
Cisco Unity Express can partake in the same messaging network to achieve the routing of messages and
the exchange of subscriber and directory information within a unified messaging network.
10-7
Chapter 10
Table 10-2
Service
Centralized
Distributed
Inter-Networked
Geographical
Diversity
Yes
Yes
Yes
Yes
No
Yes
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Cisco Expressway
Yes
Yes
Yes
Yes
Enterprise Edition
Because call processing is a fundamental service, the basic call processing deployment models are
introduced in this chapter. For a detailed technical discussion on Cisco Unified Communications
Manager call processing, refer to the chapter on Call Processing, page 9-1.
Campus Deployments
In this call processing deployment model, the Unified Communications services and the endpoints are
co-located in the campus, and the QoS-enabled network between the service nodes, the endpoints, and
applications is considered highly available, offering bandwidth in the gigabit range with less than 15 ms
of latency end-to-end. Likewise, the quality and availability of power are very high, and services are
hosted in an appropriate data center environment. Communications between the endpoints traverses a
LAN or a MAN, and communications outside the enterprise goes over an external network such as the
PSTN. An enterprise would typically deploy the campus model over a single building or over a group of
buildings connected by a LAN or MAN. (See Figure 10-1.)
10-8
Chapter 10
Figure 10-1
Monitoring/Scheduling
Applications
PSTN/ISDN
Unified CM
Media Resources
Cisco
Expressway-E
Conferencing Resources
Internet
348618
Cisco
Expressway-C
Campus
Single Cisco Unified CM cluster (Enterprise or Business Edition 7000). Some campus call
processing deployments may require more than one Unified CM cluster, for instance, if scale calls
for more endpoints than can be serviced by a single cluster or if a cluster needs to be dedicated to
an application such as a call center.
Alternatively for smaller deployments, Cisco Business Edition 6000 may be deployed in the
campus.
Maximum of 40,000 configured and registered Skinny Client Control Protocol (SCCP) or Session
Initiation Protocol (SIP) IP phones, softphones, analog ports, video endpoints, SIP-based
TelePresence endpoints and room-based TelePresence conferencing systems, mobile clients, and
Cisco Virtualization Experience Clients (VXC) per Unified CM cluster.
Maximum of 2,100 gateways and trunks (that is, the total number of H.323 gateways, H.323 trunks,
digital MGCP devices, and SIP trunks) per Unified CM cluster.
Trunks and/or gateways (IP or PSTN) for all calls to destinations outside the campus.
Multipoint control unit (MCU) resources are required for multipoint conferencing.
Co-located digital signal processor (DSP) resources for conferencing, transcoding, and media
termination point (MTP).
Other Unified Communications services, such as messaging (voicemail), presence, and mobility are
typically co-located.
10-9
Chapter 10
Interfaces to legacy voice services such as PBXs and voicemail systems are connected within the
campus, with no operational costs associated with bandwidth or connectivity.
SIP-based video ISDN gateways are needed to communicate with videoconferencing devices on the
public ISDN network.
Cisco Expressway-C and Cisco Expressway-E provide a collaboration edge function that enables
secure business-to-business telepresence and video communications, and enterprise access for
remote and mobile workers over the internet.
Cisco TelePresence Video Communication Server (VCS) may also be used to register legacy H.323
and third-party telepresence endpoints. However, to avoid the dial plan and call admission control
complexities that dual call control introduces (see Design Considerations for Dual Call Control
Deployments, page 10-38), Cisco recommends using SIP to register all TelePresence endpoints and
room-based TelePresence conferencing systems with Cisco Unified Communications Manager
High-bandwidth audio is available (for example, G.711 or G.722) between devices within the site.
High-bandwidth video (for example, 1.5 Mbps with 4CIF or 720p, to 2 Mbps with 1080p) is
available between devices within the site.
Ensure that the infrastructure is highly available, enabled for QoS, and configured to offer resiliency,
fast convergence, and inline power.
Know the calling patterns for your enterprise. Use the campus model if most of the calls from your
enterprise are within the same site or to PSTN users outside your enterprise.
Use G.711 codecs for all endpoints. This practice eliminates the consumption of digital signal
processor (DSP) resources for transcoding, and those resources can be allocated to other functions
such as conferencing and media termination points (MTPs).
Implement the recommended network infrastructure for high availability, connectivity options for
phones (in-line power), Quality of Service (QoS) mechanisms, and security. (See Network
Infrastructure, page 3-1.)
Follow the provisioning recommendations listed in the chapter on Call Processing, page 9-1.
10-10
Chapter 10
mobility are often hosted at the central site as well to reduce the overall costs of administration and
maintenance. In situations where the availability of the WAN is unreliable or when WAN bandwidth
costs are high, it is possible to consider decentralizing some Unified Communications services such as
voice messaging (voicemail) so that the service's availability is not impacted by WAN outages.
Note
Figure 10-2
In each solution for the centralized call processing model presented in this document, the various sites
connect to an IP WAN with QoS enabled.
Remote
Office
Central Site
Monitoring and Scheduling
Applications
PSTN/ISDN
Unified CM
Media Resources
IP WAN
Cisco
E
Expressway-E
Conferencing Resources
Internet
348619
Cisco
Expressway-C
The multisite model with centralized call processing has the following design characteristics:
Single Unified CM cluster (Enterprise or Business Edition 7000). Some centralized call processing
deployments may require more than one Unified CM cluster, for instance, if scale calls for more
endpoints than can be serviced by a single cluster or if a cluster needs to be dedicated to an
application such as a call center.
Cisco Business Edition 6000 may be deployed in centralized call processing configurations for up
to 49 remote sites.
Maximum of 40,000 configured and registered Skinny Client Control Protocol (SCCP) or Session
Initiation Protocol (SIP) IP phones, softphones, analog ports, video endpoints, SIP-based
TelePresence endpoints and room-based TelePresence conferencing systems, mobile clients, and
Cisco Virtualization Experience Clients (VXC) per Unified CM cluster.
10-11
Chapter 10
Maximum of 2,100 gateways and trunks (that is, the total number of H.323 gateways, H.323 trunks,
digital MGCP devices, and SIP trunks) per Unified CM cluster.
Digital signal processor (DSP) resources for conferencing, transcoding, and media termination point
(MTP) are distributed locally to each site to reduce WAN bandwidth consumption on calls requiring
DSPs.
Multipoint control unit (MCU) resources are required for multipoint conferencing. These resources
may all be located at the central site or may be distributed to the remote sites if local conferencing
resources are required.
Capability to integrate with legacy private branch exchange (PBX) and voicemail systems.
Connections to legacy voice services such as PBXs and voicemail systems can be made within the
central site, with no operational costs associated with bandwidth or connectivity. Connectivity to
legacy systems located at remote sites may require the operational expenses associated with the
provisioning of extra WAN bandwidth.
SIP-based video ISDN gateways are needed to communicate with videoconferencing devices on the
public ISDN network. ISDN video gateways can be centralized and/or deployed at each remote site.
Cisco Expressway-C and Cisco Expressway-E provide a collaboration edge function that enables
secure business-to-business telepresence and video communications, and VPN-less enterprise
access for remote and mobile workers over the internet.
Cisco TelePresence Video Communication Server (VCS) may also be used to register legacy H.323
and third-party telepresence endpoints. However, to avoid the dial plan and call admission control
complexities that dual call control introduces (see Design Considerations for Dual Call Control
Deployments, page 10-38), Cisco recommends using SIP to register all TelePresence endpoints and
room-based TelePresence conferencing systems with Cisco Unified Communications Manager
The system allows for the automated selection of high-bandwidth audio (for example, G.711 or
G.722) between devices within the site, while selecting low-bandwidth audio (for example, G.729)
between devices in different sites.
The system allows for the automated selection of high-bandwidth video (for example, 1.5 Mbps with
4CIF or 720p, to 2 Mbps with 1080p) between devices in the same site, and low-bandwidth video
(for example, 384 kbps with 448p or CIF) between devices at different sites.
A minimum of 1.5 Mbps or greater WAN link speed should be used when video is to be placed on
the WAN.
For voice and video calls, automated alternate routing (AAR) provides the automated rerouting of
calls through the PSTN when call admission control denies a call between endpoints within a cluster
due to lack of bandwidth. AAR relies on a gateway being available to route the call from the calling
phone toward the PSTN, and another gateway to accept the call from the PSTN at the remote site,
to be connected to the called phone.
Call Forward Unregistered (CFUR) functionality provides the automated rerouting of calls through
the PSTN when an endpoint is considered unregistered due to a remote WAN link failure. CFUR
relies on a gateway being available to route the call from the calling phone toward the PSTN, and
another gateway to accept the call from the PSTN at the remote site, to be connected to the called
phone.
10-12
Chapter 10
Survivable Remote Site Telephony (SRST) for video. Video endpoints located at remote sites
become audio-only devices if the WAN connection fails. Starting with Cisco IOS release 15.3(3)M,
using phone load firmware 9.4.1 or later, Enhanced SRST enables video survivability on SIP video
endpoints (Cisco Unified IP Phone 9900, for example) during WAN failure. For SRST video support
with a particular phone model, refer to the respective Cisco Unified IP Phone Administration Guide
available at http://www.cisco.com.
Cisco Unified Communications Manager Express (Unified CME) may be used for remote site
survivability (Enhanced SRST) instead of an SRST router.
Cisco Unified Communications Manager Express (Unified CME) can be integrated with the Cisco
Unity Connection server in the branch office or remote site. The Cisco Unity Connection server is
registered to the Unified CM at the central site in normal mode and can fall back to Enhanced SRST
mode when Unified CM is not reachable, or during a WAN outage, to provide the users at the branch
offices with access to their voicemail with MWI.
With multisite centralized call processing model, PSTN routing through both central and remote site
gateways is supported. Providing a local gateway at a remote site for local PSTN breakout might be
a requirement for countries that provide emergency services for users located at remote sites. In this
case, the local gateway at the remote site provides call routing to the local PSAP for emergency calls.
Local PSTN breakout at remote sites might also be required for countries having strict regulations
that require the separation of the IP telephony network from the PSTN. Where regulations allow,
local PSTN breakout through the remote site gateway can be used to enable toll bypass or tail-end
hop off (TEHO).
Leased lines
Frame Relay
Routers that reside at the WAN edges require quality of service (QoS) mechanisms, such as priority
queuing and traffic shaping, to protect the voice and video traffic from the data traffic across the WAN,
where bandwidth is typically scarce. In addition, a call admission control scheme is needed to avoid
oversubscribing the WAN links with voice and/or video traffic and deteriorating the quality of
established calls. For centralized call processing deployments, Enhanced Location CAC or
RSVP-enabled locations configured within Unified CM provide call admission control (CAC). (Refer to
the chapter on Bandwidth Management, page 13-1, for more information on locations.)
A variety of Cisco gateways can provide the remote sites with TDM and/or IP-based PSTN access. When
the IP WAN is down, or if all the available bandwidth on the IP WAN has been consumed, calls from
users at remote sites can be rerouted through the PSTN. The Cisco Unified Survivable Remote Site
Telephony (SRST) feature, available for both SCCP and SIP phones, provides call processing at the
branch offices for Cisco Unified IP Phones if they lose their connection to the remote primary, secondary,
or tertiary Unified CM or if the WAN connection is down. Cisco Unified SRST is available on Cisco IOS
gateways or on Cisco Unified CME running Enhanced SRST. Unified CME running Enhanced SRST
provides more features for the phones than SRST on a Cisco IOS gateway.
10-13
Chapter 10
Minimize delay between Unified CM and remote locations to reduce voice cut-through delays (also
known as clipping).
Configure Enhanced Locations CAC in Unified CM to provide call admission control into and out
of remote branches. See the chapter on Bandwidth Management, page 13-1, for details on how to
apply this mechanism to the various WAN topologies.
The number of IP phones and line appearances supported in Survivable Remote Site Telephony
(SRST) mode at each remote site depends on the branch router platform, the amount of memory
installed, and the Cisco IOS release. SRST on a Cisco IOS gateway supports up to 1,500 phones,
while Unified CME running Enhanced SRST supports 450 phones. (For the latest SRST or
Unified CME platform and code specifications, refer to the SRST and Unified CME documentation
available at http://www.cisco.com.) Generally speaking, however, the choice of whether to adopt a
centralized call processing or distributed call processing approach for a given site depends on a
number of factors such as:
IP WAN bandwidth or delay limitations
Criticality of the voice network
Feature set needs
Scalability
Ease of management
Cost
If a distributed call processing model is deemed more suitable for the customer's business needs, the
choices include installing a Unified CM cluster at each site or running Unified CME at the remote
sites.
At the remote sites, use the following features to ensure call processing survivability in the event of
a WAN failure:
For SCCP phones, use SRST or Enhanced SRST.
For SIP phones, use SIP SRST or Enhanced SRST.
For deployments with centralized voicemail, use Survivable Remote Site Voicemail (SRSV).
SRST, Enhanced SRST, SIP SRST, SRSV, and MGCP Gateway Fallback can reside with each other
on the same Cisco IOS gateway.
10-14
Chapter 10
Table 10-3
Strategy
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
The first two solutions listed in Table 10-3 provide high availability at the network infrastructure layer
by adding redundancy to the IP WAN access points, thus maintaining IP connectivity between the remote
IP phones and the centralized Unified CM at all times. These solutions apply to both data and voice
services, and are entirely transparent to the call processing layer. The options range from adding a
redundant IP WAN link at the branch router to adding a second branch router platform with a redundant
IP WAN link.
The third and forth solutions in Table 10-3 use an ISDN backup link to provide survivability during
WAN failures. The two deployment options for ISDN backup are:
10-15
Chapter 10
Figure 10-3
Unified CM
PSTN/ISDN
Media
Central Site
Signaling
Remote Office
348620
IP WAN
Under normal operations shown in Figure 10-3, the remote office connects to the central site via an IP
WAN, which carries data traffic, voice traffic, and call signaling. The IP phones at the remote office
exchange call signaling information with the Unified CM cluster at the central site and place their calls
across the IP WAN. The remote office router or gateway forwards both types of traffic (call signaling
and voice) transparently and has no knowledge of the IP phones.
If the WAN link to the remote office fails, as shown in Figure 10-4, or if some other event causes loss of
connectivity to the Unified CM cluster, the remote office IP phones re-register with the remote office
router in SRST mode. The remote office router, using SRST or Enhanced SRST, queries the IP phones
for their configuration and uses this information to build its own configuration automatically. The remote
office IP phones can then make and receive calls either within the remote office network or through the
PSTN. The phone displays the message Unified CM fallback mode, and some advanced Unified CM
features are unavailable and are grayed out on the phone display.
10-16
Chapter 10
Figure 10-4
Unified CM
PSTN/ISDN
Media
Central Site
Signaling
Remote Office
348621
IP
IPWAN
WAN
When WAN connectivity to the central site is reestablished, the remote office IP phones automatically
re-register with the Unified CM cluster and resume normal operation. The remote office SRST router
deletes its information about the IP phones and reverts to its standard routing or gateway configuration.
Routers using Enhanced SRST at the remote office can choose to save the learned phone and line
configuration to the running configuration on the Unified CME router by using the auto-provision
option. If auto-provision none is configured, none of the auto-provisioned phone or line configuration
information is written to the running configuration of the Unified CME router. Hence, no configuration
change is required on Unified CME if the IP phone is replaced and the MAC address changes.
Note
When WAN connectivity to the central site is reestablished, or when Unified CM is reachable again,
phones in SRST mode with active calls will not immediately re-register to Unified CM until those active
calls are terminated.
Enhanced SRST
Enhanced SRST provides more call processing features for the IP phones than are available with the
SRST feature on a router. In addition to the SRST features such as call preservation, auto-provisioning,
and failover, Enhanced SRST also provides most of the Unified CME telephony features for phones,
including:
Paging
Conferencing
Hunt groups
10-17
Chapter 10
Cisco IP Communicator
Integration with Cisco Unity with MWI support at remote sites, with distributed Microsoft
Exchange or IBM Lotus Domino server
Enhanced SRST provides call processing support for SCCP and SIP phones in case of a WAN failure.
However, Enhanced SRST does not provide fallback support for MGCP phones or endpoints. To enable
MGCP phones to fall back if they lose their connection to Unified CM, or if the WAN connection fails,
you can additionally configure the MGCP Gateway Fallback feature on the same Unified CME server
running as the SRST fallback server.
Use the Unified CME IP address as the IP address for SRST reference in the Unified CM
configuration.
The Connection Monitor Duration is a timer that specifies how long phones monitor the WAN link
before initiating a fallback from SRST to Unified CM. The default setting of 120 seconds should be
used in most cases. However, to prevent phones in SRST mode from falling back and re-homing to
Unified CM with flapping links, you can set the Connection Monitor Duration parameter on
Unified CM to a longer period so that phones do not keep registering back and forth between the
SRST router and Unified CM. Do not set the value to an extensively longer period because this will
prevent the phones from falling back from SRST to Unified CM for a long amount of time.
Phones in SRST fallback mode will not re-home to Unified CM when they are in active state.
Phones in SRST fallback mode revert to non-secure mode from secure conferencing.
Configure auto-provision none to prevent writing any learned ephone-dn or ephone configuration
to the running configuration of the Unified CME router. This eliminates the need to change the
configuration if the IP phone is replaced or the MAC address changes.
For more information on Enhanced SRST, refer to the Cisco Unified Communications Manager Express
System Administrator Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps4625/products_installation_and_configuratio
n_guides_list.html
For more information on MGCP Gateway fallback, refer to the information on MGCP gateway fallback
in the Cisco Unified Communications Manager and Interoperability Configuration Guide, Cisco IOS
Release 15M&T, available at
http://www.cisco.com/en/US/docs/ios-xml/ios/voice/cminterop/configuration/15-mt/cminterop-15mt-book.html
10-18
Chapter 10
For supporting a maximum of 1,500 phones on a single SRST router. (Enhanced SRST supports a
maximum of 450 phones.)
For up to 3,000 phones, use two SRST routers. Dial plans must be properly configured to route the
calls back and forth between the SRST routers.
For SRTP media encryption, which is available only in Cisco Unified SRST (Secure SRST).
For routing calls to and from phones that are unreachable or not registered to the SRST router, use the
alias command.
10-19
Chapter 10
Figure 10-5
Unified CM
Cisco Unified
SRST Manager
PSTN/ISDN
Enhanced
SRST
Central Site
Remote Office
348622
IP WAN
Cisco Unified SRST Manager consumes bandwidth from the WAN link when uploading the Unified CM
configurations to provision the remote office router. The Cisco Unified SRST Manager software does
not perform packet marking, therefore the Cisco Unified SRST Manager traffic will travel as best-effort
on the network. Cisco recommends maintaining this best-effort marking, which is IP Precedence 0
(DSCP 0 or PHB BE), to ensure that it does not interfere with real-time high priority voice traffic. To
ensure that Cisco Unified SRST Manager traffic does not cause congestion and to reduce the chances of
packet drop, Cisco recommends scheduling the configuration upload to take place during non-peak hours
(for example, in the evening hours or during the weekend). The configuration upload schedule can be set
from the Cisco Unified SRST Manager web interface.
Consider the following guidelines when you deploy Cisco Unified SRST Manager:
Cisco Unified SRST Manager is not supported with the Cisco Unified Communications 500 Series
platform.
The remote office voice gateway must be co-resident with (reside on) the SRST router.
There is no high availability support with Cisco Unified SRST Manager. If Cisco Unified SRST
Manager is unavailable, configuration upload is not possible.
Cisco Unified SRST Manager is not supported in deployments where NAT is used between the
headquarters and branch locations.
10-20
Chapter 10
Figure 10-6
Cisco Unified
Session Management Edition
Centralized
PSTN
Centralized
Applicationss
Campus
Campus
Remote
Office
Campus
IP PBX
Cisco
Unified
CME
10-21
Chapter 10
Each site in the distributed call processing model can be one of the following:
A single site with its own call processing agent, which can be either:
Cisco Unified Communications Manager (Enterprise or Business Edition 7000)
Cisco Business Edition 6000
Cisco Unified Communications Manager Express (Unified CME)
A third-party IP PBX
A legacy PBX with Voice over IP (VoIP) gateway
A centralized call processing site and all of its associated remote sites
The multisite model with distributed call processing has the following design characteristics:
A centralized platform for trunk and dial plan aggregation is commonly deployed. This platform is
typically a Cisco Unified Communications Manager Session Management Edition (SME) cluster,
although a Session Initiation Protocol (SIP) Proxy Server could also be used to provide intercluster
call routing and dial plan aggregation in multisite distributed call processing deployments.
These services can be deployed centrally, thus benefiting from centralized management and
economies of scale. Services that need to track end-user status (for example, Cisco IM and Presence)
must connect to the Unified CM cluster for the users that they serve.
High-bandwidth audio (for example, G.711 or G.722) between devices in the same site, but
low-bandwidth audio (for example, G.729) between devices in different sites.
High-bandwidth video (for example, 1.5 Mbps with 4CIF or 720p, to 2 Mbps with 1080p) between
devices in the same site, but low-bandwidth video (for example, 384 kbps with 448p or CIF)
between devices at different sites.
Minimum of 1.5 Mbps or greater WAN link speeds. Video is not recommended on WAN connections
that operate at speeds lower than 1.5 Mbps.
An IP WAN interconnects all the distributed call processing sites. Typically, the PSTN serves as a backup
connection between the sites in case the IP WAN connection fails or does not have any more available
bandwidth. A site connected only through the PSTN is a standalone site and is not covered by the
distributed call processing model. (See Campus Deployments, page 10-8.)
Connectivity options for the IP WAN include:
Leased lines
Frame Relay
10-22
Chapter 10
Cisco Unified Communications Manager Session Management Edition is commonly used for
intercluster call routing and dial plan aggregation in distributed call processing deployments. Intercluster
call routing can be number based using standard numeric route patterns, or URI and number based using
the Intercluster Lookup Service (ILS) and Global Dial Plan Replication (GDPR) (see Global Dial Plan
Replication, page 14-48). Unified CM Session Management Edition uses exactly the same code and user
interface as Unified CM but leverages support for multiple trunk protocols (SIP, H.323, and MGCP) as
well as sophisticated trunk, digit manipulation, and call admission control features. Unified CM Session
Management Edition cluster deployments typically consist of many trunks (SIP trunks are
recommended; see Cisco Unified CM Trunks, page 6-1) and no Unified Communications endpoints.
Unified CM Session Management Edition clusters can use all of the high availability features (such as
clustering over the WAN, and Run on all Unified CM Nodes) that are available to Unified CM clusters.
SIP Proxy Deployments
SIP proxies such as the Cisco Unified SIP Proxy provide call routing and SIP signaling normalization.
The following best practices apply to the use of SIP proxies:
Note
Ensure that the SIP proxies have the capacity for the call rate and number of calls required in the
network.
Because Session Management Edition (SME) uses exactly the same code and GUI as Unified CM and
can also share intercluster features such as ILS, GDPR, and Enhanced Locations Call Admission Control
(ELCAC), SME is the recommended trunk and dial plan aggregation platform in multi-site distributed
call processing deployments.
Leaf Unified Communications Systems for the Distributed Call Processing Model
Your choice of call processing agent will vary, based on many factors. The main factors, for the purpose
of design, are the size of the site and the functionality required.
For a distributed call processing deployment, each site may have its own call processing agent. The
design of each site varies with the call processing agent, the functionality required, and the fault
tolerance required. For example, in a site with 500 phones, a Unified CM cluster containing two servers
can provide one-to-one redundancy, with the backup server being used as a publisher and Trivial File
Transfer Protocol (TFTP) server.
10-23
Chapter 10
The requirement for IP-based applications also greatly affects the choice of call processing agent
because only Unified CM provides the required support for many Cisco IP applications.
Table 10-4 lists recommended call processing agents.
Table 10-4
Recommended Size
Up to 450 phones
Up to 2,500 phones
50 to 40,000 phones
Comments
IP PBX
SIP trunks
SIP trunks are recommended for SME and leaf Unified Communications systems because SIP offers
additional features and functionality over H.323 and MGCP trunks. (For more information, see the
chapter on Cisco Unified CM Trunks, page 6-1.)
10-24
Chapter 10
Cisco Unified CM Session Management Edition supports the following call types:
Voice calls
Video calls
Encrypted calls
Fax calls
Unified CM Session Management Edition may also be used to connect to the PSTN and third-party
unified communications systems such as PBXs and centralized unified communications applications.
(See Figure 10-7.) As with any standard Unified CM cluster, third-party connections to Unified CM
Session Management Edition should be system tested for interoperability prior to use in a production
environment.
Figure 10-7
Multisite Distributed Call Processing Deployment with Unified CM Session Management Edition
348624
Cisco Unified
Session Management Edition
Centralized
PSTN
Centralized
Applications
SIP
Campus
Campus
Remote
Office
TDM
PBX
IP PBX
Cisco
Unified
CME
10-25
Chapter 10
Note
Running Intercluster Lookup Service (ILS) and Global Dial Plan Replication (GDPR) on
SME and Unified CM leaf clusters further simplifies dial plan administration because
individual directory numbers, E.164 numbers corresponding to DNs, route patterns (for
internal and external number ranges), and URIs can be distributed using the ILS service.
This approach simplifies dial plan administration by reducing the required number of route
patterns to one SIP route pattern per call control system (Unified CM cluster, for example),
instead of a route pattern for each unique number range. For more information on ILS and
GDPR, see Intercluster Lookup Service (ILS) and Global Dial Plan Replication (GDPR),
page 10-29.
Centralize applications
The deployment of a Unified CM Session Management Edition enables commonly used applications
such as conferencing or voicemail to connect directly to the Session Management cluster, thus
reducing the overhead of managing multiple trunks to leaf systems.
Differences Between Unified CM Session Management Edition and Standard Unified CM Clusters
The Unified CM Session Management Edition software is exactly the same as Unified CM. Unified CM
Session Management Edition is designed to support a large number of trunk-to-trunk connections, and
as such it is subject to the following design considerations:
Capacity
It is important to correctly size the Unified CM Session Management cluster based on the expected
BHCA traffic load between leaf Unified Communications systems (for example, between Unified CM
clusters and PBXs), to and from any centralized PSTN connections, and to any centralized applications.
Determine the average BHCA and Call Holding Time for users of your Unified Communications system
10-26
Chapter 10
and share this information with your Cisco account Systems Engineer (SE) or Cisco Partner to size your
Unified CM Session Management Edition cluster correctly. For more information on SME sizing, see
the chapter on Collaboration Solution Sizing Guidance, page 25-1.
Trunks
Although SME supports SIP, H.323, and MGCP trunks, Cisco recommends using SIP as the trunk
protocol for SME and Unified CM leaf clusters running Cisco Unified CM 8.5 and later releases.
SIP trunks provide a number of unique features that greatly simplify trunk designs and Unified
Communications deployments, such as:
OPTIONS Ping
Lua scripts, which allow SIP Message and Session Description Protocol (SDP) content modification
for interoperability
Using only SIP trunks in the SME cluster allows you to deploy a "media transparent" cluster where
media resources, when required, are inserted by the end or leaf Unified Communications system and
never by SME. Using only SIP trunks also allows you to use extended round trip times (RTTs) between
SME nodes when clustering over the WAN.
Both leaf Unified CM cluster SIP trunks and SME SIP trunks should be configured as Best Effort Early
Offer trunks. For more details on SIP trunks and Best Effort Early Offer, see the chapter on Cisco
Unified CM Trunks, page 6-1.
Media Resources
When a media resource such as an MTP or transcoder is needed to allow a call to proceed successfully,
these resources should ideally be allocated by the leaf Unified Communications systems. If SME trunk
media resources are used for a call traversing the SME cluster, the media path call will hairpin through
the SME media resource. By using SIP trunks only and either Best Effort Early Offer or MTP-less
Early Offer, you can deploy an SME cluster without media resources. If or when media resources are
required, they can be allocated by the leaf Unified Communications system.
Clustering over the WAN
SME deployments can support extended round-trip times (RTTs) of up to 500 ms between SME cluster
nodes. (See Figure 10-8.) This extended RTT applies only to SME clusters (80 ms is the maximum RTT
for a standard Unified CM cluster designs) and is subject to the following design restrictions:
Extended round-trip times for SME deployments with clustering over the WAN are supported where
only SIP trunks are configured in the SME cluster. All SIP trunks must be configured as either Best
Effort Early Offer or MTP-less Early Offer and must use the Run on all Unified CM Nodes
feature so that calls are not routed between nodes within the SME cluster. (For more information,
see the chapter on Cisco Unified CM Trunks, page 6-1.) MGCP, SCCP, and H.323 protocols do not
support extended round-trip times for SME deployments with clustering over the WAN.
No media resources such as MTPs, Trusted Relay Points (TRPs), RSVP agents, or transcoders are
configured or registered to the SME cluster. (To disable media resources hosted on Unified CM
nodes, deactivate the IPVMS service on each node within the cluster.)
A minimum of 1.544 Mbps (T1) bandwidth is required for Intra-Cluster Communication Signaling
(ICCS) traffic between sites.
10-27
Chapter 10
In addition to the bandwidth required for Intra-Cluster Communication Signaling (ICCS) traffic, a
minimum of 1.544 Mbps (T1) bandwidth is required for database and other inter-server traffic
between the publisher and every remote subscriber node.
Like all other SME designs, your SME design must be reviewed and approved by the Cisco SME
team prior to deployment.
Note
The upgrade process for an SME cluster consists of two key parts: Version switch-over, where the call
processing node is re-booted and initialized with the new software version (this takes approximately
45 minutes per server), and database replication, where the subscriber's database is synchronized with
that of the publisher node. The time taken to complete this database replication phase depends on the
RTT between the publisher and subscriber nodes and the number of subscribers in the cluster. The
database replication process has a minimal impact of the subscriber's call processing capability and
typically can be run as a background process during normal SME cluster operation. Avoid making
changes to the SME cluster configuration during the database replication phase because this increases
the time it takes to complete the replication.
For SME clusters deployed with extended RTTs, before upgrading the cluster, run the following Admin
level CLI command on the publisher node:
utils dbreplication setprocess 40
This command improves replication setup performance and reduces database replication times.
Figure 10-8
Unified CM Session Management Edition Clustering over the WAN with Extended Round Trip Times
North America
Europe
< 500 mS
RTT
Latin America
Publisher Node
Asia Pac
SIP Trunks
Best Effort Early Offer
348625
Cisco Unified
Session Management Edition
Cluster
10-28
Chapter 10
Unified CM Versions
Using the latest Cisco Unified Communications System release and SIP trunks across all Unified CM
leaf clusters and the SME cluster allows your Unified Communications deployment to benefit from
common cross-cluster features such as Codec Preference Lists, Intercluster Lookup Service (ILS),
Global Dial Plan Replication (GDPR), and Enhanced Locations Call Admission Control (ELCAC). If
you do not wish to upgrade to the latest Unified Communications version on all clusters, the lowest
recommended version is Cisco Unified CM 8.5 using SIP trunks, because this version includes features
that improve and simplify call routing through Unified CM and Session Management Edition clusters.
Interoperability
Even though most vendors do conform to standards, differences can and do exist between protocol
implementations from various vendors. As with any standard Unified CM cluster, Cisco strongly
recommends that you conduct end-to-end system interoperability testing with any unverified third-party
unified communications system before deploying the system in a production environment. The
interoperability testing should verify call flows and features from Cisco and third-party leaf systems
through the Unified CM Session Management cluster. To learn which third-party unified
communications systems have been tested by the Cisco Interoperability team, refer to the information
available on the Cisco Interoperability Portal at
http://www.cisco.com/en/US/solutions/ns340/ns414/ns728/interOp_ucSessionMgr.html
For SIP trunk interoperability issues, Lua scripting can be used to modify inbound and outbound SIP
messages and SDP content.
Load Balancing for Inbound and Outbound Calls
Configure trunks on the Unified CM Session Management Edition and leaf unified communications
systems so that inbound and outbound calls are evenly distributed across the Unified CM servers within
the Session Management cluster. As a general rule, always enable the Run on All Unified CM Nodes
feature if it is available. For more information on load balancing for trunk calls, refer to the chapter on
Cisco Unified CM Trunks, page 6-1.
Design Guidance and Assistance
For detailed information on trunk configuration for Unified CM Session Management Edition designs
and deployments, refer to the chapter on Cisco Unified CM Trunks, page 6-1.
Note
Before deployment, Unified CM Session Management Edition designs should be reviewed by your
Cisco SE in conjunction with the Cisco Unified CM Session Management Team.
Intercluster Lookup Service (ILS) and Global Dial Plan Replication (GDPR)
Global Dial Plan Replication (GDPR) uses the Intercluster Lookup Service (ILS) to share dial plan
information between participating ILS-enabled clusters. GDPR allows each cluster to distribute
information about its associated URIs, +E.164 numbers, enterprise numbers, +E.164 patterns, enterprise
patterns, and PSTN failover numbers. Each participating cluster shares a common Global Dial Plan
catalogue, which contains every number and URI advertised with GDPR and a corresponding route
string that identifies in which cluster (or end Unified Communications system) the number or URI
resides. (See Figure 10-9.)
10-29
Chapter 10
Figure 10-9
Session
Management
Edition
Third-Party UC System
London
Bob@cisco.com
88242222
+442088242000
+44208XXXXXXX
+ 121255512XX
Ann@cisco.com
Bob@cisco.com
San-Jose.CA.US
London.UK.EU
84081111
88242222
Bob@cisco.com
88242222
+442088242000
+44208XXXXXXX
San-Jose.CA.US
Anne@cisco.com
84081111
+14089021000
+1408XXXXXXX
Third-Party-UC
Third-Party-UC
London.UK.EU
+121255512XX
PSTN
+121255512XX
Intercluster Lookup Service
SIP Trunk
+1408XXXXXXX
+442088242000
PSTN
+44208XXXXXXX
348626
+14089021000
With GDPR, each cluster advertises its dial plan information (numbers and URIs) with a location
attribute, known as a route string. When a call is placed to a number or URI, Unified CM checks to see
whether the number or URI is associated to a device within the cluster. If it is not, Unified CM searches
its GDPR catalogue for the number or URI. If a match is found in the Global Dial Plan catalogue, GDPR
returns the route string that corresponds to the cluster where the number or URI resides. Unified CM
uses the returned route string as a candidate to match to an existing SIP route pattern and corresponding
SIP trunk. (See Figure 10-10 and Figure 10-11.)
10-30
Chapter 10
Figure 10-10
Diane calls
+442088242000
No
Ann@cisco.com
Session
Management
Edition
San-Jose.CA.US
Anne@cisco.com
84081111
+14089021000
+1408XXXXXXX
Is +442088242000 a
locally registered DN ?
San-Jose.CA.US
London.UK.EU
Bob@cisco.com
88242222
+442088242000
+44208XXXXXXX
Third-Party UC System
+ 121255512XX
Bob@cisco.com
Yes
London.UK.EU
Can +442088242000 be
found in ILS ?
88242222
84081111
Return Route String
London.UK.EU for DN
+442088242000
London.UK.EU
Third-Party-UC
+121255512XX
Third-Party-UC
PSTN
+121255512XX
PSTN
+442088242000
+44208XXXXXXX
SIP Trunks
+1408XXXXXXX
348627
+14089021000
Bob@cisco.com
88242222
+442088242000
+44208XXXXXXX
San-Jose.CA.US
Anne@cisco.com
84081111
+14089021000
+1408XXXXXXX
10-31
Chapter 10
Diane calls
+442088242000
Third-Party UC System
London.UK.EU
Yes
Ann@cisco.com
Session
Management
Edition
348628
Figure 10-11
Bob@cisco.com
88242222
+442088242000
+44208XXXXXXX
+ 121255512XX
Bob@cisco.com
San Jose.CA.US
London.UK.EU
84081111
+14089021000
88242222
London.UK.EU
Bob@cisco.com
88242222
+442088242000
+44208XXXXXXX
Third-Party-UC
San-Jose.CA.US
Anne@cisco.com
84081111
+14089021000
+1408XXXXXXX
+442088242000
Third-Party-UC
+121255512XX
+121255512XX
PSTN
PSTN
SIP Trunks
+1408XXXXXXX
+44208XXXXXXX
Using GDPR is significantly different from using a standard dial plan with numeric route patterns.
Instead of requiring a route pattern for each unique number range within the Unified Communications
network, GDPR distributes the numbers, number patterns, and URIs, and only a single SIP route pattern
is required for each cluster within the Unified Communications network. Numbers and URIs associated
with third-party unified communications systems (and Unified CM clusters that do not support ILS and
GDPR) can be imported as catalogues into GDPR and distributed through ILS with a route string that
corresponds to each unified communications system. Because both individual numbers and route
patterns corresponding to groups of numbers can be advertised with GDPR, this abstraction of numbers
and number ranges away from numeric route patterns allows GDPR to simply and easily support highly
fragmented dial plans with many number ranges. Each cluster using ILS and GDPR can block and purge
individual numbers and number ranges advertised from other participating clusters.
Each GDPR number type (+E.164 number, enterprise number, +E.164 pattern, or enterprise pattern) is
placed into a specific partition when learned through ILS, allowing per-user or per-device class of
service to be applied based on number type partitions and calling search spaces.
Cisco Unified Border Element also supports number and URI call routing using dial peers that match on
a GDPR route string value, which is sent to Cisco Unified Border Element during call setup over a
Unified CM SIP trunk. GDPR route string matching with Cisco IOS dial peers is supported with
Cisco IOS releases 15.3(3)M, 15.4(1)T (ISR), 15.3(3)S (ASR), and later releases.
10-32
Chapter 10
VPN-less access
Business-to-business communications
IP PSTN access
These four deployment options for the collaboration edge are discussed in the sections that follow.
Mobile devices such as laptops, tablets, and smartphones can deploy the Cisco AnyConnect VPN
client to access Unified Communications services (for example, Unified CM, Cisco IM and
Presence, Cisco Unity, and others) as well as business application services (for example, the
corporate email system and internal websites) within the enterprise network. With this VPN
connection established, Unified Communications soft clients such as Cisco Jabber and Cisco IP
Communicator can register with Unified CM and make voice, video, and encrypted calls between
enterprise devices.
Home office workers with one or more enterprise devices can deploy a Cisco Virtual Office (CVO)
Integrated Services Router (ISR) to extend the enterprise network to their homes over a VPN. The
CVO VPN connection provides connected devices with access to Unified Communications services
(for example, Unified CM, Cisco IM and Presence, Cisco Unity, and others) as well as business
application services (for example, the corporate email system and internal websites) within the
enterprise network. With the CVO VPN connection established, Unified Communications soft
clients and IP phones can register with Unified CM and make voice, video, and encrypted calls
between enterprise devices.
The Cisco VPN client for Cisco Unified IP Phones provides enterprise access for a subset of Cisco
Unified IP Phone models. For more information on the devices supporting the Cisco VPN client for
Cisco Unified IP Phones, see the chapter on Collaboration Endpoints, page 8-1. The phone's VPN
client creates a tunnel (for the phone only), allowing to it register with Unified CM and to make
voice, video, and encrypted calls between enterprise devices. A computer connected to the phone's
PC port is responsible for authenticating and establishing its own tunnel to the enterprise with VPN
client software.
VPN access gives the user access to all Unified Communications and business applications within the
enterprise by creating a secure encrypted tunnel from the device to the VPN head-end. All traffic,
including traffic destined for the internet and media for calls between VPN users, must always traverse
the enterprise network rather than be established directly from the device over the internet to its
destination. (See Figure 10-12.)
10-33
Chapter 10
Figure 10-12
Signaling
Media
VPN Tunnel
Data
IP PSTN
Unified CM
UC Applications
Coffee Shop
Data Applications
E-Mail & Web Services
Central Site
Home Office
Home Office
348629
Internet
All of the above devices use their VPN client to connect to the enterprise network via a VPN head-end
platform such as a Cisco Adaptive Security Appliance (ASA 5500) or a Cisco VPN aggregation router.
For more information on VPN access solutions, refer to the Teleworking and BYOD solutions guides
available at
http://www.cisco.com/en/US/netsol/ns982/networking_solutions_program_home.html#~slng
10-34
Chapter 10
This deployment type uses Cisco Expressway-C and Expressway-E. Cisco Expressway-E can be placed
either in a DMZ or in the public internet, and it communicates by means of Cisco Expressway-C to the
Unified CM cluster in the enterprise network. (See Figure 10-13.) Cisco Expressway supports VPN-less
access primarily for Cisco Jabber clients and TelePresence endpoints. Voice, video, encrypted calls, and
IM and Presence are supported between enterprise endpoints. Media and signaling for calls between
remote VPN-less devices traverse Cisco Expressway-C and Expressway-E. For specific information on
the range of endpoints supported with Cisco Expressway VPN-less enterprise access, see the chapter on
Collaboration Endpoints, page 8-1. For more information on Cisco Expressway VPN-less client access,
refer to the documentation available at the following locations:
Figure 10-13
http://www.cisco.com/en/US/netsol/ns1246/index.html
http://www.cisco.com/en/US/products/ps13435/index.html
http://www.cisco.com/c/en/us/support/unified-communications/expressway-series/products-installati
on-and-configuration-guides-list.html
Media
Secure Signaling
Data
IP PSTN
UC Applications
Unified CM
IM and
Presence
Coffee Shop
Central Site
Cisco
Expressway-E
Home Office
Home Office
348630
Internet
Cisco
Expressway-C
Business-to-Business Communications
Both Cisco Expressway and Cisco Unified Border Element (CUBE) support Internet based
business-to-business unified communications connections between enterprises. Both Cisco Expressway
and CUBE use SIP or H.323 trunks for business-to-business unified communications signaling. Cisco
Expressway supports voice calls, video calls, and IM and Presence federation (see Figure 10-14); while
CUBE supports voice calls and video calls only (see Figure 10-15).
10-35
Chapter 10
Figure 10-14
UC
Applications
pp cations
UC
Applications
Applica
IP PSTN
Unified CM
Unified CM
IM &
Presence
IM &
Presence
Internet
Cisco
o
Expressway-E
Cisco
C
Expressway-E
Cisco
Expressway-C
348780
Cisco
Expressway-C
DNS
Enterprise A
Figure 10-15
Enterprise B
UC
Applications
pplicatio
IP PSTN
Unified CM
Unified CM
Internet
CUBE
U
348781
CUBE
U
DNS
Enterprise A
Enterprise B
IP PSTN Deployments
IP PSTN deployments are increasing in popularity and are gradually replacing existing TDM-based
PSTN access. SIP is commonly used as the IP PSTN access protocol, and today many service providers
offer a voice-only service to the IP PSTN through a session border controller such a Cisco Unified
Border Element. Session border controllers are SIP Back-to-Back User Agents (B2BUAs) and are
typically used in flow-through mode, where both the voice media and SIP signalling for each call flow
through Cisco Unified Border Element. (See Figure 10-16.) As a B2BUA in flow-through mode, Cisco
Unified Border Element can implement sophisticated QoS marking and call admission control policies
while also providing support for transcoding, encryption, media forking for call recording applications,
10-36
Chapter 10
and scripting that allows SIP messages and SDP content to be modified for interoperability. For more
info on Cisco Unified Border Element features and functions, refer to the latest version of the Cisco
Unified Border Element Enterprise Edition data sheet, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps5640/products_data_sheets_list.html
Cisco Unified Border Element is supported on wide range of Cisco routing platforms, from the Cisco
800 Series Integrated Services Routers (ISR) to the Cisco 1000 Series Aggregation Service Routers
(ASR). Depending on the hardware platform, Cisco Unified Border Element can provide session
scalability from 4 to 16,000 concurrent voice calls. Cisco Unified Border Element also provides
redundancy on the following platforms:
Note
Figure 10-16
The Cisco ISR-G2 platforms, which can provide box-to-box redundancy with media preservation
for stable active calls (Cisco IOS Release 15.1.2T or later).
The Cisco ASR platforms, which can provide box-to-box or in-box redundancy with media and
signaling preservation (stateful failover) for stable active calls.
Access to the IP PSTN and access to the enterprise for VPN-less clients can be deployed on the same
Cisco Unified Border Element platform.
Unified CM
Device Signaling
Voice Media
SIP
Applications
IP PSTN
Unified CM
Cisco Unified
Border Element
Media Resources
Cisco Unified
Border Element
Central Site
Remote Office
348632
IP WAN
10-37
Chapter 10
SIP trunks may be connected to IP PSTN service providers in several different ways, depending on the
desired architecture. The two most common architectures for this connectivity are centralized trunks and
distributed trunks.
Centralized trunks connect to the service provider (SP) through one logical connection (although there
may be more than one physical connection for redundancy) with session border controllers (SBCs) such
as the Cisco Unified Border Element. All IP PSTN calls to and from the enterprise use this set of trunks,
and for most calls, media and signaling traverse the enterprise WAN to connect devices in the enterprise
to those in the PSTN.
Distributed trunks connect to the service provider through several logical connections. Each branch of
an enterprise may have its own local trunk to the service provider. With distributed trunks, media from
the branch no longer needs to traverse the enterprise WAN but can flow directly to the service provider
through a local SBC.
Each connectivity model has its own advantages and disadvantages. Centralized trunks are generally
easier to deploy in terms of both physical equipment and configuration complexity. Distributed trunks
have the advantage of local hand-off of media and better number portability from local providers.
Alternatively, a hybrid connectivity model that combines some centralized and distributed IP PSTN
access can capture the advantages of both forms of IP PSTN deployment.
10-38
Chapter 10
Figure 10-17
Dual Call Control Deployments with Centralized and Distributed Third-Party Systems
Bandwidth for third-party voice and video
Remote Office A
Unified CM
Third-Party
Call Control
348633
Remote Office B
Central Site
10-39
Chapter 10
Use an explicit route pattern and corresponding trunk for each of the unique number ranges
associated with each call control.
Within the Unified CM (and SME, if used) deployment, use the Intercluster Lookup Service (ILS)
and Global Dial Plan Replication (GDPR) to share information about the number ranges supported
by each Unified CM cluster and each third-party unified communications system. For third-party
systems and their associated devices, import each unique number range into GDPR and associate
each imported number range with a route string (a label that identifies the call control system). When
a Unified CM user dials a number, Unified CM checks to see if the number is registered to its cluster.
If the number is not registered to the Unified CM cluster, Unified CM searches ILS for the called
number and its corresponding route string. The route string identifies the call control cluster where
the number resides, which is used to match a SIP route pattern that then forwards the call over a SIP
trunk toward its destination.
10-40
Chapter 10
If alphanumeric URIs are used to address and call endpoints registered to Unified CM and the third-party
call control system, then call routing can be implemented in either of the following ways, depending on
the deployment:
For deployments where only a single third-party call control system exists with a single SIP trunk
to a Unified CM cluster, a default SIP route can be configured on Unified CM and the third-party
call control system, so that calls to endpoints that are not found on one call control are sent to the
other call control.
If multiple third-party call control systems are deployed, use the Intercluster Lookup Service (ILS)
and Global Dial Plan Replication (GDPR) to share information about the URIs supported by each
Unified CM cluster and each third-party unified communications system. For URI-based call
routing when a Unified CM user dials a URI, Unified CM checks to see if the URI is registered to
its cluster. If it is not, Unified CM searches the ILS for the called URI and its corresponding
route-string. The route string identifies the call control cluster where the URI resides, and it is used
to match a SIP route pattern, which then forwards the call over a SIP trunk toward the destination
URI. For URI-based endpoints registered to a third-party call control system, the list of URIs
registered to the third-party call control system must be imported manually into ILS along with the
corresponding route string for the third-party call control system.
Note
Remote failover deployments might require higher bandwidth because a large amount of intra-cluster
traffic flows between the subscriber servers.
You can also use a combination of the two deployment models to satisfy specific site requirements. For
example, two main sites may each have primary and backup subscribers, with another two sites
containing only a primary server each and utilizing either shared backups or dedicated backups at the
two main sites.
Some of the key advantages of clustering over the WAN are:
Single point of administration for users for all sites within the cluster
Feature transparency
10-41
Chapter 10
These features make this solution ideal as a disaster recovery plan for business continuance sites or as a
single solution for multiple small or medium sites.
WAN Considerations
For clustering over the WAN to be successful, you must carefully plan, design, and implement various
characteristics of the WAN itself. The Intra-Cluster Communication Signaling (ICCS) between
Unified CM servers consists of many traffic types. The ICCS traffic types are classified as either priority
or best-effort. Priority ICCS traffic is marked with IP Precedence 3 (DSCP 24 or PHB CS3). Best-effort
ICCS traffic is marked with IP Precedence 0 (DSCP 0 or PHB BE). The various types of ICCS traffic
are described in Intra-Cluster Communications, page 10-43, which also provides further guidelines for
provisioning. The following design guidelines apply to the indicated WAN characteristics:
Delay
The maximum one-way delay between any two Unified CM servers should not exceed 40 ms, or
80 ms round-trip time. Measuring the delay is covered in Delay Testing, page 10-44. Propagation
delay between two sites introduces 6 microseconds per kilometer without any other network delays
being considered. This equates to a theoretical maximum distance of approximately 6,000 km for
40 ms delay or approximately 3,720 miles. These distances are provided only as relative guidelines
and in reality will be shorter due to other delay incurred within the network.
Jitter
Jitter is the varying delay that packets incur through the network due to processing, queue, buffer,
congestion, or path variation delay. Jitter for the IP Precedence 3 ICCS traffic must be minimized
using Quality of Service (QoS) features.
Bandwidth
Provision the correct amount of bandwidth between each server for the expected call volume, type
of devices, and number of devices. This bandwidth is in addition to any other bandwidth for other
applications sharing the network, including voice and video traffic between the sites. The bandwidth
provisioned must have QoS enabled to provide the prioritization and scheduling for the different
classes of traffic. The general rule of thumb for bandwidth is to over-provision and under-subscribe.
Quality of Service
The network infrastructure relies on QoS engineering to provide consistent and predictable
end-to-end levels of service for traffic. Neither QoS nor bandwidth alone is the solution; rather,
QoS-enabled bandwidth must be engineered into the network infrastructure.
10-42
Chapter 10
Intra-Cluster Communications
In general, intra-cluster communications means all traffic between servers. There is also a real-time
protocol called Intra-Cluster Communication Signaling (ICCS), which provides the communications
with the Cisco CallManager Service process that is at the heart of the call processing in each server or
node within the cluster.
The intra-cluster traffic between the servers consists of the following:
Note
Database traffic from the IBM Informix Dynamic Server (IDS) database that provides the main
configuration information. The IDS traffic may be re-prioritized in line with Cisco QoS
recommendations to a higher priority data service (for example, IP Precedence 1 if required by the
particular business needs). An example of this is extensive use of Extension Mobility, which relies
on IDS database configuration.
Firewall management traffic, which is used to authenticate the subscribers to the publisher to access
the publisher's database. The management traffic flows between all servers in a cluster. The
management traffic may be prioritized in line with Cisco QoS recommendations to a higher priority
data service (for example, IP Precedence 1 if required by the particular business needs).
ICCS real-time traffic, which consists of signaling, call admission control, and other information
regarding calls as they are initiated and completed. ICCS uses a Transmission Control Protocol
(TCP) connection between all servers that have the Cisco CallManager Service enabled. The
connections are a full mesh between these servers. This traffic is priority ICCS traffic and is marked
dependant on release and service parameter configuration.
CTI Manager real-time traffic is used for CTI devices involved in calls or for controlling or
monitoring other third-party devices on the Unified CM servers. This traffic is marked as priority
ICCS traffic and exists between the Unified CM server with the CTI Manager and the Unified CM
server with the CTI device.
For detailed information on various types of traffic between Unified CM servers, refer to the TCP and
UDP port usage documents at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html.
Unified CM Publisher
The publisher server replicates a partial read-only copy of the master database to all other servers in the
cluster. Most of the database modifications are done on the publisher. If changes such as administration
updates are made in the publishers master database during a period when another server in the cluster
is unreachable, the publisher will replicate the updated database when communications are
re-established. Database modifications for user-facing call processing features are made on the
subscriber servers to which the IP phones are registered. These features include:
Privacy Enable/Disable
10-43
Chapter 10
Device Mobility
CTI Certificate Authority Proxy Function (CAPF) status for end users and application users
Each subscriber replicates these changes to every other server in the cluster. Any other configuration
changes cannot be made on the database during the period when the publisher is unreachable or offline.
Most normal operations of the cluster, including the following, will not be affected during the period of
publisher failure:
Call processing
Failover
Other services or applications might also be affected, and their ability to function without the publisher
should be verified when deployed.
Delay Testing
The maximum round-trip time (RTT) between any two servers must not exceed 80 ms. This time limit
must include all delays in the transmission path between the two servers. Verifying the round trip delay
using the ping utility on the Unified CM server will not provide an accurate result. The ping is sent as a
best-effort tagged packet and is not transported using the same QoS-enabled path as the ICCS traffic.
Therefore, Cisco recommends that you verify the delay by using the closest network device to the
Unified CM servers, ideally the access switch to which the server is attached. Cisco IOS provides a
extended ping capable to set the Layer 3 type of service (ToS) bits to make sure the ping packet is sent
on the same QoS-enabled path that the ICCS traffic will traverse. The time recorded by the extended ping
is the round-trip time (RTT), or the time it takes to traverse the communications path and return.
The following example shows a Cisco IOS extended ping with the IP Precedence set to 3 (ToS byte value
set to 96):
Access_SW#ping
Protocol [ip]:
Target IP address: 10.10.10.10
Repeat count [5]:
Datagram size [100]:
Timeout in seconds [2]:
Extended commands [n]: y
Source address or interface:
Type of service [0]: 96
Set DF bit in IP header? [no]:
Validate reply data? [no]:
Data pattern [0xABCD]:
Loose, Strict, Record, Timestamp, Verbose[none]:
Sweep range of sizes [n]:
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.10.10.10, timeout is 2 seconds:
!!!!!
10-44
Chapter 10
Error Rate
The expected error rate should be zero. Any errors, dropped packets, or other impairments to the IP
network can have an impact on the call processing performance of the cluster. This may be noticeable
by delay in dial tone, slow key or display response on the IP phone, or delay from off-hook to connection
of the voice path. Although Unified CM will tolerate random errors, they should be avoided to avoid
impairing the performance of the cluster.
Troubleshooting
If the Unified CM subscribers in a cluster are experiencing impairment of intra-cluster communications
due to higher than expected delay, errors, or dropped packets, some of the following symptoms might
occur:
IP phones, gateways, or other devices on a remote Unified CM server within the cluster might
temporarily be unreachable.
Users might experience longer than expected delays before hearing dial tone.
The time taken to upgrade a subscriber and synchronize its database with the publisher will increase.
Verify that sufficient bandwidth is provisioned for the queues across the WAN to support all the
traffic.
10-45
Chapter 10
Figure 10-18
Site 1
Publisher
Sub A
Site 2
Sub B
Sub E
Sub F
Backup G
Backup H
WAN
TFTP/ MoH
Backup C
Backup D
TFTP/ MoH
348634
Observe the following guidelines when implementing the local failover model:
Configure each site to contain at least one primary Unified CM subscriber and one backup
subscriber.
Configure Unified CM groups and device pools to allow devices within the site to register with only
the servers at that site under all conditions.
Cisco highly recommends that you replicate key services (TFTP, DNS, DHCP, LDAP, and IP Phone
Services), all media resources (transcoders, conferencing resources, annunciator, and music on
hold), and gateways at each site to provide the highest level of resiliency. You could also extend this
practice to include a voicemail system at each site.
Under a WAN failure condition, sites without access to the publisher database will lose some
functionality. For example, system administration at the remote site will not be able to add, modify,
or delete any part of the configuration. However, users can continue to access the user-facing
features listed in the section on Unified CM Publisher, page 10-43.
Under WAN failure conditions, calls made to phone numbers that are not currently communicating
with the subscriber placing the call, will result in either a fast-busy tone or a call forward (possibly
to voicemail or to a destination configured under Call Forward Unregistered).
10-46
Chapter 10
The maximum allowed round-trip time (RTT) between any two servers in the Unified CM cluster is
80 ms.
Note
At a higher round-trip delay time and higher busy hour call attempts (BHCA), voice
cut-through delay might be higher, causing initial voice clipping when a voice call is
established.
A minimum of 1.544 Mbps (T1) bandwidth is required for Intra-Cluster Communication Signaling
(ICCS) between each site and every other site that is clustered over the WAN. For example, if three
sites are clustered over the WAN, each site would require 2 * 1.544 Mbps of WAN bandwidth for
call control traffic. This minimum bandwidth requirement for call control traffic accounts for up to
for 10,000 busy hour call attempts (BHCA) from one site to another site and applies only to
deployments where directory numbers are not shared between sites that are clustered over the WAN.
The following equation may be used as a guideline to calculate the bandwidth for more than 10,000
BHCA between non-shared directory numbers at a specific delay:
Total Bandwidth (Mbps) = (Total BHCA/10,000) (1 + 0.006 Delay), where
Delay = RTT delay in ms
This call control traffic is classified as priority traffic. Priority ICCS traffic is marked with IP
Precedence 3 (DSCP 24 or PHB CS3).
In addition to the bandwidth required for Intra-Cluster Communication Signaling (ICCS) traffic, a
minimum of 1.544 Mbps (T1) bandwidth is required for database and other inter-server traffic
between the publisher and every subscriber node within the cluster.
For customers who also want to deploy CTI Manager over the WAN (see Figure 10-19), the
following formula can be used to calculate the bandwidth (Mbps) for the CTI Intra-Cluster
Communication Signaling (ICCS) traffic between the Unified CM subscriber running the
CTI Manager service and the Unified CM subscriber to which the CTI controlled endpoint is
registered:
CTI ICCS bandwidth (Mbps) = (Total BHCA/10,000) 0.53
Figure 10-19
Site 1
Site 2
WAN
CTI
Application
Subscriber
running
CTI Manager
for Application
SIP/ SCCP
Subscriber
for Phone
Registration
CTI
Controlled
Device
348635
JTAPI
10-47
Chapter 10
For deployments where the J/TAPI application is remote from the Unified CM subscriber (see
Figure 10-20), the following formula can be used to calculate the Quick Buffer Encoding (QBE)
J/TAPI bandwidth for a typical J/TAPI application:
J/TAPI bandwidth (Mbps) = (Total BHCA/10,000) 0.28
The bandwidth may vary depending on the J/TAPI application. Check with the application developer
or provider to validate the bandwidth requirement.
Figure 10-20
Site 1
JTAPI
Site 2
WAN
CTI
Application
Subscriber
for phone
registration
and running
CTI Manager for
application
CTI
Controlled
Device
348636
Maximum RTT is
application dependent
and based on testing
SIP/ SCCP
Consider two sites, Site 1 and Site 2, with Unified CM clustered over the WAN across these two sites
that are 80 ms round-trip time apart. Site 1 has one publisher, one combined TFTP and music on hold
(MoH) server, and two Unified CM subscriber servers. Site 2 has one TFTP/MoH server and two
Unified CM subscriber servers. Site 1 has 5000 phones, each having one DN; and Site 2 has 5000
phones, each having one DN. During the busy hour, 2500 phones in Site 1 call 2500 phones in Site 2,
each at 3 BHCA. During that same busy hour, 2500 phones in Site 2 also call 2500 phones in Site 1, each
at 3 BHCA. In this case:
Total BHCA during the busy hour = 25003 + 25003 = 15,000
Total bandwidth required between the sites = Total ICCS bandwidth + Total database bandwidth
Because total BHCA is 15,000 (greater than 10,000), we can use the formula to calculate:
Total ICCS bandwidth = (15,000/10,000) (1 + 0.00680) = 2.22 Mbps
Total database bandwidth = (Number of servers remote to the publisher) 1.544 = 3 1.544
= 4.632 Mbps
Total bandwidth required between the sites = 2.22 Mbps + 4.632 Mbps = 6.852 Mbps
(Approximately 7 Mbps)
10-48
Chapter 10
When directory numbers are shared between sites that are clustered over the WAN, additional
bandwidth must be reserved. This overhead or additional bandwidth (in addition to the minimum
1.544 Mbps bandwidth) for 10,000 BHCA between shared DNs can be calculated using the
following equation:
Overhead = (0.012 Delay Shared-line) + (0.65 Shared-line), where:
Delay = RTT delay over the IP WAN, in ms
Shared-line = Average number of additional phones on which a directory number is shared
across the WAN.
The following equation may be used as a guideline to calculate the bandwidth for more than
10,000 BHCA between shared directory numbers at a specific delay:
Total bandwidth (Mbps) = (Total BHCA/10,000) (1 + 0.006 Delay
+ 0.012 Delay Shared-line + 0.65 Shared-line), where:
Delay = RTT delay in ms
Shared-line = Average number of additional phones on which a directory number is shared
across the WAN.
Example 10-2 Bandwidth Calculation for Two Sites with Shared Directory Numbers
Consider two sites, Site 1 and Site 2, with Unified CM clustered over the WAN across these two sites
that are 80 ms round-trip time apart. Site 1 has one publisher, one combined TFTP and music on hold
(MoH) server, and two Unified CM subscriber servers. Site 2 has one TFTP/MoH server and two
Unified CM subscriber servers. Site 1 has 5000 phones, each having one DN; and Site 2 has 5000
phones, each sharing a DN with the 5000 phones in Site 1. Thus, each DN is shared across the WAN with
an average of one additional phone. During the busy hour, 2500 phones in Site 1 call 2500 phones in
Site 2, each at 3 BHCA. This also causes the phones in Site 1 to ring. During that same busy hour, 2500
phones in Site 2 call 2500 phones in Site 1, each at 3 BHCA. This also causes the phones in Site 2 to
ring. In this case:
Total BHCA during the busy hour = 2500 3 + 2500 3 = 15,000
Total bandwidth required between the sites = Total ICCS bandwidth + Total database bandwidth
Because total BHCA is 15,000 (greater than 10,000), we can use the formula to calculate:
Total ICCS bandwidth = (15,000/10,000) (1 + 0.00680 + 0.012801 + 0.651) = 4.635 Mbps
Total database bandwidth = (Number of servers remote to the publisher) 1.544 = 3 1.544
= 4.632 Mbps
Total bandwidth required between the sites = 4.635 Mbps + 4.632 Mbps = 9.267 Mbps
(Approximately 10 Mbps)
Note
The bandwidth requirements stated above are strictly for ICCS, database, and other inter-server traffic.
If calls are going over the IP WAN, additional bandwidth must be provisioned for media traffic,
depending on the voice and video codecs used for the calls. For details see Bandwidth Provisioning,
page 3-52.
Subscriber servers in the cluster read their local database. Database modifications can occur in both
the local database as well as the publisher database, depending on the type of changes. Informix
Dynamic Server (IDS) database replication is used to synchronize the databases on the various
servers in the cluster. Therefore, when recovering from failure conditions such as the loss of WAN
connectivity for an extended period of time, the Unified CM databases must be synchronized with
any changes that might have been made during the outage. This process happens automatically when
10-49
Chapter 10
database connectivity is restored to the publisher and other servers in the cluster. This process can
take longer over low bandwidth and/or higher delay links. In rare scenarios, manual reset or repair
of the database replication between servers in the cluster might be required. This is performed by
using the commands such as utils dbreplication repair all and/or utils dbreplication reset all at
the command line interface (CLI). Repair or reset of database replication using the CLI on remote
subscribers over the WAN causes all Unified CM databases in the cluster to be re-synchronized.
With longer delays and lower bandwidth between the publisher and subscriber nodes, it can take
longer for database replication repair or reset to complete.
Note
If remote branches using centralized call processing with clustering over the WAN are connected to
the central sites via the same WAN path that is used for clustering over the WAN traffic, pay careful
attention to the configuration of call admission control to avoid oversubscribing the links used for
clustering over the WAN.
If the bandwidth is not limited on the links used for clustering over the WAN (that is, if the
interfaces to the links are OC-3s or STM-1s and there is no requirement for call admission
control), then the remote sites may be connected to any of the main sites because all the main
sites should be configured as location Hub_None. This configuration still maintains
hub-and-spoke topology for purposes of call admission control.
If you are using the Multiprotocol Label Switching (MPLS) Virtual Private Network (VPN)
feature, all sites in Unified CM locations and the remote sites may register with any of the main
sites.
If bandwidth is limited between the main sites, call admission control must be used between
sites, and all remote sites must register with the main site that is configured as location
Hub_None. This main site is considered the hub site, and all other remote sites and
clustering-over-the-WAN sites are spokes sites.
During a software upgrade, all servers in the cluster should be upgraded during the same
maintenance period, using the standard upgrade procedures outlined in the software release notes.
The software upgrade time will increase for higher round-trip delay time over the IP WAN. Publisher
to subscriber bandwidth lower than the required 1.544 Mbps for each subscriber node can also cause
the software upgrade process to take longer to complete. If a faster upgrade time is desired,
additional bandwidth above the required 1.544 Mbps per remote subscriber can be provisioned
during the upgrade period.
10-50
Chapter 10
To improve redundancy and upgrade times, Cisco recommends that you enable the Cisco Trivial File
Transfer Protocol (TFTP) service on two Unified CM servers. More than two TFTP servers can be
deployed in a cluster, however this configuration can result in an extended period for rebuilding all the
TFTP files on all TFTP servers.
You can run the TFTP service on either a publisher or a subscriber server, depending on the site and the
available capacity of the server. The TFTP server option must be correctly set in the DHCP servers at
each site. If DHCP is not in use or if the TFTP server is manually configured, you should configure the
correct TFTP address for the site.
Other services, which may affect normal operation of Unified CM during WAN outages, should also be
replicated at all sites to ensure uninterrupted service. These services include DHCP servers, DNS
servers, corporate directories, and IP phone services. On each DHCP server, set the DNS server address
correctly for each location.
IP phones may have shared line appearances between the sites. During a WAN outage, call control for
each line appearance is segmented, but call control returns to a single Unified CM server once the WAN
is restored. During the WAN restoration period, there is additional traffic between the two sites. If this
situation occurs during a period of high call volume, the shared lines might not operate as expected
during that period. This situation should not last more than a few minutes, but if it is a concern, you can
provision additional prioritized bandwidth to minimize the effects.
10-51
Chapter 10
Clustering over the WAN, Remote Failover Model with Four Sites
Site 1
Site 2
WAN
Site 3
Primary Subscriber Registration
Backup Subscriber Registration
348637
Site 4
When implementing the remote failover model, observe all guidelines for the local failover model (see
Local Failover Deployment Model, page 10-45), with the following modifications:
Configure each site to contain at least one primary Unified CM subscriber and an optional backup
subscriber as desired. If a backup subscriber over the IP WAN is not desired, a Survivable Remote
Site Telephony (SRST) router may be used as a backup call processing agent.
You may configure Unified CM groups and device pools to allow devices to register with servers
over the WAN as a second or third choice.
Signaling or call control traffic requires bandwidth when devices are registered across the WAN with
a remote Unified CM server in the same cluster. This bandwidth might be more than the ICCS traffic
and should be calculated using the bandwidth provisioning calculations for signaling, as described
in Bandwidth Provisioning, page 3-52.
10-52
Chapter 10
Note
You can also combine the features of these two types of deployments for disaster recovery purposes. For
example, Unified CM groups permit configuring up to three servers (primary, secondary and tertiary).
Therefore, you can configure the Unified CM groups to have primary and secondary servers that are
located at the same site and the tertiary server at a remote site over the WAN.
http://www.cisco.com/en/US/products/ps10265/index.html
http://www.cisco.com/go/uc-virtualized
For sizing aspects of Unified Communications systems on virtualized servers, use the Cisco
Collaboration Sizing Tool, available to Cisco partners and employees (with valid login authentication) at
http://tools.cisco.com/cucst
Hypervisor
A hypervisor is a thin software system that runs directly on the server hardware to control the hardware,
and it allows multiple operating systems (guests) to run on a server (host computer) concurrently. A
guest operating system (such as that of Cisco Unified CM) runs on another level above the hypervisor.
Hypervisors are one of the foundation elements in cloud computing and virtualization technologies, and
they consolidate applications onto fewer servers.
Most of the Cisco Collaboration Systems applications are supported only with virtualization. This means
that deploying the VMware vSphere ESXi hypervisor is required for those applications and that they
cannot be installed directly on the server (bare metal).
VMware vCenter is a tool that helps to manage your virtual environment. With Tested Reference
Configurations, VMware vCenter is not mandatory; however, it is strongly recommended when
deploying a large number of hosts. With specification-based hardware, VMware vCenter is required.
10-53
Chapter 10
Tested Reference Configurations (TRC), which are selected hardware configurations based on Cisco
Unified Computing System (UCS) platforms. They are tested and documented for specific
guaranteed performance, capacity, and application co-residency scenarios running "full-load" Cisco
Collaboration System virtual machines.
Specification-based hardware that provides more hardware flexibility and that, for example, adds
support for other Cisco UCS and third-party servers that are listed in the VMware Compatibility
Guide (available at (http://www.vmware.com/resources/compatibility/search.php).
For more details on the Cisco Unified Computing System architecture, refer to the documentation
available at
http://www.cisco.com/en/US/netsol/ns944/index.html
Cisco UCS E-Series server modules are blade servers designed to be deployed in Cisco Integrated
Services Routers Generation 2 (ISR G2). Some Cisco Collaboration applications are supported on Cisco
UCS E-Series, but the support might be limited (specification-based hardware support instead of TRC,
for instance).
10-54
Chapter 10
Figure 10-22
Unified Communications
Application Virtual Machine
vNIC
Hypervisor Layer
!
Console
UCS B200 M1
B Series Blade
Reset
Mezzanine CNA
!
UCS B200 M1
UCS B200 M1
UCS 5108
2
!
Reset
Console
UCS B200 M1
Reset
Console
UCS B200 M1
4
!
Reset
Console
UCS B200 M1
Reset
Console
UCS B200 M1
B Series Chassis
6
!
OK
FAIL
Console
Reset
Console
UCS B200 M1
Reset
OK
FAIL
Reset
Console
OK
FAIL
Console
UCS B200 M1
Reset
OK
FAIL
I/O Module
LAN Cloud
FCoE
SAN Cloud
FC
LAN Switch
Fabric
Interconnect
FC
SAN Switch
Storage Array
348638
Ethernet
IP
This section briefly describes the primary UCS components and how they function in a Unified
Communications solution. For details about the Cisco UCS B-Series Blade Servers, refer to the model
comparison at
http://www.cisco.com/en/US/products/ps10280/prod_models_comparison.html
10-55
Chapter 10
Design Considerations for Running Virtual Unified Communications Applications on B-Series Blade Servers
This section highlights some design rules and considerations that must be followed for running Unified
Communications services on virtualized servers.
Blade Server
The Cisco B-Series Blade Servers support multiple CPU sockets, and each CPU socket can host multiple
multi-core processors. For example, one B200 blade has two CPU sockets that can host up to two
multi-core processors. This provides the ability to run multiple Unified Communications applications on
a single blade server. Each Unified Communications application should be allotted dedicated processing
and memory resources to ensure that the resources are not oversubscribed.
SAN and Storage Arrays
Tested Reference Configurations based on the Cisco UCS B-Series platform require the virtual machines
to run from a Fibre Channel SAN storage array. The SAN storage array must satisfy the requirements of
the VMware hardware compatibility list. Other storage options such as iSCSI, FCoE SAN, and NFS NAS
are supported with the specification-based hardware support. For more details, refer to the
documentation available at
http://www.cisco.com/go/uc-virtualized
Design Considerations for Running Virtual Unified Communications Applications on C-Series Rack-Mount Servers
Unlike with UCS B-Series, the Tested Reference Configurations based on UCS C-Series support storage
for the hypervisor and the applications virtual machines locally on the directly attached storage drives,
not on an FC SAN storage array. It is possible to use an external storage array with a C-Series server,
but the server would then be considered as specifications-based hardware and not as a TRC.
For more details, refer to the documentation available at
http://www.cisco.com/go/uc-virtualized
10-56
Chapter 10
For MoH live audio source feed, consider using Cisco IOS-based gateway multicast MoH for live
audio source connectivity.
There is no alternative option for the Cisco Messaging Interface (CMI) service for Simplified Message
Desk Interface (SMDI) integrations.
Call Routing and Dial Plan Distribution Using Call Control Discovery
(CCD) for the Service Advertisement Framework (SAF)
The Cisco Service Advertisement Framework (SAF) is a Cisco IOS service routing protocol that can be
used to share call routing and dial plan information automatically between call processing platforms.
SAF allows non-Cisco call processing platforms (such as TDM PBXs) to partake in the Service
Advertisement Framework when they are interconnected through a Cisco IOS gateway.
The Service Advertisement Framework (SAF) enables networking applications to advertise and discover
information about networked services within an IP network. SAF consists of the following functional
components and protocols:
The SAF Client Protocol Used between SAF Clients and SAF Forwarders.
The nature of the advertised service is unimportant to the network of SAF Forwarders. The SAF
Forwarder protocol is designed to dynamically distribute information about the availability of services
to SAF client applications that have registered to the SAF network.
Services that SAF Can Advertise with Call Control Discovery (CCD)
In theory, any service can be advertised through SAF. The first service to use SAF is Cisco Unified
Communications Call Control Discovery (CCD). CCD uses SAF to distribute and maintain information
about the availability of internal directory numbers (DNs) hosted by call control agents such as Cisco
Unified CM and Unified CME. CCD also distributes the corresponding number prefixes that allow these
internal directory numbers to be reached from the PSTN ("To PSTN" prefixes).
10-57
Chapter 10
Call Routing and Dial Plan Distribution Using Call Control Discovery (CCD) for the Service Advertisement Framework
Note
SAF CCD supports the distribution of internal enterprise DN ranges only, unlike GDPR which supports
the distribution of internal enterprise DN ranges, external (PSTN) DN ranges, and URIs.
The dynamic nature of SAF and the ability for call agents to advertise the availability of their hosted DN
ranges and To PSTN prefixes to other call agents in a SAF network, provides distinct advantages over
other static and more labor-intensive methods of dial plan distribution.
The following Cisco products support the Call Control Discovery (CCD) service for SAF:
Cisco Unified Communications Manager Express (Unified CME) on a Cisco Integrated Services
Router (ISR)
CCD is supported on Cisco ISR platforms running Cisco IOS Release 15.0(1)M or higher.
For more information on SAF CCD in Unified Communications networks, refer to the SAF sections in
the Unified Communications Deployment Models chapter of the Cisco Unified Communications
System 9.0 SRND, available at
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/srnd/9x/models.html
For more information on SAF itself, refer to the Service Advertisement Framework (SAF) section in the
Network Infrastructure chapter of the Cisco Collaboration 9.x Solution Reference Network Designs
(SRND), available at
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/srnd/collab09/netstruc.html
Note
Up to 100,000 learned DN patterns per Unified CM cluster (Default value = 20,000 learned patterns)
Up to 125 advertised DN patterns per Unified CME, Cisco Unified Border Element, or Cisco IOS
Gateway
Up to 6,000 learned DN patterns per Unified CME, Cisco Unified Border Element, Cisco IOS
Gateway, or SRST (platform-dependant)
For SAF deployments using a single SAF autonomous system (AS) and consisting of Cisco Unified CM
and SAF CCD running on a Cisco IOS platform, SAF CCD system-wide scalability is limited to 6,000
learned DN patterns.
10-58
CH A P T E R
11
Conferencing is an essential component of any collaboration system, especially when serving remote
users and/or a large user base. Cisco Rich Media Conferencing offers features such as instant,
permanent, and scheduled audio and video conferencing, as well as content sharing.
Conference bridges provide the conferencing function. A conference bridge is a resource that joins
multiple participants into a single call (audio or video). It can accept any number of connections for a
given conference, up to the maximum capacity allowed for a single conference on that device. The output
display for a given party shows all connected parties minus the viewers own input.
Cisco Rich Media Conferencing solutions utilize various infrastructures to provide audio and video
conferencing capability and, in some cases, content sharing. The conferencing infrastructure can be
Cisco Unified CM using software or DSP resources, Cisco TelePresence, or Cisco WebEx Collaboration
Cloud, and this chapter covers the design details pertaining to each solution.
For a conferencing deployment using Cisco IP Voice Media Streaming Application from Unified CM or
hardware DSPs (PVDM3 or newer) hosted on a Cisco Integrated Services Router (ISR), the design
guidance remains the same as in previous versions. For details, refer to the Cisco Rich Media
Conferencing chapter in the Cisco Collaboration System 10.x SRND, available at,
http://www.cisco.com/go/ucsrnd
Cisco Rich Media Conferencing solutions are available as on-premises, cloud, or hybrid deployments.
This allows an organization to integrate with the Collaboration solution in which they have already
invested or, alternatively, to implement a service that is hosted "in the cloud." This is one of the more
important distinctions between the various solutions, and it is the first decision point when determining
which solution is the best fit for an organization.
Cisco WebEx Software as a Service (SaaS) offers a completely off-premises solution, while Cisco
Collaboration Meeting Room (CMR) Hybrid is a hybrid solution with a mix of on-premises and
off-premises equipment. Organizations that have deployed Cisco Collaboration System will benefit most
from leveraging an on-premises solution. The later sections of this chapter provide more detailed
deployment options for each conferencing solution.
Table 11-1 summarizes available solutions from an on-premises cloud perspective.
11-1
Chapter 11
Table 11-1
Audio
Solution
On-premises
Video
Cloud
On-premises
1
Content Sharing
Cloud
On-premises
Cloud
No
Yes
No
No
Yes
No
Yes
No
Yes 1
No
Yes
Yes
No
Yes
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
Yes
No
Yes
1. Cisco WebEx webcam video only, and no support with standards-based video.
To provide a satisfactory end-user experience, careful planning and design should be done when
deploying Cisco Conferencing solutions so that users are enabled with the conferencing functionality
they require.
To aid in the design, this chapter starts with an introduction of the different types of conferences
supported in the Cisco Conferencing solutions, followed by detailed discussions of the following main
topics for each solution:
Architecture
This section introduces the main components of the Conferencing solution and describes its
advantages as well as the different conferencing mechanisms available through the various
components of a collaboration system. Supported deployment models, solutions, and
recommendations are discussed here as well.
High availability
This section discusses best practices for designing a resilient Cisco Conferencing solution; it also
contains guidance for redundancy and load balancing.
Capacity planning
This section provides best practices and design information related to capacity limits and scalability
for the Cisco Conferencing solution.
Design considerations
This section discusses general recommendations and best practices for the Cisco Conferencing
solution design.
11-2
Chapter 11
Described in:
Revision Date
Cisco Collaboration Meeting Room (CMR) Cloud Cisco Collaboration Meeting Room Cloud,
page 11-69
Types of Conferences
The Cisco Rich Media Conferencing solution supports the following types of conferences:
Instant conference
An instant audio or video conference (also referred to as an ad hoc conference) is an impromptu
conference. Instant conferences are not scheduled or arranged prior to the conference. For example,
a point-to-point call escalated to a multipoint conference is considered to be an instant conference.
Permanent conference
Permanent conferences (also referred to as meet-me, static, or rendezvous conferences) are
predefined addresses that allow conferencing without previous scheduling. The conference host
shares the address with other users, who can call in to that address at any time.
Permanent conference resources are used on a first-come-first-served basis (non-assured). For a
guaranteed conference resource (assured), scheduled conferences should be used.
Scheduled conference
A scheduled conference is started by its initiator through a scheduling management system called
Cisco TelePresence Management Suite (TMS). Conferences are booked via Cisco TMS with a start
and end time and optionally with a predefined set of participants.
11-3
Chapter 11
Cisco Rich Media Conferencing consists of the conferencing solutions described below. The details
pertaining to each solution are described in each individual section that follows.
11-4
Chapter 11
Figure 11-1
Peter, Attendee
Centralized Number
1-800-553-6387
Alice, Host, PIN
Meeting ID 1000
PSTN
Gateway
Ben, Attendee
IVR
MOH
348889
Signal path
Media path
The administrator can enable the Conference Now option for a user. If enabled, the user gets a meeting
number and must configure a host PIN to start the meeting. Also, an optional attendee access code can
be configured for attendees to join the meeting. Prior to the meeting, the conference host distributes the
meeting number and the optional access code to all the participants. To start the meeting, the host dials
into Conference Now and enters both the meeting number and the host PIN. To join the meeting, the
attendees dial into Conference Now and enter the meeting number along with the optional attendee
access code. If the attendee dials into the meeting before the host, the attendee will be placed on Music
on Hold (MoH). Conference Now uses conference bridges configured in the media resource group
(MRG) and media resource group list (MRGL) associated with the host's calling device to perform the
conferencing function. Ensure that both the conference bridge and the IVR resources are available to
Unified CM in order to use the Conference Now feature.
Using a conference bridge other than the software-based Cisco IP Voice Media Streaming Application
(IPVMS) from Unified CM might not provide the conference party entry and exit tone. For the best user
experience, we recommend using the software-based Cisco IPVMS conference bridge for Conference
Now. For detail on conference party entry and exit tone support, refer to the latest version of the Feature
Configuration Guide for Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
Consider the following points when implementing Cisco Conference Now:
The IVR supports Out of Band DTMF only. Use an MTP to convert any DTMF capability mismatch.
The IVR supports G.711 (a-law and mu-law), G.729, and Wide Band 256K. IPVMS supports G.711
and Wide Band 256K. For other codec support, use a transcoder.
Conference Now does not support any advanced functionality such as a roster list or muting and
un-muting attendees.
11-5
Chapter 11
Increased efficiency by allowing instant, permanent, and CMR conferences as well as scheduled
conferences to use the same TelePresence Servers
Better user experience through advanced TelePresence Server features such as ActiveControl and
dynamic optimization of resources
Note
Cisco TelePresence MCU does not support multiparty licensing. However, if TelePresence
Conductor has multiparty licensing enabled, the TelePresence MCU can be added but it requires
screen licenses to be installed and continues to operate as port-based licenses.
TelePresence Conductor optimizes TelePresence Server resources dynamically when the Optimize
resources setting is enabled in the TelePresence Conductor conference template. This enforces maximum
resource usage of a participant based on the maximum receive bandwidth advertised by the resources at
conference join. This can reduce the amount of resources conference calls use and allows more
concurrent connections to take place. For more information, see the TelePresence Server release notes
available at
http://www.cisco.com/c/en/us/support/conferencing/telepresence-server/products-release-notes-list
.html
11-6
Chapter 11
Flexible layouts, and views optimized for the capabilities of each device
Enhanced user experience with features including Cisco TelePresence ActivePresence screen
layout, individual participant identifiers, and Cisco ClearPath
The ability to extend scale and reach to more participants to join meetings by extending meetings to
WebEx Meeting Center users
Cisco TelePresence Server is available as a virtualized application compatible with standard Cisco
Unified Computing System (UCS) servers, or you can deploy it on dedicated hardware platforms.
Flexible licensing options are offered to enable you to deploy Cisco TelePresence Server capabilities in
the way that best suits your needs. You can license Cisco TelePresence Server in conjunction with Cisco
TelePresence Conductor on a per-user basis for high-quality small group instant conferencing, with
Cisco Personal Multiparty licensing, or on a concurrent call (screen) basis to enable the whole enterprise
without restrictions.
TelePresence Management Suite user interface Provides complete control and advanced settings
and is typically used by administrators.
Cisco TelePresence Management Suite Extension for Microsoft Exchange (TMSXE) TMSXE is
an extension of TMS that enables Cisco TelePresence scheduling through Microsoft Outlook. It does
this by replicating Cisco TMS scheduled meetings to Microsoft Exchange room calendars. This
extension enables conference organizers to set up conferences using their Microsoft Outlook client.
Cisco TelePresence Management Suite Extension for IBM Lotus Notes (TMSXN) This extension
enables conference organizers to set up conferences using their Lotus Notes client.
11-7
Chapter 11
Cisco TelePresence Management Suite Extension Booking API (TMSBA) This extension gives
developers access to TMS booking functionality for custom integration with third-party calendaring
applications. This enables conference organizers to set up conferences using their existing corporate
calendaring interface.
TMSXE, TMSXN, and TMSBA are optional plug-ins installed on the calendaring server to achieve
calendar integration. Client machines do not have to be modified.
11-8
Chapter 11
Figure 11-2 illustrates a design where TelePresence Conductor manages both scheduled and
non-scheduled conference bridges, and it is the current recommended design. In this architecture,
scheduled conferences can be dedicated for scheduling as indicated in Figure 11-2 or, as indicated in
Figure 11-3, shared (not dedicated) with a pool of TelePresence Servers (General TelePresence Server
in the illustration) that are grouped together and used on a first-come-first-served basis by TelePresence
Conductor for instant, permanent, and scheduled conferences (see Deployment Considerations,
page 11-13, for more information).
Figure 11-3 shows multiple shared TelePresence Servers used for both scheduled and non-scheduled
conferences as well as Cisco WebEx conferences. In this design, conference bridges are connected via
SIP to TelePresence Conductor, and Unified CM engages TelePresence Conductor directly via SIP to
establish the video conferences.
11-9
Chapter 11
Figure 11-3
As shown in Figure 11-2 and Figure 11-3, conferencing resources can be dedicated for scheduled
conferencing or they can be shared resources for both scheduled and non-scheduled conferences when
deployed with TelePresence Conductor. There are some advantages and disadvantages of using either
method (see Table 11-3):
Dedicated TelePresence Servers Deploy one or more TelePresence Servers that are dedicated just
for scheduled conferences, with each TelePresence Server in a separate bridge pool and service
preference of its own. Optionally, a second bridge and pool combination can be used as a backup.
Shared TelePresence Servers Allow TelePresence Servers to be used for non-scheduled as well
as scheduled conferences. In this case, resource availability for scheduled conferences cannot be
guaranteed because the necessary resources might already be in use by non-scheduled conferences.
All TelePresence Servers can be configured into a single bridge pool if they are of identical size.
11-10
Chapter 11
Table 11-3
Type of
Resource
Dedicated
Advantages
Disadvantages
Shared
For video conferencing, Cisco recommends using TelePresence Conductor for centralized management
of the video conferencing solution with TelePresence Server as the conference bridge of choice for both
audio-only and video conferencing, and any combination thereof achieved through screen licenses.
For more information about screen licenses per call type, including audio-only licensing, refer to the
Cisco TelePresence Server Data Sheet available at
http://www.cisco.com/c/en/us/products/conferencing/telepresence-server/datasheet-listing.html
11-11
Chapter 11
Note
Figure 11-4
Conference bridge selection based on IP Zone will not function because TMS recognizes only IP Zones
configured for TelePresence Conductor and not for individual bridges.
Cisco TMS does not support TelePresence Conductor clustering because it can recognize only one
TelePresence Conductor cluster node, and manual intervention is required to fail-over to the backup
TelePresence Conductor node. If the primary cluster node should fail, the Cisco TMS scheduling service
will be unavailable until that node is brought back up or until Cisco TMS is updated to communicate
with a different TelePresence Conductor node in the cluster.
11-12
Chapter 11
Licensing
Cisco Personal Multiparty is a user-based licensing approach that provides instant, permanent,
scheduled, and CMR conferences that a named host (entitled user) may use. Two variations of this
licensing model exist Basic and Advanced. Personal Multiparty requires dedicated TelePresence
Servers managed by TelePresence Conductor because the Personal Multiparty licenses may not be
shared with other licensing types. TelePresence Servers licensed with Personal Multiparty do not require
screen licenses to be purchased, although licenses received through Personal Multiparty licensing must
still be applied to the devices.
Cisco Personal Multiparty Basic:
Includes a named host, four-party license for multiparty video and audio with content sharing
Supports video resolution up to 720p at 30 frames per second (fps) and content resolution up to
720p at 5 fps
Supports an unrestricted number of participants in a meeting (within the limits of the fair use policy)
Supports Microsoft Lync and business-to-business interoperability with Cisco Expressway Rich
Media Session (RMS) licenses
Deployment Considerations
The physical location of a TelePresence Server is important to consider because media traffic flows
between it and each participant in the conference. To provide the best experience for participants,
centralize the location of the TelePresence Servers in each region where they will be deployed.
Figure 11-5 illustrates the deployment differences between remotely and locally managed modes. Set the
TelePresence Servers to remotely managed mode, which is a system-wide setting that enables a more
advanced API and requires that API to be used for all operations. Remotely managed mode is the only
mode available on the Cisco TelePresence Server on Virtual Machine and Cisco TelePresence Server on
Multiparty Media 310 or 320, and it requires TelePresence Conductor, while other variants of the
TelePresence Server support both locally and remotely managed modes.
11-13
Chapter 11
Figure 11-5
Caution
In remotely managed mode, certain features are not available from the TelePresence Server interface,
and management of conferences and pre-configuration of endpoints are done at the TelePresence
Conductor level instead. Changing the operating mode requires the TelePresence Server to be rebooted,
and any conferences configured on the TelePresence Server in locally managed mode are lost when the
unit reboots.
TelePresence Conductor manages the TelePresence Servers via an XML-RPC over HTTPs API.
TelePresence Conductor also routes SIP signaling to the TelePresence Servers via its Back-to-Back User
Agent (B2BUA).
The TelePresence Server has the ability to use a secure connection for signal and secure RTP for media
transmission. To use secure RTP between TelePresence Conductor and the TelePresence Server, the
encryption feature key is required.
Interworking H.323 Endpoints into a SIP Environment
Many designs require the ability to incorporate H.323 endpoints into the architecture and to ensure
interworking with a SIP-based design. The Cisco Video Communication Server (VCS) can be used to
interwork these H.323 endpoints into a SIP-based design to provide features such as H.323-to-SIP
conversion for call control, H.239-to-BFCP conversion for content sharing, and H.235-to-SIP for SRTP
conversion. Figure 11-6 illustrates a conferencing design with VCS interworking the H.323 endpoints
with SIP to Unified CM to access the SIP-based conferencing solution.
11-14
Chapter 11
Figure 11-6
Scheduled
TelePresence Server
Expressway-C
Expressway-E
Internet
Cisco TMS
VCS
Unified CM
SIP
Media &
Content
HTTPs
General
TelePresence Server
H.323
348979
TelePresence
Conductor
Audio Conferencing
If you are deploying a large-scale video conferencing solution, Cisco recommends also deploying audio
conferencing as part of that solution. The video conferencing solution requires TelePresence Conductor
and TelePresence Server, which can also be used for audio conferencing. This reduces the complexity
and total cost of ownership of the overall conferencing solution.
Note
TelePresence Server does not support any advanced audio bridge functionality such as join or leave
tones, or any form of in-meeting control.
It is important to note that a video multipoint resource can be used for audio-only conferences, but
audio-only conferencing resources cannot be used for the audio portion of video conferences.
11-15
Chapter 11
Video Conferencing
When integrated with the Cisco CMR Premises architecture, video-capable endpoints provide the
capability to conduct video conferences that function similar to audio conferences. Video conferences
can be instant, permanent, or scheduled. This section discusses the following main topics:
Videoconferencing resources are hardware or software types, and currently the main difference between
software and hardware video resources is capacity:
Meeting Experience
The video portion of the conference can operate in one of three meeting experience modes, depending
on the conferencing device:
In addition, video conferencing can use any of the following methods to select the dominant speaker:
Automatic Participant List Cycling, page 11-17, (Not available for Cisco TelePresence Server)
Voice-activated conferences take in the audio and video streams of all the participants, decide which
participant is the dominant speaker, and send only the dominant speaker's video stream back out to all
other participants. The participants then see a full-screen image of the dominant speaker (and the current
speaker sees the previous dominant speaker). The audio streams from the participants (four in the case
of the Cisco TelePresence MCU and Cisco TelePresence Server) are mixed together, so everyone hears
everyone else, but only the dominant speaker's video is displayed. This mode is optimal when one
participant speaks most of the time, as with an instructor teaching or training a group. Speaker (segment)
switching and room switching fall under this category.
Continuous Presence
Continuous-presence conferences display some or all of the participants together in a composite view.
The view can display the participants in a variety of layouts. Each layout offers the ability to make one
of the squares voice-activated, which is useful if there are more participants in the conference than there
are squares to display them all in the composite view. For instance, if you are using a four-way view but
there are five participants in the call, only four of them will be displayed at any given time. You can make
one of the squares in this case voice-activated so that participants 5 and 6 will switch in and out of that
11-16
Chapter 11
square, depending on who is the dominant speaker. The participants displayed in the other three squares
would be fixed, and all of the squares can be manipulated through the conference control web-based user
interface, DTMF (in the case of the Cisco TelePresence MCU and Cisco TelePresence Server), and Far
End Camera Control (FECC, in the case of the Cisco TelePresence Server).On the other hand, if you are
using the "equal panels layout family," the layout would change to 3x3 when the sixth participant joins.
The audio portion of the conference follows or tracks the dominant speaker. Continuous presence is more
popular than voice switching, and it is optimal for conferences or discussions between speakers at
various sites.
Cisco ActivePresence
The Cisco ActivePresence capability of Cisco TelePresence Server enables the delivery of
next-generation multipoint conferencing by offering a view of all attendees in a meeting while giving
prominence to the active speaker. While the active speaker occupies most of the screen, an overlay of
others in the call appears in the lower third of the screen. This maintains the immersive feel of the
life-size main speakers while giving participants a more natural view of everyone else sitting around the
virtual table.
Voice Activation Mode
Using this mode, the video conference bridge automatically selects the dominant speaker by determining
which conference participant is speaking the loudest and the longest. To determine loudness, the MCU
calculates the strength of the voice signal for each participant. As conditions change throughout the
conversation, the MCU automatically selects a new dominant speaker and switches the video to display
that participant. A hold timer prevents the video from switching too hastily. To become the dominant
speaker, a participant has to speak for a specified number of seconds and be more dominant than all other
participants.
Manual Selection of the Dominant Speaker
The dominant speaker might be selected through the MCU's web-based conference control user
interface. A user with privileges to log onto the MCU's web page, highlights a participant and selects
that person as the important or dominant speaker. This action disables voice activity detection, and the
dominant speaker remains constant until the chairperson either selects a new dominant speaker or
re-enables voice activation mode.
Automatic Participant List Cycling
With this method, the MCU is configured to cycle through the participant list automatically, one
participant at a time. The MCU stays on each participant for a configured period of time and then
switches to the next participant in the list. The conference controller (or chairperson) can turn this
feature on and off (re-enable voice activation mode) from the web interface.
11-17
Chapter 11
Note
For H.323 and SIP clients with built-in MCUs, Unified CM allows functionality of the endpoints
built-in MCU only if the client is SIP.
MultiSiteTM
Certain endpoints are capable of escalating point-to-point calls with two endpoints to conferences with
three or more endpoints, without the need for an external dedicated device, and this is referred to as
MultiSiteTM. The conference created using MultiSite is considered instant because it usually happens
without prior planning and scheduling (see Figure 11-7). An option key is required to unlock the
MultiSite feature. MultiSite conferences use the embedded resources in the endpoint for the conference
creation. Endpoints with MultiSite capability that have the key installed can invoke this conference type
whether they are managed by Cisco Unified CM or Cisco VCS.
Figure 11-7
1
1
3
348980
11-18
Chapter 11
While many video endpoints on the market today are incapable of hosting conferences themselves and
require an additional device to handle mixing the multiple video and audio streams, key factors for
selecting dedicated video resources over embedded resources are bandwidth usage centralization,
scalability, and cost efficiency. These multipoint resources are shared by a number of endpoints and are
capable of hosting many conferences at the same time. The method in which a dedicated resource is
invoked depends on the endpoints and the call control device(s) involved. In the case of users utilizing
devices that can and are registered natively to Cisco Unified CM, the users can initiate an instant video
conference with dedicated resources by using the Conf, Join, or cBarge keys.
Figure 11-8 illustrates one example of an instant conference using an external resource.
Figure 11-8
Conference
Bridge Pool
348890
2
Media (Video and Audio)
The Cisco TelePresence Server and MCU, used in conjunction with Cisco TelePresence Conductor, are
the dedicated resources that enable instant calls for TelePresence endpoints controlled by Cisco
Unified CM.
11-19
Chapter 11
348891
Conference
Bridge Pool
Dial 2001
The Cisco TelePresence multipoint devices may use different names for permanent conferences,
referring to them as permanent conferences or static conferences.
11-20
Chapter 11
Dial-in conferences can optionally use an interactive voice response (IVR) system to prompt users to
enter the conference ID and the password (if one is configured) of the conference they want to join. You
can use either of the following types of IVRs with the Cisco MCUs:
If you want to have a single dial-in number and then prompt the user for the conference ID, you can use
Cisco Unified IP IVR in conjunction with the MCU. Cisco Unified IP IVR has the following
characteristics:
Can prompt for the conference ID and the password (among other things)
Can be highly customized to provide more flexible menus and other advanced functionality
Customizations can include such things as verifying the user's account against a back-end database
before permitting that user to enter into the conference, or queuing the participants until the
chairperson joins.
Note
Because Cisco Unified IP IVR supports only out-of-band signaling, it will not work with endpoints that
use in-band DTMF tones.
With Cisco Unified IP IVR, users dial a CTI route point that routes the call to the Cisco Unified IP IVR
server instead of dialing a route pattern that routes directly to the MCU. After collecting the DTMF digits
of the conference ID, the Cisco Unified IP IVR then transfers the call to the route pattern that routes the
call to the MCU. This transfer operation requires that the calling device supports having its media
channels closed and reopened to a new destination. For example, an H.323 video device that calls the
Cisco Unified IP IVR will initially negotiate an audio channel to the Cisco Unified IP IVR server and
then, after entering the appropriate DTMF digits, it will be transferred to the MCU, at which point
Unified CM will invoke the Empty Capabilities Set (ECS) procedure to close the audio channel between
the endpoint and the Cisco Unified IP IVR server and open new logical channels between the endpoint
and the MCU. If the H.323 video endpoint does not support receiving an ECS from Unified CM, it will
react by misbehaving or disconnecting the call.
11-21
Chapter 11
Launch Method
Description
Automatic Connect
All endpoints are automatically connected at the specified date and time.
Manual Connect or Hosted Conference cannot begin until a specific endpoint (usually the
conference organizer's endpoint) connects. After this endpoint connects,
the remaining endpoints are either automatically connected or allowed to
dial in manually.
No Connect
Reservation
Scheduling ensures endpoint and port resource availability and provides convenient methods to connect
to TelePresence conferences. Most organizations already use calendaring applications to schedule
conferences. In this case, calendaring integration enables users to schedule conferences with their
existing calendaring client. TelePresence deployments often include a large quantity of endpoints and
different infrastructure components. Without a centralized management component, provisioning,
monitoring, and resource allocation are difficult if not impossible. Management platforms greatly
simplify these processes.
The Cisco TelePresence scheduling and management options you choose depend on the type of
calendaring your organization uses, the type of TelePresence deployment selected or already
implemented by your organization, and the requirements or preferences of your organization.
Scheduled meetings work by integrating TelePresence resources and endpoints with corporate
calendaring applications (see Figure 11-10). Cisco TelePresence Management Suite (TMS) resides
between endpoints and calendaring applications to locate the proper multipoint resource for each
scheduled conference, and to provide resource reservation. Both the TelePresence Server and the MCU
are capable of creating scheduled conferences without the aid of the TelePresence Management Suite;
however, in this case only the multipoint device is scheduled, but the endpoints themselves are not
guaranteed to be available. For this reason Cisco recommends deploying scheduled conferencing with
the TelePresence Management Suite and creating conferences by scheduling three or more endpoints.
11-22
Chapter 11
Figure 11-10
Conference organizer
schedules 3 or more
endpoints.
Cisco TMS
348892
Conference
Bridge Pool
Conferencing Resources
A conferencing resource is the entity that performs the media conferencing, multiplexing, or media
switching functions for the conference. The actual entity that hosts a multipoint conference may reside
within a video endpoint or be separate from the endpoint in a dedicated device whose resources are
shared by many endpoints. In many customer environments both options can be deployed. It is also
important to understand that the video conferencing functionality is achieved by media transcoding and
media switching. This section provides overall guidance on the most appropriate uses of each of the
following conferencing resources:
11-23
Chapter 11
The dedicated conferencing device mixes the audio and video streams from each video endpoint and
transmits a single audio and video stream back to each endpoint. In the case of multi-screen endpoints,
multiple audio and video streams may be sent and received by the conferencing device.
Using a dedicated device has the following benefits:
Greater scalability
Enhanced feature sets (auto attendants, scheduling, presenter modes, and so forth)
The Cisco TelePresence Server is a transcoding multipoint device that is available as an appliance, a
blade running in the Cisco MSE 8000 Series chassis, or a virtual machine running on Cisco supported
virtualized platforms. The TelePresence Server can connect many video and audio devices using a
variety of protocols, including SIP, H.323, and TIP; and it supports video resolution up to 1080p.
Currently it is the only Cisco multipoint device that supports both TIP and non-TIP multi-screen
systems. The TelePresence Server is capable of instant, permanent, and scheduled conferences when
configured in remotely managed mode and installed behind Cisco TelePresence Conductor. In certain
cases when TelePresence Servers are managed by TelePresence Conductor, TelePresence Conductor
optimizes conference resources dynamically by freeing up unused resources that were initially allocated
to endpoints. Cisco recommends using Cisco TelePresence Servers for all video conferencing.
Depending on the model, Cisco TelePresence Servers can run as a cluster for redundancy or increased
capacity. For more information about clustering with TelePresence Servers, refer to the latest version of
the Cisco TelePresence Server Product User Guide, available at
http://www.cisco.com/en/US/partner/products/ps11339/products_user_guide_list.html
Cisco TelePresence Multipoint Control Unit (MCU)
The Cisco TelePresence Multipoint Control Unit (MCU) portfolio offers flexibility for a variety of video
deployments. The MCUs are designed for single-screen Cisco and non-Cisco video endpoints using
standards-based H.323 and SIP call signaling. The user's experience can vary depending upon which of
the many custom screen layouts is chosen for a particular conference. The MCU supports instant,
permanent, and scheduled conferences. Depending on the model, Cisco MCUs can support resolutions
from QCIF up to 1080p in 4:3 and 16:9 aspect ratios and can support the video modes listed in
Table 11-5.
A global configuration setting on the MCU enables one of these modes, which in turn affects the
supported resolutions and capacity. Port count depends on which mode is enabled because HD, HD+,
and Full HD settings require more hardware resources than the lower-resolution SD setting. Table 11-5
describes the differences between these options.
11-24
Chapter 11
Table 11-5
Mode
Description
Full HD
The number of 1080p30 or 720p60 (symmetric) video ports available. This allows the
MCU to transmit and receive at these resolutions.
HD+
The number of 1080p30 or 720p60 (asymmetric) video ports available. This mode
allows the MCU only to transmit at these resolutions. This mode is not available for
Cisco TelePresence MCU 5300 Series.
HD
The number of 720p30 or w448p60 (symmetric) video ports available. This mode
allows the MCU to transmit and receive at these resolutions.
SD
The number of standard definition (up to 448p) symmetric video ports available.
nHD
Cisco recommends using dedicated video resources for greater conference scalability and when a large
number of endpoints will be deployed in the organization. Cisco TelePresence MCUs can be pooled
together and managed by Cisco TelePresence Conductor for increased flexibility.
11-25
Chapter 11
Feature Comparison
When deciding which multipoint device to deploy, it is important not only to understand their audio and
video capabilities, but also to know which features are supported on each device. Table 11-6 summarizes
feature support for the multipoint devices.
Table 11-6
Feature
TelePresence
Server
MCU
Yes
Yes
TIP support
Yes
No
CMR Hybrid
Yes
Yes
Individual transcoding
Yes
Yes
TelePresence Management
Suite scheduling
Yes
Yes
ActivePresence
Yes
No
Custom layouts
Yes
Yes
Yes
Yes
Director controls
No
No
Clustering
Yes
Yes
Cascading
Yes
Yes
Lecture Mode
No
Yes
Content sharing
Yes
Yes
Auto attendant
Yes
Yes
Resource Optimization
Yes
No
Virtualization Support
Yes
No
Comments
11-26
Chapter 11
Limited scale. The host endpoint has to mix the audio and video from every other endpoint;
therefore, the size of the conference is limited by that host endpoint's capacity.
Endpoints with this capability require more bandwidth than other endpoints.
Depending on the endpoint model, there might be limitations in video resolution, which can degrade
the overall user experience when compared to calls hosted on a dedicated multipoint device.
For the reasons listed above, Cisco recommends using the MultiSite feature only when limited size
instant conferences are needed in the organization. Careful analysis also should be done to weigh the
cost benefits of the number of endpoints requiring the MultiSite key versus the benefits of a full MCU.
Table 11-7 indicates which devices are capable of MultiSite, and thus of hosting multipoint conferences.
Table 11-7
Endpoint or Codec
SX20 Codec
576p
4 video + 1 audio
SX80 Codec
720p
5 video + 1 audio
1080p
4 video + 1 audio
C40 Codec
576p
4 video + 1 audio
C60 Codec
720p
4 video + 1 audio
C90 Codec
1080p
4 video + 1 audio
576p
4 video + 1 audio
720p
5 video + 1 audio
1080p
4 video + 1 audio
720p
4 video + 1 audio
EX90
11-27
Chapter 11
Endpoint
creates an
instant
conference by
joining 3
participants
Figure 11-12
Unified CM
initiates an
instant
conference by
using the
MRG and/or
MRGL
Unified CM
routes the call
to conductor
Conductor
accepts the
call and
creates a
conference
Conductor
routes the call
to the
conference
bridge
348807
Unified CM
matches the
route pattern
by using the
route list
and/or route
group
Endpoint dials
a permanent
conference
number
Unified CM
routes the call
to conductor
via a SIP trunk
construct
Conductor
matches the
called number
to an alias and
creates a
conference
Conductor
routes the call
to the
conference
bridge
348808
Figure 11-11
Once the steps in Figure 11-11 or Figure 11-12 are complete, the call is set up and media flows between
the endpoint and the conference bridge.
In the case of MCU management, TelePresence Conductor can automatically cascade an active
conference to a separate MCU to expand total capacity, and this is transparent to the users.
Because of its inherent high availability, Cisco TelePresence Conductor is well suited for organizations
where video conferencing resiliency has a premium value and organizations with a large number of
multipoint control units.
High Availability
Proper design of the CMR Premises infrastructure requires building a robust and redundant solution
from the bottom up. By structuring the solution with redundancy (redundant media resources groups,
redundant route groups, Cisco TelePresence Conductor, and redundant media resources), you can build
a highly available, fault tolerant, and redundant solution. The following sections provide design
guidance for high availability:
11-28
Chapter 11
Conference bridge selection by conference type - either secure, video, or audio conferences
Optimized use of MCU video ports that potentially would have been used for audio-only
conferences with other methods of bridge selection
All the conference bridge resources and MCUs can be in one MRGL, and Intelligent Bridge Selection
will then select the conference bridge based on the need to do just an audio conference or a video
conference.
Unified CM also supports an alternate way of selecting conference bridges, which can be specified by
service parameter configurations. In this mode, Unified CM applies the following criteria to select the
conference bridge resource to use, in the order listed here:
1.
The priority order in which the media resource groups (MRGs) are listed in the media resource
group list (MRGL)
2.
Within the selected MRG, the resource that has been used the least
If the MCU is placed at the top of the MRGL for the phone, the MCU will always be chosen even for
audio-only conferences that do not involve any video-capable participants. In this scenario, the MCU
resources might be wasted on audio-only conferences and not be available to satisfy the request for a
video conference when it occurs. This mode, however, is not recommended because it removes the
intelligence from Unified CM to select the right resource on every conference made and should be used
only by administrators who are aware of the system-wide effects of this service parameter setting.
For further information about media resource design, see the chapter on Media Resources, page 7-1.
Note
11-29
Chapter 11
The TelePresence Conductor has the ability to manage conferencing resources such as the TelePresence
Server and TelePresence MCUs. Within its configuration it has the ability to group those resources into
pools. These pools of conferencing resources contain similar types of devices: TelePresence MCUs must
be in a pool with other TelePresence MCUs only, and TelePresence Servers must be in a pool with other
TelePresence Servers. In addition, it is best practice to try to differentiate the MCUs by HD or SD to
allow for more granularity of the video services. This granularity will allow the administrative staff to
assign endpoints such as the Cisco Unified IP Phone 9971 to the SD resources while making the HD
resources available for the high definition endpoints such as the Cisco TelePresence Series devices. In
addition, all other multi-screen and immersive systems such as Cisco TelePresence MX and IX Series
endpoints can be directed by the Conductor to use the TelePresence Server, which supports
ActivePresence and TIP, to maintain the immersive experience for all participants.
The ability to group conferencing resources by pools enables the Conductor to add redundancy at the
pool level. Pool-level redundancy is achieved by having more than one device of a given type in the pool.
The Conductor will load-balance the conference placement among the bridges residing in the pool.
In addition, the Conductor has the ability to create an ordered list of pools, which is called a service
preference. It is similar to a route list or media resource group list in Unified CM. At this level of
configuration, the Conductor can use primary and secondary pools to create a redundant model of
conferencing resources. For example, if the service preference has a list of MCU pools in this order:
Pool 1 for the US and Pool 2 for EMEA (see Figure 11-13), then the Conductor will use all the resources
in the US pool of devices first and, if needed, it can make a cascaded link between the MCU in Pool 1
and the MCU in Pool 2. The Conductor does this automatically as a part of its intelligent conferencing
selection process.
11-30
Chapter 11
Figure 11-13
Service Preference TS
Pool 1
Pool 2
USA_TS
US_MCU_#1
UK_MCU
UK_TS
US_MCU_#2
Italy_MCU
348699
Pool 1
The TelePresence Conductor can also be configured to use multiple Conductors for redundancy at the
system level. This is achieved by clustering together multiple Conductors. A maximum of three
TelePresence Conductors can be used in a clustered design. Cisco recommends at least two Conductors
in all designs to ensure high availability of all the conferencing resources.
Clustering Conductors requires a low-latency connection with less than 30 ms round trip between the
Conductor nodes. During the clustering process, the Conductors performs a database synchronization of
all table entries, such as aliases, templates, service preferences, pools, and present state table of the
conference resource ports. The initial clustering process has to have a primary Conductor that is used as
the initial database for the cluster. Once the initial process is done, any Conductor in the cluster can
update the database that is synchronized across all the Conductor nodes.
For more details on clustering, refer to the Cisco TelePresence Conductor deployment guides available at
http://www.cisco.com/en/US/products/ps11775/products_installation_and_configuration_guides_l
ist.html
When integrating Unified CM with the Conductor cluster, redundancy is achieved by having multiple
connections to Unified CM for the instant and permanent conference calls. When the Conductor is
clustered in this design, it requires unique virtual IP addresses that are the termination points for the
Unified CM conference bridge and SIP trunk. Because the virtual IP addresses are unique on each
Conductor cluster node, this information cannot be replicated in the database synchronization process
and needs to be configured by the administrator. Once this configuration is complete, Cisco
recommends using different Conductor cluster nodes for the primary, secondary, or tertiary links to
Unified CM.
For example, as illustrated in Figure 11-14, a Unified CM should be configured to use the virtual IP
address of Conductor_1's instant conference configuration as the primary connection, and the secondary
connection is set to the virtual IP address of the instant conference configuration on Conductor_2. For
the permanent calls, the destination address of the primary SIP trunk should be Conductor_2's virtual IP
address for the permanent conference configuration, and the secondary connection is set to the virtual
IP address of the permanent conference configuration on Conductor_1.
11-31
Chapter 11
Figure 11-14
Unified CM Cluster
Cond_1
Instant
Permanent
10.10.10.10
10.10.10.20
10.10.10.30
Conference Bridge
Resources
Instant Bridges
Permanent
Bridges
g
Cond_2
Instant
Permanent
10.10.10.40
10.10.10.50
10.10.10.60
Primary Instant
Secondary Instant
Secondary Permanent
Note
348893
Primary Permanent
Additional redundancy can be achieved within Unified CM for instant and permanent conference calls.
For more information, refer to the sections on Media Resource Groups and Lists, page 11-29, and Route
List and Route Groups, page 11-30.
11-32
Chapter 11
Capacity Planning
Capacity planning is critical for successful conferencing deployments. Given the many features and
functions provided by the conferencing services as well as the many different types of media resources
used as part of the architecture, it is important to size the conferencing infrastructure and its individual
components to ensure they meet the capacity needs of a particular deployment.
This section provides information and best practices for sizing the media resources used in the
conferencing architecture.
Geographical location Each region served by Unified CM should have dedicated conferencing
resources. For example, there could be one central location for the US where Unified CM,
TelePresence Servers, and other servers are installed, and one central location for EMEA.
TelePresence Server platform capacities For capacity details, refer to the latest version of the
Cisco TelePresence Server release notes available at
http://www.cisco.com/c/en/us/support/conferencing/telepresence-server/products-release-note
s-list.html
TelePresence Conductor platform capacities For capacity details, refer to the latest version of the
Cisco TelePresence Conductor release notes available at
http://www.cisco.com/c/en/us/support/conferencing/telepresence-conductor/products-releasenotes-list.html
Conference resources are generally dedicated to a region in order to keep as much of the conference
media on the regional network; therefore, sizing can be considered on a region-by-region basis. To
properly size the conferencing deployment, use the Cisco Collaboration Sizing Tool, available at
http://tools.cisco.com/cucst
Note
Consult with your Cisco sales representative for assistance on sizing the conferencing resources for your
particular environment.
11-33
Chapter 11
If an instant conference that does not use the MultiSite or built-in bridge functionality is initiated by
an endpoint registered to Unified CM, the resource is allocated based on the media resource group
and list assigned to the conference initiator. All other endpoints in the conference just join their
streams to the resource selected by the conference initiator. (See Figure 11-15.)
If a conference is initiated by an endpoint registered to Unified CM, the resource is allocated based
on the media resource group and list assigned to the conference initiator. All other endpoints in the
conference just join their streams to the resource selected by the conference initiator. (See
Figure 11-15.)
If a permanent video conference is created, the conference call (all participants' individual calls) will
be processed based on the dial plan applicable to the call (permanent alias or number dialed) and
off-loaded to the applicable trunk multipoint device.
If a scheduled conference is created, the conference call (all participants' individual calls) will be
processed based on the dial plan applicable to the call and off-loaded to the applicable trunk
multipoint device.
The resource allocation logic for scheduled conferencing lies in the scheduling platform resource
selection algorithm.
In Figure 11-15, assume that the conference was initiated by endpoint A, which has its local resource as
the first option in the media resource group list. Note how, for this particular conferencing example, the
usage of the local resource is bandwidth intensive because there are more remote users than local
participants in the conference, thus causing the remote users streams to traverse the WAN.
Conference Initiated by Endpoint A
346455
Figure 11-15
For further information about the processing of media resources groups and lists, refer to the section on
Media Resource Groups and Lists, page 11-29, and the chapter on Media Resources, page 7-1.
11-34
Chapter 11
Scalability
Cisco recommends following the guidelines in this section to scale the design of your conferencing
deployment.
Media Resource Groups and Lists
Make use of media resource groups and lists whenever possible to achieve the desired scalability for
instant and MeetMe conferences in a Cisco Unified CM deployment. For example, see Figure 11-16,
where two MCUs are used to double the instant video conferencing capacity of Device Pool 2 by adding
the two containing media resource groups into the media resource group list used by the pool. For more
information on media resource groups and lists, see the chapter on Media Resources, page 7-1.
Figure 11-16
11-35
Chapter 11
Make use of Cisco TelePresence Conductor to scale video conferencing services. TelePresence
Conductor offers orchestration for multiple resources so that they can be allocated as needed, and it can
span conferences through more than one multipoint resource when its capacity is exceeded. Figure 11-17
depicts an example of how a single Conductor server can improve the scalability of a deployment for
instant video conferencing. In this example the resource allocation starts when the user (1) initiates a
conference on Cisco Unified CM. This in turn sends the request (2) to the Conductor (3). The Conductor
determines which is the most available resource to be utilized and creates the conference in the MCU
(4). It then replies to Unified CM with the conference details (5), and Unified CM starts the conference
signaling negotiations with the MCU (6). The media (RTP) flows point-to-point (directly) between the
endpoint(s) and the MCU. TelePresence Conductor pools the resources, thus making all those collective
resource available to Unified CM as if they were a single resource.
Figure 11-17
Unified CM
3
4
Multipoint Devices
348894
For further information about Cisco TelePresence Conductor, see the section on Redundancy with Cisco
TelePresence Conductor, page 11-30. The information and logic documented in that section also apply
to increasing the scalability in a deployment.
Clustering and Cascading
Make use of clustering and/or cascading if your multipoint device supports it. (For the definition and
functionality of cascading, see the section on Cascading, page 11-41.) Clustering multipoint blades in a
Cisco TelePresence MSE 8000 series chassis has resource implications because it can triple or quadruple
the amount of bandwidth that is required for that chassis. On the other hand, although cascading can be
used to increase conference capacity while maintaining a distributed multipoint deployment, note that
cascading can create an inconsistent meeting experience. Cascading is automated by Cisco TelePresence
Conductor.
Table 11-8 summarizes information about which multipoint devices are capable clustering or cascading.
11-36
Chapter 11
Table 11-8
Device
Clustering1
Cascading
Yes
Yes
Cisco MCU
Yes
Yes
1. Requires Cisco TelePresence Servers running on physical hardware, MCU 5300 Series, or MCU MSE 8510 blades in an
MSE 8000 Series chassis.
2.
The MCU 5300 can be.configured for clustering when configured in a stack.
Design Considerations
This section provides general guidance and recommendations for the conferencing design.
11-37
Chapter 11
Figure 11-18
QoS Enabled
WAN
MX200 G2
DX80
San Jose
London
IX 5000
New York
348981
Multipoint
Device
11-38
Chapter 11
London
Dallas
Multipoint
Device
Multipoint
Device
Paris
New York
MX300 G2
San Jose
SX Series
DX70
Madrid
348982
IX 5000
QoS
Enabled
WAN
11-39
Chapter 11
Cisco Unified CM SME Deployment with All Leaf Clusters in a Single Region
Session Management Edition
Conductor
TelePresence
Bridge
TelePresence
Bridge
Unified CM
Cluster
SIP Trunk
SIP/XML RPC
Unified CM
Cluster
Unified CM
Cluster
348895
TelePresence
Bridge
Figure 11-20 shows an example of a Cisco Unified CM SME deployment where all the leaf clusters are
in a single region. Within the region, the TelePresence Conductor cluster is trunked to the SME for
permanent and scheduled conferences; this is to simplify the routing process and minimize the number
of required SIP trunks. In this configuration, the leaf clusters direct the default route (SIP route pattern
and numeric route pattern) to the SME, which sends the call to the Conductor. For instant conferences,
each leaf cluster requires a connection to the Conductor because conferences are initiated from the leaf
clusters.
11-40
Chapter 11
Design Recommendations
This section provides general recommendations for conferencing deployments.
Latency
For an optimal and natural experience regardless of deployment model, multipoint devices should be
located at sites with one-way network latency of less than 150 ms between the multipoint device and any
endpoints, provisioned with adequate port capacity and provided with ample bandwidth for the number
of provisioned conferencing ports. Bandwidth requirements vary depending on the required maximum
call rate of endpoints and the number of endpoints connecting to the multipoint device. Provision based
on the maximum bandwidth that a particular endpoint requires for the desired call rate and resolution.
For details, refer to the specific endpoints' data sheets available on http://www.cisco.com.
Cascading
Cascading refers to the ability to bridge together conferences on two multipoint devices as one
conference. This increases the maximum number of endpoints supported in a conference and can reduce
bandwidth across links, depending on how the multipoint devices are deployed. Note that cascading
conferences can create an inconsistent user experience, especially when features such as continuous
presence are used by the endpoints. This is due to the remote conference appearing as just another
endpoint on the local conference.
When making use of cascading for a large briefing conference type, it is a best practice to have all slave
MCUs set to full-screen voice switched, and the master MCU set to the desired Continuous Presence
layout. All the main speakers should be located in the master MCU so that they can be provided with the
best experience.
11-41
Chapter 11
Architecture
Cisco WebEx SaaS utilizes the Cisco WebEx Collaboration Cloud to deliver the conferencing solution
to the customers. The Cisco WebEx Collaboration Cloud is a global network created with a carrier-class
information switching architecture, and only Cisco Collaboration traffic flows over this network.
Figure 11-21 shows the Cisco WebEx Collaboration Cloud architecture.
Cisco WebEx Collaboration Cloud Architecture
Meeting
Center
Training
Center
Event Center
Support
Center
Multi-Layer Security
Session Manager
Session Manager
Web Zone
SSL
Collaboration
Collaboration
Collaboration
GSB
Meeting Zone
AES
Switches
Switches
Switches
Global
Watch
Network Layer
GSLB
Data Centers
Data Centers
Figure 11-21
Data Centers
348747
Physical Layer
11-42
Chapter 11
This network is purpose-built for real-time communications and has been specially formulated to
minimize latency associated with TCP-layer flows. The network consists of application-specific
multimedia switches at key peering points to handle rapid session traffic and to guarantee a high quality
of service for WebEx meetings. These switches are housed in highly secure Cisco data centers
interconnected via dedicated lines that circumvent the public internet. These data centers are located
near the major internet access points to route meeting traffic around the globe securely and reliably. In
addition to these large data centers housing major meeting nodes, Cisco deploys nodes around the world.
The network is built on fully redundant clusters with Global Site Backup. These services and other
facilities form part of the Cisco WebEx Collaboration Cloud Operational Support System.
Users can connect to a WebEx meeting using the meeting client running on the computer or mobile
device. Once the connection is established, the WebEx Collaboration Cloud manages all synchronous
real-time interactions that make up a WebEx meeting, as depicted in Figure 11-21. Users access WebEx
applications via browsers through the WebEx Collaboration Cloud, which resides within the Web Zone.
The Applications Program Interface (API) ties the WebEx applications to the switching platform in the
Meeting Zone within the WebEx Collaboration Cloud core. Numerous clusters of interconnected and
distributed collaboration switches, their associated databases, and the logical and physical network
infrastructure make up the WebEx Collaboration Cloud core. Multi-layer security components and the
WebEx Operational Support System encircle the network with an additional layer of protection.
The WebEx Collaboration Cloud delivers real-time traffic reliably using intelligent routing, Global Site
Backup (GSB), and Global Server Load Balancing (GSLB). Based on the geographic location of WebEx
meeting participants, the WebEx Collaboration Cloud determines the point of presence that offers the
lowest latency and best performance. WebEx meeting hosts automatically get a backup site physically
located in a geographically distant Cisco data center within the same region. In the unlikely event that
the primary WebEx site becomes unavailable, GSB automatically switches all meeting activity to the
backup site. GSLB is a load-balancing design that directs traffic to the least congested switch in the
WebEx Collaboration Cloud in order to minimize the delays. Thus, if one meeting switch has congestion,
traffic is directed to an alternate switch, resulting in faster screen updates and synchronization among
participants, and a better meeting experience.
In the WebEx deployment model shown in Figure 11-22, all the content, voice, and video traffic from
every client traverses the internet and is mixed and managed in the cloud at the WebEx data center. The
WebEx data center is logically divided into the Meeting Zone and the Web Zone. The Web Zone is
responsible for things that happen before and after a web meeting. It incorporates tasks such as
scheduling, user management, billing, reporting, and streaming recordings. The Meeting Zone is
responsible for switching the actual meeting once it is in progress between the endpoints.
11-43
Chapter 11
Figure 11-22
WebEx Deployment
Enterprise
WebEx Clients
Cisco
webex
WebEx Client
Web Zone
Servers
WebEx
Database
Recording
Enterprise
Network
Internet
Meeting Zone
Session Initiation
Meeting traffic
VoIP/Video
Multimedia
Platform
253832
Collaboration
Bridge
The Meeting Zone consists of two subsystems. Within the Meeting Zone there are collaboration bridges
that switch meeting content. The multimedia platform is responsible for mixing all of the VoIP and video
streams within a meeting. To join a WebEx session, an attendee first connects to the Web Zone. The Web
Zone traffic flows only before or after the meeting, is relatively low bandwidth, and is mainly non-real
time. The real-time meeting content share flows to and from the Meeting Zone and can be bandwidth
intensive. Its real-time nature can place a heavy burden on enterprise access infrastructure. For further
details regarding network traffic planning, see Capacity Planning, page 11-49.
Meeting Center uses the H.264 AVC/SVC codec to provide high-definition video for the conference.
Higher network bandwidth is needed for those deployments. For further details regarding network traffic
optimization for high-definition video, see Capacity Planning, page 11-49.
Starting with Cisco WebEx Meeting Center WBS29.11, each WebEx Meeting Center host has a Personal
Room with a fixed customizable URL. The host can use his room to conduct meetings, and participants
can enter the room using that fixed URL. Lobby management functions are available to the host, such as
maintaining privacy by locking the room to prevent others from entering while the meeting is in progress.
Starting with Cisco WebEx Meeting Center version WBS30, Meeting Center can be integrated with
Cisco Spark for post-meeting collaboration for any host who has both WebEx and Spark entitlement.
After the meeting ends, the host is offered the option to create a Spark Room with all the invitees and
11-44
Chapter 11
attendees pre-populated. The host can then use this room to upload meeting minutes and perform further
discussion and collaboration after the meeting. This capability applies to both scheduled and personal
room meetings. For more information about Cisco Spark, refer to the Cisco Spark documentation at
http://support.ciscospark.com
For details on IM and Presence services delivered by WebEx Collaboration Cloud, see the chapter on
Collaboration Instant Messaging and Presence, page 20-1.
Security
By default, all WebEx meeting data is encrypted using 128-bit SSL encryption between the client and
Cisco's Collaboration Cloud. SSL accelerators within the cloud decrypt the content sharing information
and send it to a WebEx conference bridge that processes the content and sends it back through an SSL
accelerator, where it is re-encrypted and sent back to the attendees. All Web Zone and Meeting Zone
traffic is encrypted using 128-bit SSL where SSL accelerators are used to off-load the SSL function from
the Web and Meeting Zone servers.
After the meeting ends, no session data is retained in the WebEx cloud or an attendee's computer. Only
two types of data are retained on a long-term basis: billing and reporting information and optionally
network based recordings, both of which are accessible only to authorized enterprise users.
Some limited caching of meeting data is carried out within the Meeting Zone, and this is done to ensure
that users with connectivity issues or who may be joining the meeting after the start time receive a
current fully synchronized version of the meeting content.
Independent third parties are used to conduct external audits covering both commercial and
governmental security requirements, to ensure the WebEx cloud maintains its adherence to documented
security best practices. WebEx performs an annual SSAE 16 audit in accordance with standards
established by the AICPA, conducted by Price Waterhouse Coopers. The controls audited against WebEx
are based on ISO-27002 standards. This highly respected and recognized audit validates that WebEx
services have been audited in-depth against control objectives and control activities (that often include
controls over information technology and security related processes) with respect to handling and
processing customer data.
For customers that require enhanced security, there is also an option to perform end-to-end 256 bit AES
encryption for collaboration bridge and multimedia content so that traffic is never decrypted in the cloud.
End-to-end encryption results in some lost features such as NBRs. For more information on enhanced
WebEx security options, refer to the white paper Unleash the Power of Highly Secure, Real-Time
Collaboration, available at
http://www.cisco.com/en/US/products/ps12584/prod_white_papers_list.html
Note
End-to-end encryption options are available for Meeting Center and Support Center meetings without
additional cost.
Scheduling
With respect to scheduling and initiating meetings, WebEx provides cloud-based web scheduling
capability, but most organizations prefer to schedule from their corporate email system (Exchange, Lotus
Notes, and so forth) or other enterprise applications. The WebEx Productivity Tools is a bundle of
integrations with well known desktop tools incorporated into a single application. A WebEx
administrator can control the specific integrations that are provided through the tool to their
11-45
Chapter 11
organization's user population. It can be downloaded and installed from the WebEx site, or it can be
pushed out locally using standard desktop management tools. For more information on WebEx
Productivity Tool, refer to the WebEx Productivity Tools FAQs, available at
https://welcome.webex.com/docs/T27LD/mc0805ld/en_US/support/productivitytools_faq.htm
User Profile
There are several options for creating WebEx user profiles for an organization in the cloud. Security
considerations for the actual usernames and passwords, as well as for handling a large number of user
accounts, should be considered. A WebEx administrator can create user profiles manually by bulk import
of a CSV template or by a programmatic approach. A programmatic approach uses one or a combination
of the WebEx APIs, URL, and XML, or a Federated SSO solution. The programmatic approach can be
used by a customer portal, which is an application such as a CRM tool or a Learning Management
System that integrates directly into WebEx. In addition, the user can sign up for an account from the
company's WebEx site, and the user profile will be created after the request has been approved.
For integrating directly with an organization's LDAP directory, Federated SSO with Security Assertion
Markup Language (SAML) is the preferred approach. For more information regarding Federated SSO,
refer to the white papers and technical notes available at
https://developer.cisco.com/site/webex-developer/develop-test/sso/reference/
High Availability
The Cisco WebEx Collaboration Cloud has a very high level of redundancy built in and is managed by
Cisco. It is designed for continuous service with a very robust cut-over to the redundant meeting nodes
during outages. In addition to the primary WebEx site, every customer has a backup site physically
located in a geographically distant WebEx data center within the same region. If a customer's primary
site is unavailable, Global Site Backup (GSB) automatically moves all meeting activity to the backup
site. Neither the hosts nor the participants notice that they are being redirected to the backup site. The
GSB system facilitates continuous accessibility to WebEx meetings globally, and all attributes, address
books, preferences, meeting schedules, and other real-time data are kept in sync between the primary
and backup sites. Because of this synchronization, GSB provides redundancy and disaster recovery both
before and after the meetings.
11-46
Chapter 11
Figure 11-23
WebEx Audio
Bridge (Acve)
CUBE
CUBE
SIP
SIP
Acve
Peering Link
WebEx
Customer Site
Peering
Router 1
Peering
Router 2
SIP
Cisco Unied
Border Element 1
(Acve)
Transport
SIP
CUBE
CUBE
Customer
Data Center 1
SIP
Cisco Unied
Border Element 2
(Standby)
Customer
Data Center 2
346448
Cisco Unied
CM Cluster
As shown in Figure 11-23, a typical WebEx CCA high-level design consists of the on-premises IP
telephony network and the WebEx cloud infrastructure that are connected via the dedicated IP Peering
Connections provided by the customer. The on-premises IP telephony network consists of a Cisco
Unified Communications Manager (Unified CM) cluster and Cisco Unified Border Element. Cisco
Unified Border Elements are deployed in the WebEx cloud infrastructure and they mark the entry point
for an organization's IP telephony network. The Cisco Unified Border Elements in the cloud and at the
customer site communicate with each other via SIP. WebEx CCA requires the customer to have two IP
Peering Connections that connect with different WebEx data centers residing in geographically
separated locations for redundancy purpose. The redundant IP links are configured in active/standby
mode. All conferencing audio traffic flows through the primary link and fails-over to the secondary link
if the primary link goes down. WebEx CCA also requires the gateway routers to support Border Gateway
Protocol (BGP) and Bidirectional Forwarding Detection (BFD) protocol. BGP and BFD offer a
significant faster re-convergence time in the event of a network failure.
Note
The WebEx data center equipment, audio bridge, and servers run over the shared infrastructure along
with other customers in the WebEx CCA solution.
11-47
Chapter 11
Cisco Unified CM has a SIP connection with the WebEx cloud through the Cisco Unified Border
Element at the customer site to handle telephony signal. The conference dial-in number is owned by the
customer and is terminated at the customer site. Call routing is handled at customer the site, call
signaling and audio traffic is handled over the redundant IP peering connections, and call mixing is
handled in the cloud. When users dial the conference number within the enterprise, Cisco Unified CM
routes the call over the dedicated SIP trunk through the Cisco Unified Border Element to the WebEx
cloud without traversing through the PSTN. When the conference users request callback, WebEx sends
the call to the Cisco Unified Border Element at the customer site that routes it to the destination
end-point. If the conference users reside outside of the enterprise network, calls are routed through the
PSTN before terminating or after leaving the customer's IP telephony network. WebEx CCA supports
only the G.711 audio codec, RFC 2833 DTMF, and SIP signaling.
WebEx CCA has the highly available and fully redundant architecture that is designed to ensure
continuous service operation. Every major component has two instances in active and standby mode,
backing up each other. There are two IP Peering Connections handled by two independent pairs of
routers, two pairs of Cisco Unified Border Elements, and two audio conferencing bridges. If any of these
components fails, its standby counterpart takes over. If the active peering link fails, the network will
converge via the standby connection. All existing calls continue, but with a very brief interruption of the
media flow. Cisco Unified Border Elements use the Out-of-Dialog OPTIONS ping mechanism to
monitor the operational state of each other. Cisco Unified Border Elements at the customer site also
monitor the Cisco Unified CM cluster using the Out-of-Dialog OPTIONS ping mechanism. Failure in
responding to the ping results in removal of the unresponsive element from the dial-peer list of the
sender, which commences routing all new calls via the standby instance. In case the active WebEx audio
bridge fails, all calls associated with the bridge are terminated and the standby WebEx audio bridge is
activated. WebEx will then prompt the users with a new number to connect to the newly activated bridge,
which also re-dials all system-originated calls (callbacks) from before the failure.
Consider the following guidelines when deploying Cisco WebEx Cloud Connected Audio:
Cisco recommends using Cisco Unified CM 8.5 or later release with the WebEx CCA deployment.
Cisco recommends using a dedicated Cisco Unified Border Element for the WebEx CCA
deployment to ensure a sound architecture and easy troubleshooting.
Cisco Unified Border Element can be deployed on either a Cisco Integrated Services Router (ISR)
or an Aggregated Services Router (ASR), depending on the audio port capacity requirements.
Use an access control list (ACL) instead of packet inspection to restrict traffic in the firewall on the
IP Peering link.
The system administrator must provide at least one toll and one toll-free number for guest dial-in.
If an audio codec other than G.711 is desired, use a transcoder to transcode the audio stream to
G.711 before sending it to WebEx.
One Direct Inward Dialing (DID) Digital Number Identification Service (DNIS) must be passed to
the WebEx cloud via the Cisco Unified Border Element for all conferencing numbers.
For more information on Cisco WebEx Cloud Connected Audio, refer to the documentation available at
http://www.cisco.com/go/cwcca
11-48
Chapter 11
Capacity Planning
For a given customer, the actual number of concurrent meetings is essentially unlimited. Different
WebEx conferencing types have different capacities with respect to number of attendees. For a detailed
product comparison table, refer to the Cisco WebEx Web Conferencing Product Comparison, available at
http://www.cisco.com/en/US/prod/ps10352/product_comparison.html
Design Considerations
Observe the following design considerations when implementing a Cisco WebEx SaaS solution:
Collaborative meeting systems typically result in increased top-of-the-hour call processing loads.
Cisco partners and employees have access to capacity planning tools with parameters specific to
collaborative meetings to help calculate the capacity of the Cisco Unified Communications System
for large configurations. Contact your Cisco partner or Cisco Systems Engineer (SE) for assistance
with sizing of your system. For Cisco partners and employees, the Cisco Unified Communications
Sizing Tool is available at http://tools.cisco.com/cucst.
All connections from WebEx clients are initiated out to the cloud. Typically, opening pinholes in
network firewalls is not required as long as the firewalls allow intranet devices to initiate TCP
connections to the Internet.
Provision sufficient bandwidth for conference video and data traffic. See Network Traffic Planning,
page 11-57, for details.
Based upon business requirements, design decisions have to be made about the following:
User creation and authentication options (see User Profile, page 11-46, for details)
Meetings scheduling options (see Scheduling, page 11-45, for details)
Cisco WebEx SaaS uses the multi-layer security model, and security extends from the WebEx
infrastructure to the organization and individual meeting layer. There are various security options
available, and depending on the business requirements., an organization can implement different
levels of security. For security options and considerations, refer to the white paper Unleash the
Power of Highly Secure, Real-Time Collaboration, available at
http://www.cisco.com/en/US/products/ps12584/prod_white_papers_list.html
11-49
Chapter 11
For more details on the various Cisco Collaboration client offerings and how they fit into Cisco
conferencing solutions, see the chapter on Collaboration Endpoints, page 8-1.
Utilize Unified CM's SIP trunk secure connection support for secure conferencing
Leverage Unified CM's dual stack (IPv4 and IPv6) capability to support IPv6
Architecture
Cisco WebEx Meetings Server is a fully virtualized, software-based solution that runs on Cisco Unified
Computing System (UCS). It uses the virtual appliance technology for rapid deployment of services.
Virtual appliance simplifies the task of managing the system. For example, using the hypervisor
technology, system components can easily be moved around for maintenance, or system components can
easily be rolled back to a working version if problem arises. The virtual appliance is distributed in the
form of an industry standard format, Open Virtual Appliance (OVA). All the software components
required to install WebEx Meetings Server are packaged inside the OVA. Traditionally, using an
executable installer to install individual software components would take hours to deploy the software.
However, using OVA can significantly reduce the amount of time required to deploy the software
because all software components are pre-packaged inside the file. Thus, virtual appliance technology can
help tremendously to reduce the deployment time for Cisco WebEx Meetings Server.
Figure 11-24 shows the high-level architecture for Cisco WebEx Meetings Server using the non-split
horizon network topology. (For details on the non-split horizon network topologies, refer to the Cisco
WebEx Meetings Server Planning Guide, available at
http://www.cisco.com/en/US/products/ps12732/products_installation_and_configuration_guides_list.h
tml.) Inside the virtual appliance, there could be one or more virtual machines (VMs) running. These are
the administration, web, and media virtual machines. The administration and web virtual machines serve
as the back-end processing for the administration and WebEx sites. These sites handle tasks that happen
before and after the meeting, such as configuration, scheduling/joining meetings, and recording
playback. The media virtual machine provides resource allocation, teleconference call control, and
media processing (voice, video, and data) during the meeting. The number of virtual machines running
inside the virtual appliance depends on the capacity desired and on whether high availability is needed.
This provides various options for deployment size.
11-50
Chapter 11
Figure 11-24
Data Center
Virtual Appliance
Internet Reverse
Proxy VM
Web/Video/PC
Audio Traffic from
Mobile Users
Guest and
Mobile
Users
Telephony
Audio
Web/Video
Traffic from
Internal Users
Administration
VM
Administration
VM
Web VM
Web VM
Media VM
Media VM
SAML 2.0
Single Sign On
WebEx
API
SIP
Internal
Users
External Firewall
Internal Firewall
Corporate
Directory
Cisco Jabber
Cisco Unified CM
348749
DMZ
Cisco WebEx Meetings Server offers the option of deploying the Internet Reverse Proxy (or edge
servers) in the DMZ to facilitate external access. This option provides two advantages. First, all external
participants can securely access the WebEx conferences from the internet without going through a VPN.
Second, mobile users can join the meetings from a mobile device anywhere as long as there is internet
connectivity. Note that the Internet Reverse Proxy is mandatory if mobile client access is enabled.
Internet Reverse Proxy is used to terminate all inbound traffic from the internet inside the DMZ. The
content is then forwarded to the internal virtual machines through an encrypted Secure Socket Layer
(SSL) or Transport Layer Security (TLS) tunnel. This encrypted tunnel is established by the internal
virtual machines connecting outbound to the Internet Reverse Proxy. Therefore, there is no need to open
TCP ports inbound from the DMZ to the internal network on the internal firewall. However, some
outbound ports from the internal network need to be opened on the internal firewall to allow
communication with the Internet Reverse Proxy in the DMZ.
All end-user sessions are 100% encrypted using industry standard Secure Socket Layer (SSL) and
Transport Layer Security (TLS). All traffic between the virtual machines is sent over the secure channel.
Federal Information Processing Standard (FIPS) encryption can also be turned on by a single policy
setting, providing US Department of Defense (DoD) level security. Alternatively, the Internet Reverse
Proxy can be deployed behind the internal firewall as shown in Figure 11-25.
11-51
Chapter 11
Figure 11-25
Data Center
Virtual Appliance
Internet Reverse
Proxy VM
Administration
VM
Web/Video/PC
Audio Traffic from
Mobile Users
Web VM
Guest and
Mobile
Users
Telephony
Audio
Web/Video
Traffic from
Internal
Users
Media VM
SAML 2.0
Single Sign On
DMZ
External Firewall
SIP
WebEx
API
Internal Firewall
Corporate
Directory
Cisco Jabber
Cisco Unified CM
348750
Internal
Users
For security concerns, an organization would typically take several months to get approval in deploying
a component inside the DMZ. Using this methodology, it could eliminate any DMZ components and
bypass the approval process to get the WebEx Meetings Server deployment done quickly. All internet
traffic (HTTP on port 80 and SSL on port 443) to the external firewall should be forwarded to the internal
firewall. This will minimize the number of ports that need to be opened in the external and internal
firewalls. However, placing the Internet Reverse Proxy inside the internal network implies that inbound
internet traffic will terminate in the internal network. Although direct internet access to the internal
network could be controlled by the firewalls, not all organizations allow terminating internet traffic
directly on their internal network. Ensure that this deployment does not violate your organization's IT
policy before choosing this option.
In a large enterprise deployment, an organization would require the Single Sign On (SSO) capability to
allow end users to sign in using their corporate credentials. Cisco WebEx Meetings Server can connect
to the corporate LDAP directory using the industry standard SAML 2.0 for SSO.
Note
Note
Starting with Cisco WebEx Meetings Server 1.1, Cisco Jabber integrated with the Cisco Unified CM IM
and Presence Service can be used to join or start meetings hosted on WebEx Meetings Server. For Cisco
Jabber support details, refer to the Cisco WebEx Meetings Server System Requirements, available at
http://www.cisco.com/en/US/products/ps12732/prod_installation_guides_list.html.
11-52
Chapter 11
Cisco Unified CM integrates with WebEx Meetings Server by means of SIP trunks to provide inbound
and callback call control. Customer can choose to turn on security and run Transport Layer Security
(TLS) and Secured Real-time Transport Protocol (SRTP) over the SIP trunk connection. A SIP trunk is
configured in Unified CM with a destination address of the Load Balancer in WebEx Meetings Server,
and then a route pattern (match the call-in access number configured in WebEx Meetings Server) must
be used to route calls via the SIP trunk. A second SIP trunk is configured in Unified CM with a
destination address of the Application Server in WebEx Meetings Server, and then a SIP route pattern
must be used to route calls via the SIP trunk. When an attendee dials the access number to join the
meeting, the first SIP trunk is used to send the call. After the call is connected and the caller enters the
meeting ID, the Load Balancer issues a SIP REFER to Unified CM to send the caller to the Application
Server that hosts the meeting via the second SIP trunk.
The system administrator can configure a SIP trunk in WebEx Meetings Server that points to a
Unified CM to perform callback. Attendees can provide a callback number and have the system out-dial
the number to the attendees to join the bridge. In the case of attendees requesting callback, the WebEx
Meetings Server sends the SIP request to Unified CM along with the callback number via the configured
SIP trunk. It is imperative for Unified CM to be able to resolve all dial strings received from a callback
request to join the meetings. Callbacks may also be disabled system-wide by means of site
administration settings. Unified CM is in control of all toll restrictions to various countries or other
numbers that most enterprises will block, because WebEx Meetings Server does not have any toll
restriction blocking itself.
WebEx Meetings Server supports the bidirectional SIP OPTIONS ping mechanism. The ping response
from the remote end indicates that the remote end is active and whether it is ready to accept calls. Based
on the response, WebEx Meetings Server or Unified CM can determine whether to send calls on the
current SIP trunk or look for an alternate SIP trunk (if configured) to send calls. Note that SIP OPTIONS
ping is supported in Cisco Unified CM 8.5 and later releases. Due to this reason, Cisco recommends
using a compatible Cisco Unified CM version that supports SIP OPTIONS ping for Cisco WebEx
Meetings Server deployment. For the list of compatible Unified CM versions, refer to the compatibility
matrix in the Cisco WebEx Meetings Server System Requirements, available at
http://www.cisco.com/en/US/products/ps12732/prod_installation_guides_list.html
Note
Cisco WebEx Meetings Server supports SIP trunk connection with Cisco Unified CM only.
11-53
Chapter 11
Allow an organization to adopt the new technology gradually, at its own pace
Protect the customer's investment in existing technology while allowing them to migrate to Cisco
technology gradually
For further details on PBX interoperability with Unified CM, refer to the documentation available at
http://www.cisco.com/en/US/solutions/ns340/ns414/ns728/networking_solutions_products_generi
ccontent0900aecd805b561d.html
IPv6 Support
Cisco WebEx Meetings Server supports IPv4 only or dual stack (IPv4 and IPv6) addressing for telephony
audio, while telephony signaling remains at IPv4. Audio streams can be IPv4, IPv6, or a mix of IPv4 and
IPv6 in the same meeting. Cisco WebEx Meetings Server supports Alternate Network Address Types
(ANAT) to enable both IPv4 and IPv6 media addressing in the Session Description Protocol (SDP)
during the SIP Offer and Answer exchange on the SIP trunk with Unified CM to establish a media
connection using the preferred addressing scheme.
Both IPv4 and IPv6 devices can be used for teleconferencing. With IPv6 devices, Cisco WebEx Meetings
Server leverages Unified CM's capacity to translate the IPv6 signaling to IPv4 and transport it over a SIP
trunk to the Cisco WebEx Meetings Server. With the telephony media addressing, Cisco WebEx
Meetings Server can convert between IPv4 and IPv6. Therefore, Cisco WebEx Meetings Server can
support IPv6 without any expensive MTP resources.
With ANAT, Cisco WebEx Meetings Server can support IPv6 telephony audio without the support of
IPv6 telephony signaling. However, ANAT must be supported on both ends of the Unified CM SIP trunk.
Be sure to enable ANAT on the Unified CM SIP trunk, otherwise there will be a failure to establish the
call when attendees request callback or attempt to dial in.
If the WebEx Meetings Server has IPv6 enabled, ANAT headers will be included in the media offer.
WebEx Meetings Server will always answer with ANAT headers if the media offer includes ANAT
headers. The following paragraphs describe the media address version selection process between the
IPv6-enabled WebEx Meetings Server and the dual-stack Unified CM using the ANAT header.
When WebEx Meetings Server sends a call to Unified CM, the SDP offer contains both IPv4 and IPv6
media addresses. If the called device is IPv6, Unified CM chooses IPv6 for the media connection and
answers with the IPv6 media address in the SDP; if the called device is dual-stack, Unified CM uses the
IP Addressing Mode Preference for Media parameter to determine the address version in the answer
SDP. If the parameter is set to IPv6, then IPv6 will be used for the media connection.
11-54
Chapter 11
When Unified CM sends a call to the WebEx Meetings Server through the SIP trunk, WebEx Meetings
Server receives the SDP offer with an ANAT header. If the SDP offer contains both IPv6 and IPv4 media
addresses, WebEx Meetings Server answers with the higher precedence address version specified in the
ANAT header, which would be IPv6 in this case. If the SDP contains only an IPv6 address, WebEx
Meeting Server answers with an IPv6 media address.
For information on deploying IPv6 in a Cisco Unified Communications system, refer to the latest version
of Deploying IPv6 in Unified Communications Networks with Cisco Unified Communication Manager,
available at
http://www.cisco.com/go/ucsrnd
High Availability
Cisco WebEx Meetings Server uses the N+1 redundancy scheme to ensure system availability in the
event of component failures. High availability is achieved by adding a local, redundant system to the
primary system within the same data center. At the system level, virtual machines and components inside
run in active/active mode. If one component goes down, the system restarts the component. Status
information is exchanged between system components. Using this status information, the system is able
to distribute the requests evenly among the active components. Depending on the deployment size, the
number of virtual machines in the backup or redundant system might or might not be the same as in the
primary system.
In the high availability system, when the virtual machine hosting the meeting goes down, affected
meeting clients will automatically reconnect to the available service within a short period of time.
However, depending on the nature of the failure and which component has failure, not all clients and
meetings would be affected. For descriptions of high availability system behavior after a component
failure, refer to the latest version of the Cisco WebEx Meetings Server Administration Guide, available at
http://www.cisco.com/c/en/us/support/conferencing/webex-meetings-server/products-installationguides-list.html
Virtual IP Address
Inside the high availability system, there is a second network interface in the active administration and
Internet Reverse Proxy virtual machine that is configured with the virtual IP address. The administration
and WebEx site URLs use this virtual IP address to access the administration and WebEx sites. In the
event of failover, the virtual IP address is moved over to the new active virtual machine. Thus, it provides
access redundancy to the administration and WebEx site.
11-55
Chapter 11
peers over an encrypted SSL link. Administrators use a single URL to manage the systems, and
participants use a single URL or one set of dial-in numbers to join the meeting. When participant join a
meeting via the client, the system automatically chooses the data center closest to the participant to host
the meeting, and the meeting is cascaded across data centers.
In the event of failure, if one component goes down in the data center, the system restarts that component.
If the whole data center goes down, the surviving data center takes over without any manual intervention,
and the system still runs with full capacity. When this happens, affected meeting clients automatically
reconnect to the service in the surviving data center within a short period of time. However, depending
on the nature of the failure and state of the client, the recovery mechanism might be different and would
follow the same behavior as the high availability system. For detail descriptions, refer to the latest
version of the Cisco WebEx Meetings Server Administration Guide, available at
http://www.cisco.com/c/en/us/support/conferencing/webex-meetings-server/products-installationguides-list.html
Consider the following information when using the multiple data center design:
A multi-data center license is required for the WebEx Meetings Server system in each data center.
Install the licenses onto the primary data center system before joining the data centers.
A deployment size of 50 users per system is not supported, but larger system sizes are supported.
Running a high availability system within the data center is not supported.
Joining the systems together will not increase the total system capacity.
Either both data centers or neither data center can have Internet Reverse Proxy deployed.
Capacity Planning
The capacity of WebEx Meetings Server depends on the platform of choice and the number of
conferencing nodes running in the deployment. For capacity planning details, see the section on
Collaborative Conferencing, page 25-42.
Storage Planning
If recording meetings is a requirement, sufficient disk space should be allocated on the Network
Attached Storage (NAS) device to store the recordings. For disk space allocation detail, refer to the
Meeting Recordings section in the Cisco WebEx Meetings Server Planning Guide, available at
http://www.cisco.com/en/US/products/ps12732/products_installation_and_configuration_guides_l
ist.html
11-56
Chapter 11
Design Consideration
The following additional design considerations apply to WebEx Meetings Server deployments:
For scenarios where any WebEx Meetings Server components are separated by network firewalls, it
is imperative to ensure the correct pinholes are opened for all required traffic.
Collaborative meeting systems typically result in increased top-of-the-hour call processing load.
Capacity planning tools with specific parameters for WebEx Meetings Server are available to Cisco
partners and employees to help calculate the capacity of the Cisco Unified Communications System
for large configurations. Contact your Cisco partner or Cisco Systems Engineer (SE) for assistance
with sizing of your system. For Cisco partners and employees, the Cisco Unified Communications
Sizing Tool is available at http://tools.cisco.com/cucst.
Using Transport Layer Security (TLS) and Secured Real-time Transport Protocol (SRTP) have no
effect to the WebEx Meetings Server capacity. However, using TLS and SRTP does have an impact
on Cisco Unified CM capacity.
WebEx Meetings Server has no built-in line echo cancellation. Use an external device such as a
Cisco Integrated Service Router (ISR) to provide echo cancellation functionality.
11-57
Chapter 11
For more details on the various Cisco Collaboration client offerings and how they fit into Cisco
conferencing solutions, see the chapter on Collaboration Endpoints, page 8-1.
Call admission control with WebEx Meetings Server is performed by Unified CM. With
locations-based call admission control, Unified CM can control bandwidth to the WebEx Meetings
Server system by placing the SIP trunk specific to WebEx Meetings Server in a location with a set
amount of audio bandwidth allowed. Alternatively, Unified CM supports the use of Resource
Reservation Protocol (RSVP), which can also provide call admission control. For further
information regarding call admission control strategies, see the chapter on Bandwidth Management,
page 13-1.
Cisco recommends marking both the audio streams and video streams from WebEx Meetings Server
as AF41 (DSCP 0x22) to preserve lip-sync. These values are configurable in WebEx Meetings
Server Administration.
Web conferencing traffic is encrypted in SSL and is always marked best-effort (DSCP 0x00).
Reference Document
For network requirements, network topology, deployment size options, and other deployment
requirements and options for WebEx Meetings Server, refer to the Cisco WebEx Meetings Server
Planning Guide, available at
http://www.cisco.com/en/US/products/ps12732/products_installation_and_configuration_guides_l
ist.html
11-58
Chapter 11
Architecture
As shown in Figure 11-26, the high-level architecture of Cisco CMR Hybrid consists of the enterprise
collaboration network and the WebEx Cloud infrastructure that are connected through an IP connection.
The enterprise collaboration network consists of Cisco Unified Communications Manager
(Unified CM), Cisco Expressway-C and Expressway-E, TelePresence Bridge pools that are managed by
TelePresence Conductor, and Cisco TelePresence Management Suite (TMS). Cisco Unified CM is the
call processing platform that provides call routing and call control for the TelePresence endpoints within
the enterprise. Cisco Expressway-C and Expressway-E route calls between the enterprise network and
WebEx Cloud. Cisco Unified CM connects with Cisco Expressway-C and Cisco TelePresence Conductor
over separate Best Effort Early Offer SIP trunks.
For details on integrating Cisco Unified CM with Cisco Expressway, refer to the latest version of the
Cisco Expressway and CUCM via SIP Trunk Deployment Guide, available at
http://www.cisco.com/c/en/us/support/unified-communications/expressway-series/products-install
ation-and-configuration-guides-list.html
Note
For existing Cisco VCS customers, deployment using Cisco VCS Control and Expressway in place of
Cisco Expressway-C and Expressway-E is supported.
Note
Deployment using a Best Effort Early Offer SIP trunk between Unified CM and the TelePresence Bridge
without TelePresence Conductor is supported, but using TelePresence Conductor is recommended.
Cisco TelePresence Conductor selects a TelePresence Bridge from the pool to host the TelePresence
conference. The TelePresence Bridge mixes the audio from the TelePresence endpoint participants and
sends the mixed audio, the active speaker video, and the content sharing video to the WebEx Cloud using
SIP. Similarly, the TelePresence Bridge receives the media (mixed audio, active speaker, and content
sharing video) from the WebEx Cloud, cascades the audio into the TelePresence conference, and sends
the content sharing video to the TelePresence endpoints. If the TelePresence Bridge detects that the
active speaker is from the WebEx side, it switches the TelePresence endpoints to the active speaker
video. If the active speaker is from the TelePresence side, the TelePresence Bridge sends the previous
active speaker video to the TelePresence endpoint of the current active speaker.
11-59
Chapter 11
Figure 11-26
WebEx Client
Cisco
Expressway-C
Cisco
Expressway-E
TelePresence
Conductor
SIP Signaling
Cisco
Unified CM
348789
Cisco TelePresence
Management Suite
In the DMZ, Cisco Expressway-E handles the traversal calls between the enterprise and WebEx Cloud,
and it allows the signal and media to traverse through the internal and external firewalls. Cisco
Expressway-E connects with the WebEx Cloud through the configured DNS Zone and routes calls to
WebEx via DNS lookup. Cisco Expressway-E communicates with WebEx Cloud via an encrypted
connection using TLS and secured RTP for SIP signal and media. Customers have an option to turn on
encryption for the SIP signal and media traffic within the enterprise. TelePresence endpoints outside of
the enterprise can register with Unified CM through Expressway-C and Expressway-E, and thus
participants on these endpoints can join the CMR Hybrid meetings.
When the WebEx Cloud receives the traversal calls and media sent from the enterprise network, the
WebEx audio bridge cascades the audio into the WebEx conference, and WebEx switches to the active
speaker video and displays the content sharing on the WebEx meeting clients. Similarly, WebEx Cloud
sends the conference mixed audio, the active speaker, and content sharing video from the WebEx side to
the enterprise via Cisco Expressway-E and Expressway-C, which routes them to the TelePresence
Bridge.
Cisco CMR Hybrid supports H.264 video for active speaker and content sharing. It utilizes Binary Floor
Control Protocol (BFCP) for content sharing and G.711 codec for audio. While Cisco WebEx uses H.264
video and G.711 audio codec, TelePresence can still use other video formats or codecs that are supported
by the endpoints. The TelePresence Bridge will handle the audio and video interoperability between the
TelePresence endpoints and WebEx meeting clients. In addition, there is a flow control on the link
between the TelePresence Bridge and WebEx Cloud that regulates the bandwidth available for handling
the media. For media from WebEx, the TelePresence Bridge always allocates 4 Mbps to ensure that
WebEx sends the best quality of video possible to the TelePresence Bridge. For media from the
TelePresence Bridge, the WebEx meeting client has a video floor of 180p for active speaker video at the
minimum bit rate of 1.2 Mbps. If the minimum bit rate cannot be maintained due to network conditions
(severe packets loss, for example), the WebEx client will stop receiving the active speaker video but still
11-60
Chapter 11
receives content sharing as well as conference audio and sends its video to other participants. Starting
with WBS 29.11, the WebEx client will periodically perform bandwidth retest and automatically
reestablish active speaker video when network conditions stabilize. Depending on the capability of the
device running the WebEx meeting client and on bandwidth available, the WebEx client supports active
speaker video up to HD 720p at 30 frames per second (fps) and content video up to 1080p. During the
meeting, WebEx allocates the bandwidth based upon the least capable device among all WebEx clients
in the conference (excluding devices running below the video floor), with a maximum bandwidth of
4 Mbps. However, if the least capable device leaves the conference, the bandwidth will be re-allocated
based on the next least capable device that runs the WebEx meeting client. The allocated bandwidth
determines the resolution and frame rates used to display TelePresence video on WebEx clients.
Depending on the TelePresence endpoints deployed, video resolution required, screen layout desired,
and deployment options chosen, customers can deploy the TelePresence Bridge using the Cisco
TelePresence Server (appliance or virtualized platforms) or Cisco TelePresence MCU, but the pool must
consists of bridges of the same type only (either TelePresence Server or TelePresence MCU). For
TelePresence Conductor deployment details, see to the section on Cisco Collaboration Meeting Room
Premises, page 11-6.
WebEx and TelePresence participants can join the CMR Hybrid meeting from within the enterprise or
anywhere from the internet. For WebEx participants, they join the meeting using the WebEx meeting
clients with either PSTN or VoIP audio. For TelePresence participants, they join the meeting via the One
Button To Push (OBTP) or Auto Connect feature with the supported endpoints or by calling directly into
the TelePresence Bridge. Once the participants successfully join the meeting, they can see the live video
of each other from the endpoints and meeting clients. For presentation sharing with a WebEx user, either
the user can make himself the presenter or the host can assign the presenter privilege to the user before
he can start sharing the presentation. There is the WebEx site configuration to control this behavior. For
presentation sharing with a TelePresence user, the user can connect the video display cable to his
computer or press a button on the endpoint to start sharing his presentation without involving the host.
Note
Staring with Cisco TMS 14.6 and TMSPE 1.4, Cisco Collaboration Meeting Room Premises can be
integrated with Cisco WebEx, allowing participants to join a meeting in the user's personal room from
the WebEx meeting client.
Scheduling
Cisco TelePresence Management Suite (TMS) is the key component for scheduling Cisco CMR Hybrid
meetings. It provides a control link to the Cisco WebEx meeting scheduler. This link enables Cisco TMS
to create new meetings on Cisco WebEx calendar and to obtain Cisco WebEx meeting information that
is distributed to meeting participants. The following options are available to schedule CMR Hybrid
meetings:
11-61
Chapter 11
Smart Scheduler
Smart Scheduler is a web-based tool that is hosted on Cisco TelePresence Management Suite
Provisioning Extension (TMSPE), and it allow users to schedule CMR Hybrid meetings using a
browser. This could provide an option for users who would like to schedule meetings on mobile
devices.
Note
As long as the Cisco TMSPE option key has been installed, there is no extra license required
for using Smart Scheduler.
For Cisco TMS configuration details with these options, refer to the Cisco Collaboration Meeting Rooms
(CMR) Hybrid Configuration Guide, available at
http://www.cisco.com/en/US/products/ps11338/products_installation_and_configuration_guides_l
ist.html
Scheduling a CMR Hybrid meeting is a two-steps process. First, a request is sent to the WebEx Cloud
to schedule the meeting on the WebEx calendar, and the WebEx Cloud responds with the meeting details
that are passed to Cisco TMS. Second, Cisco TMS schedules the TelePresence meeting in its calendar.
When it is the meeting start time, Cisco TMS pushes the meeting details to the TelePresence Bridge for
joining the meeting on WebEx. The meeting details returned from WebEx include the date and time for
the meeting, dial-in information, subject, meeting number, URL for joining the meeting, and so forth.
Once the meeting has been scheduled, details for the WebEx and TelePresence portions of the meeting
are sent to the host, and the host can forward the details to all participants. However, if the productivity
tool is used, the meeting details are automatically included in the invitation that the host creates and
sends to the meeting participants.
11-62
Chapter 11
Single Sign On
Cisco CMR Hybrid supports scheduling the WebEx portion of the meeting in Cisco TMS using Single
Sign On (SSO). This feature requires the WebEx site to have Cisco TMS provisioned as the delegated
partner and to have the Partner Delegated Authentication configured. With SSO enabled in Cisco TMS,
only the user's WebEx username is stored in the Cisco TMS user profile without the need of the WebEx
password. When the user schedules a CMR Hybrid meeting, WebEx trusts Cisco TMS and requires only
the WebEx username stored in Cisco TMS to schedule the meeting in the WebEx calendar. For
Cisco TMS configuration details with SSO, refer to the Cisco Collaboration Meeting Rooms (CMR)
Hybrid Configuration Guide, available at
http://www.cisco.com/en/US/products/ps11338/products_installation_and_configuration_guides_l
ist.html
For more information regarding SSO with Cisco WebEx, refer to the white papers and technical notes
available at
http://developer.cisco.com/web/webex-developer/sso-reference
Security
All communications between the enterprise network and the WebEx Cloud are encrypted (using TLS and
secured RTP). Customers also have an option to turn on encryption for the SIP signal and media within
the enterprise. A certificate has to be uploaded to the Cisco Expressway-E to ensure that proper
handshaking takes place for the TLS connection to be functional. That certificate must be signed by a
trusted Root Certificate Authority. For the list of the trusted Root Certificate Authorities, refer to the
Cisco Collaboration Meeting Rooms (CMR) Hybrid Configuration Guide, available at
http://www.cisco.com/en/US/products/ps11338/products_installation_and_configuration_guides_l
ist.html
A password is required when the TelePresence Bridge calls into WebEx to join the meeting. The
password is allocated for each CMR Hybrid meeting scheduled on the WebEx calendar and is embedded
in the SIP URI that is returned as part of the meeting details from the WebEx Cloud. This password is
encoded into 22 bytes and qualifies for the security standards. At the start of the meeting, the
TelePresence Bridge calls into WebEx using this SIP URI, and WebEx validates the password to
authorize the call to join the meeting.
Deployment Options
When it is the start time for the CMR Hybrid meeting, Cisco TMS initiates the conference on the
TelePresence Bridge through TelePresence Conductor for the TelePresence participants. The
TelePresence Bridge makes a SIP call through TelePresence Conductor out to the WebEx Cloud using
the SIP URI that was returned as part of the scheduling process and to join the conference on the WebEx
side. As a result, the TelePresence Bridge establishes separate audio, active speaker video, and content
sharing video streams with the cloud for the meeting. The active speaker video, content sharing video,
and conference control always travels over the IP network, but the audio can travel over either the IP
network or the PSTN, depending on the deployment options chosen. The various audio options available
for CMR Hybrid are:
WebEx Audio Using SIP, page 11-64, including Cloud Connected Audio
11-63
Chapter 11
11-64
Chapter 11
Figure 11-27
WebEx Client
Cisco
Expressway-C
Cisco
Expressway-E
TelePresence
Conductor
SIP Signaling
Cisco TelePresence
Management Suite
Cisco
Unified CM
348790
PSTN
Conference Audio
Voice Gateway
The dial-out number returned from WebEx is in full E.164 number format (for example, +14085551212).
The dial plan design in Cisco Unified CM should take into account the handling of E.164 numbers. For
dial plan design with Cisco Unified CM, see the chapter on Dial Plan, page 14-1.
11-65
Chapter 11
TelePresence
Bridge Pool
Cisco
Expressway-E
Cisco
Expressway-C
TelePresence
Conductor
SIP Signaling
Cisco TelePresence
Management Suite
Teleconferencing
Service Provider
Participant Video
Cisco
Unified CM
Conference Audio
Voice Gateway
PSTN
WebEx Client
During the scheduling process, in addition to the dial-out number and meeting number, extra digits for
navigating through the IVR prompts on the TSP audio bridge are passed from WebEx to Cisco TMS. At
the scheduled meeting start time, the TelePresence Bridge initiates a SIP call to the WebEx Cloud to
establish the video connections. At the same time, the TelePresence Bridge dials out to the TSP audio
bridge through the PSTN. Then the TelePresence Bridge plays out the meeting number as a DTMF dial
sequence, along with additional DTMF digits to navigate through the IVR prompts on the audio bridge
to start the meeting. On the WebEx side, WebEx participants start the WebEx session using the meeting
client and dial into the TSP audio bridge or have callback from the audio bridge. Thus, the audio streams
from TelePresence and WebEx participants are cascaded. From this point onward, information about the
loudest speaker, participant list, and so forth in the WebEx side, is passed from the TSP to WebEx
through the TSP link and then into the enterprise collaboration network.
11-66
Chapter 11
The dial-out number returned from WebEx is in full E.164 number format (for example, +14085551212).
The dial plan design in Cisco Unified CM should take into account the handling of E.164 numbers. For
dial plan design with Cisco Unified CM, see the chapter on Dial Plan, page 14-1.
High Availability
There are two areas that must be considered when designing high availability for CMR Hybrid: the
enterprise collaboration network and the WebEx Cloud. The WebEx Cloud is managed by Cisco and
already has the redundancy built into the infrastructure. For details, see the section on Cisco WebEx
Software as a Service, page 11-42.
In the enterprise collaboration network, utilize the clustering option from Cisco Unified CM and Cisco
Expressway to provide redundancy for call control and call routing on the TelePresence endpoints. In
case the primary server fails, the backup server can take over the call control and call routing functions.
In addition, resiliency of the TelePresence conferencing infrastructure must be considered to handle
failure of conference bridges.
For Cisco Unified CM clustering, see the chapter on Call Processing, page 9-1.
For Cisco Expressway clustering, refer to the latest version of the Cisco Expressway Cluster Creation
and Maintenance Deployment Guide, available at
http://www.cisco.com/c/en/us/support/unified-communications/expressway-series/products-install
ation-and-configuration-guides-list.html
For resiliency of the TelePresence conferencing infrastructure, see the section on Cisco Collaboration
Meeting Room Premises, page 11-6.
Capacity Planning
The WebEx Cloud has the built-in capability to evenly distribute the traffic and dynamically add more
capacity if thresholds are exceeded. Capacity planning for Cisco CMR Hybrid involves sizing of the
components running within the enterprise. The components include:
TelePresence Conferencing
The Cisco TelePresence Conductor, Cisco TelePresence Server, or Cisco TelePresence MCU must
provide enough resources to handle the conference traffic. For details, see the section on Capacity
Planning, page 11-33.
Cisco Expressway
Cisco Expressway must provide enough resources to handle the traversal call traffic for the
deployment. For capacity details, see the chapter on Collaboration Solution Sizing Guidance,
page 25-1.
11-67
Chapter 11
For optimal SIP audio and video quality between the TelePresence Bridge and the WebEx Cloud, Cisco
recommends setting up the video bandwidth of at least 1.3 Mbps in the region associated with each
endpoint registering with Cisco Unified CM.
Design Considerations
The following design considerations apply to Cisco CMR Hybrid deployments:
Upgrade from previous versions of CMR Hybrid that use the Cisco TelePresence MultiPoint Switch
infrastructure is not supported, and customers using those previous versions should plan for
migration.
Every user who wants to schedule a CMR Hybrid meeting must have a host account with Cisco
TelePresence Session type assigned in the WebEx site.
Any endpoints that can register with Cisco Unified CM and that are supported by the TelePresence
Bridge can be used to join the Cisco CMR Hybrid meeting.
Only devices managed by the Cisco TelePresence Management Suite (TMS) can use One Button to
Push (OBTP) or the Auto Connect feature to join the CMR Hybrid meeting.
Ensure that the Cisco Unified CM Neighbor Zone in Cisco Expressway-C is configured with Binary
Floor Control Protocol (BFCP) enabled.
Provision Hybrid Audio in the WebEx site to allow the use of SIP audio for the TelePresence Bridge
and PSTN audio for WebEx participants.
Cisco CMR Hybrid does not support Cisco WebEx Meetings Server.
The TelePresence Bridge becomes the default host if no host is present when it joins the
CMR Hybrid meeting, and the host privilege is reassigned to the host when he joins using the
WebEx meeting client.
The TelePresence Bridge will call into the WebEx Cloud at meeting start time even if no
TelePresence or WebEx participant has joined yet.
The organizer's WebEx account and Outlook time zone should match; otherwise, the meeting
scheduled in WebEx and in the Cisco TMS calendar will have different start times.
Enable UDP for media streaming in the firewalls for the optimal video experience.
11-68
Chapter 11
Architecture
Figure 11-29 illustrates the Cisco CMR Cloud architecture using SIP video. This architecture consists
of the enterprise collaboration network and the WebEx Collaboration Cloud where all the conferencing
resources are hosted, and they are connected via the Internet. The enterprise collaboration network
encompasses Cisco Unified Communications Manager (Unified CM) and Cisco Expressway, and
Unified CM connects with Cisco Expressway-C over a SIP trunk. Cisco Unified CM provides the call
routing and call control functions for the registered video devices. Cisco Expressway provides a secure
firewall traversal mechanism for calls between the enterprise and WebEx Cloud, and it routes the video
calls to WebEx Cloud via the DNS zone configured inside Cisco Expressway-E. In addition, Cisco
Expressway provides mobile and remote access capability to the supported Cisco video endpoints so
they can register with Unified CM outside of the enterprise. In order for a participant to join the meeting
and share content, the SIP device must support URI dialing and Binary Floor Control Protocol (BFCP).
Without BFCP, content cannot be shared and will be seen embedded in the main video.
Note
For existing Cisco VCS customers, using VCS Control as a SIP Registrar for SIP endpoints and VCS
Expressway for firewall traversal is supported with the deployment.
11-69
Chapter 11
Figure 11-29
Cisco
Expressway-C
Cisco
Expressway-E
SIP Signal
DNS
Third-Party
Cisco
Unified CM
Internet
348797
Third-Party
Cisco CMR Cloud architecture also support H.323 video devices (see Figure 11-30). In this architecture,
Cisco VCS Control is the gatekeeper and provides call control for the registered H.323 endpoints. Cisco
VCS Expressway provides a secure firewall traversal mechanism for calls between the enterprise and
WebEx Cloud, and it routes the video calls to WebEx Cloud via the DNS zone configured inside Cisco
VCS Expressway. In order for a participant to join the meeting and share content, the H.323 device must
support Annex O for URI dialing and H.239 for content sharing. Without H.239, content cannot be
shared and will be seen embedded in the video. In addition, H.323 devices must support either the H.245
User Input or RFC 2833 method of DTMF signaling in order to use interactive voice response (IVR) to
start a meeting as a host or to join a meeting before the host.
Figure 11-30
Cisco
H.323 Endpoint
Cisco
VCS Control
Cisco VCS
Expressway
DNS
348798
Internet
Third-Party
H.323 Endpoint
11-70
Chapter 11
Alternatively, Cisco CMR Cloud can be deployed using H.323 video without a call control system (see
Figure 11-31). In this architecture, the H.323 device does not register to any gatekeeper; and when the
user dials the URI, the call is routed using DNS through the firewall to the WebEx Cloud. Make sure the
necessary ports on the firewall are opened so that signaling and media can pass through.
Cisco Collaboration Meeting Room Cloud Architecture Using H.323 Video Without
Call Control System
H.323 Endpoint
DNS
Internet
348799
Figure 11-31
Irrespective of SIP or H.323 devices used in the deployment, WebEx Cloud can perform the interworking
between protocols. There are requirements for video devices to be used in a CMR Cloud deployment.
For details, refer to Cisco Collaboration Meeting Rooms (CMR Cloud) Enterprise Deployment Guide,
available at
http://www.cisco.com/c/en/us/support/conferencing/webex-meeting-center/products-installation-a
nd-configuration-guides-list.html
For each participant on a video device, the audio, video, and content sharing are sent over the IP
connection to WebEx Cloud, where the media are mixed with other participants, and the mixed audio,
active speaker video, and content sharing are sent back to the device for display.
Cisco CMR Cloud uses H.264 video for active speaker and content sharing. Depending on the capability
of the device and the bandwidth available, CMR Cloud supports active speaker video up to 720p at
30 frames per second (fps) and content video up to 720p on video devices as well as WebEx clients.
WebEx meeting client has a video floor of 180p for active speaker video at the minimum bit rate of
1.2 Mbps. If the minimum bit rate cannot be maintained due to network condition (severe packets loss,
for example), WebEx client will stop receiving the active speaker video but still receives content sharing
as well as conference audio and sends its video to other participants. Starting with release WBS 29.11,
WebEx client will periodically perform bandwidth retest and automatically reestablish active speaker
video when network conditions stabilize. During the meeting, WebEx allocates the bandwidth based
upon the least capable device among all WebEx clients in the conference (excluding devices running
below the video floor), with a maximum bandwidth of 4 Mbps. However, if the least capable device
leaves the conference, the bandwidth will be reallocated based upon the next least capable device that
runs the WebEx meeting client. The allocated bandwidth determines the resolution used to display the
video on the WebEx clients.
11-71
Chapter 11
Each CMR Cloud meeting has an associated video address URI and URL. Participants dial the URI on
the video device or click on the URL to bring up the WebEx meeting client to join the meeting. A
CMR Cloud meeting can be one of the following types:
Scheduled meeting
Users can use WebEx Productivity Tools to schedule Cisco CMR Cloud meetings. Productivity
Tools is a suite of tools, including an Outlook plug-in, that allows users to schedule meetings quickly
and easily within the email client. This tool suite provides seamless integration with the user's
calendar, and users can schedule meetings and send the invitations to all participants directly inside
the email client with a single transaction. Alternatively, user can schedule CMR Cloud meetings
from the WebEx portal but the host has to first schedule the meeting from WebEx, and then create
an invitation with meeting detail attached and send it to all the participants.
Permanent meeting
Meetings can be hosted in the user's personal room. Personal rooms can be enabled at the site level
or per-user level in the WebEx site. When enabled, a fixed URI and URL are assigned to the user,
and participants can use them to join the user's personal room. This personal room belongs to the
designated user and is always on. Thus, the user can use his room for his meetings and can send an
invitation to all participants with his room's URI and URL attached.
Instant meeting
A user can create an instant meeting from the WebEx portal or by using the WebEx Productivity
Tools, and the meeting will start immediately. Using the Meet Now configuration option, the instant
meeting can be instantiated from the Meeting Center or the user's personal room.
Security
Cisco CMR Cloud supports encrypted signaling and media, or a combination of encrypted and
non-secure signaling and media, between the enterprise network and WebEx Cloud. For end-to-end
encryption, customers can turn on encrypted signaling and media in the enterprise and use encrypted
signaling and media between the enterprise network and WebEx Cloud. A certificate has to be uploaded
to Cisco Expressway-E to ensure that proper handshaking takes place for encrypted signaling to be
functional. That certificate can be either self-signed or signed by a trusted Root Certificate Authority
(CA). For the list of the trusted Root Certificate Authorities, refer to Cisco Collaboration Meeting
Rooms (CMR Cloud) Enterprise Deployment Guide, available at
http://www.cisco.com/c/en/us/support/conferencing/webex-meeting-center/products-installation-a
nd-configuration-guides-list.html
For SIP based calls, Cisco CMR Cloud supports four levels of security (in order of preference):
Encrypted TLS signaling with CA-signed certificates and SRTP media encryption
Encrypted TLS signaling with self-signed certificates and SRTP media encryption
Make sure to open the network ports on the firewall so that inbound and outbound traffic for signaling
and media can pass through. For port range details, refer to Cisco Collaboration Meeting Rooms (CMR
Cloud) Enterprise Deployment Guide.
All CMR Cloud meetings require the presence of the host to start the meeting. If the guests join before
the host, they will be in the waiting room and cannot talk to each other until the host joins. In addition,
a host PIN is required when the host joins the meeting from a video device.
11-72
Chapter 11
Inside the user's personal meeting room, a Lock Room button is available that can be used to lock the
room and prevent other participants from entering the user's personal room. When the room is locked
and a participant tries to enter the room, that participant will be blocked until the host admits him or
unlocks the room. This button is useful in case a user's personal room is used for back-to-back meetings
and the host has not finished with the first meeting. The host can lock the room to prevent participants
of the second meeting from entering until he finishes with the first meeting and unlocks the room.
High Availability
In the enterprise collaboration network, utilize the clustering option with Cisco Unified CM and Cisco
Expressway to provide redundancy for call control with video devices and firewall traversal calls. If the
primary server fails, the backup server can take over the call control and call handling functions.
For Cisco Unified CM clustering, see the chapter on Call Processing, page 9-1.
For Cisco Expressway clustering, refer to the latest version of the Cisco Expressway Cluster Creation
and Maintenance Deployment Guide, available at
http://www.cisco.com/c/en/us/support/unified-communications/expressway-series/products-install
ation-and-configuration-guides-list.html
Capacity Planning
Cisco CMR Cloud meetings support up to 25 standards-based video devices, 500 WebEx participants
with video enabled, and 500 WebEx participants with audio only.
Note
Each screen in a multi-screen video device counts as one video device. For example, if a triple-screen
immersive system joins the CMR Cloud meeting, it consumes 3 video devices from the video device
capacity limit.
Capacity planning for Cisco CMR Cloud involves sizing of the components running within the
enterprise. The components could include:
Cisco Unified CM
Ensure that Unified CM has enough resources and capacity to handle the traffic generated by the
video endpoints and IP phones for CMR Cloud meetings. For capacity details, see the chapter on
Collaboration Solution Sizing Guidance, page 25-1.
11-73
Chapter 11
Cisco Expressway
Cisco Expressway must provide enough resources to handle the traversal call traffic for the
deployment. For capacity details, see the chapter on Collaboration Solution Sizing Guidance,
page 25-1.
Design Considerations
Consider the following recommendations when deploying Cisco CMR Cloud:
Enable UDP for media streaming in the firewalls for the optimal video experience.
Open network ports on firewalls to allow inbound and outbound signaling and media traffic. For port
range details, refer to Cisco Collaboration Meeting Rooms (CMR Cloud) Enterprise Deployment
Guide, available at
http://www.cisco.com/c/en/us/support/conferencing/webex-meeting-center/products-installati
on-and-configuration-guides-list.html
Ensure that Binary Floor Control Protocol (BFCP) is enabled in the Unified CM Neighbor Zone in
Cisco Expressway-C and that BFCP is also enabled in the SIP profile associated with the SIP trunk
between Unified CM and Expressway-C.
11-74
PART
Bandwidth Management
Dial Plan
Emergency Services
CH A P T E R
12
Once the network infrastructure has been put in place for your Cisco Unified Communications and
Collaboration System, call control and routing applications, components, and services can be layered on
top of this infrastructure. There are numerous applications and features that can, and in some cases
must, be deployed on the network infrastructure:
Bandwidth management Provides mechanisms for ensuring voice and video quality and for
preventing oversubscription of network bandwidth by limiting the number of calls that are allowed
on the network at a given time. With a combination of packet marking and re-marking, and advance
queuing mechanisms such as low latency or priority queuing, voice and video quality is guaranteed.
Similarly, call admission control enforces the overall call capacity of the call processing components
and available network bandwidth.
Dial plan Provides endpoint numbering, dialed digits analysis, and classes of restriction to limit
types of calls that a user can make.
Emergency services Provide essential information about the callers location and emergency
situation to the appropriate Public Safety Answering Point (PSAP) so that the caller receives a swift
response and the necessary help (for example, police, fire, or ambulance teams).
Directory and identity management services Lightweight Directory Access Protocol (LDAP)
provides applications with a standard method for accessing and potentially modifying the
information stored in a directory. Likewise, identity management and single sign-on ensure that user
access and identification is secure. These capabilities enable companies to centralize all user
information in a single repository available to several applications, resulting in better access to the
information and a reduction in maintenance costs through the ease of making adds, moves, and
changes.
The chapters in this part of the SRND cover the features, components, and services mentioned above.
Each chapter provides an introduction to the component or service, followed by discussions surrounding
architecture, high availability, and design considerations. The chapters focus on design-related aspects
of the applications and services rather than product-specific support and configuration information,
which is covered in the related product documentation.
This part of the SRND includes the following chapters:
12-1
Chapter 12
Architecture
at a given time to prevent oversubscription. This chapter covers quality of service (QoS) and call
admission control types, including location-based call admission control, as well as design and
deployment guidelines for successfully deploying QoS and admission control services.
Architecture
Call routing components and services such as call processing agents and IP and PSTN gateways rely on
the underlying network infrastructure for network connectivity and access. By connecting to the
underlying network infrastructure, call routing components and features are able to leverage end-to-end
network connectivity and quality of service to access both the enterprise and public telephone networks.
In turn, call routing applications and services provide basic Unified Communications and Collaboration
functions such as call control, dial plan, call admission control, and gateway services to other
applications and services in the deployment. For example, a Unified CM cluster connects to the IP
network through a switch in order to communicate with other devices and applications within the
network as well as to access other devices and services in other locations. At the same time, the
Unified CM cluster provides services such as phone registration and media resource provisioning and
allocation to call control components and services such as IP phones.
Further, just as call routing components rely on the network infrastructure for network connectivity, call
routing components and services are also often dependent upon each other for full functionality. For
example, while Unified CM provides registration and call routing services to various IP endpoints within
the network, it is completely dependent upon gateways and gateway services to route calls beyond the
enterprise.
12-2
Chapter 12
High Availability
As with the network infrastructure, critical Unified Communications and Collaboration call routing
services should be made highly available to ensure that required features and functionality remain
available if failures occur in the network or with individual call routing components. It is important to
understand the various types of failures that can occur and the design considerations around those
failures. In some cases, the failure of a single server or component (for example, a subscriber node in a
Unified CM cluster) might have little or no impact due to the redundant nature of the Unified CM
clustering mechanism. However, in other cases a single failure can impact multiple components or
services. For example, the failure of a PSTN or IP gateway could result in loss of access to the public
telephone network, and even though a call processing agent such as Unified CM is still available and
able to provide most features and services, it cannot route calls to the PSTN because there is no path
available if the gateway fails. To avoid these types of situations, you should deploy multiple PSTN
gateways to provide redundant gateway services, and you should configure the call processing agent to
handle call routing to both gateways as needed.
For features and services such as dial plan and bandwidth management, high availability considerations
include temporary loss of functionality due to network connectivity or call processing agent application
server failures, resulting in the inability of the call agent to route calls and therefore the inability for
callers to make calls. Oversubscription of the network could also occur if QoS and other call admission
control services are not available to the endpoints initiating a call. For example, if the call admission
control agent fails or loses connectivity to the network, the call may still go through but without the call
admission control service being aware of the call, thus potentially resulting in poor quality. To avoid
these types of scenarios, provide call admission control resiliency by deploying multiple call admission
control agents.
High availability considerations are also a concern for components and services such as video endpoints
and remote site survivability. For deployments with network-attached remote sites where devices are
leveraging call processing services from an agent in a central site, remote site survivability using SRST,
for example, can ensure that local phones within the remote site will still receive call processing services
in the event of a connectivity failure to the central site. Likewise, to ensure that video endpoints are
highly available, you can deploy more than one multipoint control unit (MCU) in case one fails.
Capacity Planning
The network infrastructures must be designed and deployed with consideration for the capacity and
scalability of the individual components and the overall system. Similarly, deployments of call routing
components and services must also be designed with attention to capacity and scalability considerations.
When deploying various call routing applications and services, not only is it important to consider the
scalability of the applications and services themselves, but you must also consider the scalability of the
underlying network infrastructure. Certainly the network infrastructure must have available bandwidth
and be capable of handling the additional traffic load that the call routing components will create.
Similarly, the call routing infrastructure and its components must be capable of handling all the required
device configurations and registrations as well as the call load or busy hour call attempts (BHCA),
For example, with call processing agents such as Unified CM, it is critical to assess the size of the
deployment in terms of number of users, endpoints, and calls per user per hour, and to deploy sufficient
resources to handle the required load. If a call processing agent is undersized and does not have sufficient
resources, features and services will begin to fail as the load increases. Two of the chief considerations
when attempting to size a call processing deployment are the call processing type and the call processing
hardware. Both of these are critical for sizing the system appropriately given the number of users,
locations, devices, and so on. As an example, Cisco Unified Communications Manager has a much
12-3
Chapter 12
Capacity Planning
higher capacity than Cisco Unified Communications Manager Express and should therefore be used for
larger deployments. In addition, the server platform selected to run the call processing agent will, in
many cases, determine the maximum load.
Capacity planning for remote site survivability is much the same in that it relies on backup call
processing hardware. Selecting the appropriate Cisco IOS platform to provide backup or survivable call
processing services typically begins with determining the number of devices or users that must be
supported at that site in the event that connectivity to the central site is disrupted. Equally critical in this
sizing exercise is the local PSTN gateway services. In the event of a central site connection failure, will
the local PSTN gateway have sufficient circuits to be able to route all calls without blocking during the
busiest hour? If the answer is no, adding additional gateways or trunks will be necessary to appropriately
size the remote site for backup call processing.
PSTN and IP gateways must also be sized appropriately for a deployment, so that sufficient capacity is
available to handle all calls in the busiest hour. In some cases, you might have to deploy multiple PSTN
or IP gateways to provide enough resources.
When configuring and sizing QoS and call admission control services, ensure that sufficient bandwidth
is available over network connections and in priority queues to support the required number of calls. If
sufficient bandwidth is not available, additional network capacity, gateways, and IP or telephony trunks
may be required.
Sizing dial plan services is also important. However, in most cases dial plan capacity in terms of the
number of endpoints or phone numbers, route patterns, or other dial plan constructs, is completely
dependent upon the type of call processing agent and platform used.
For components and services such as video telephony, appropriate sizing is just as critical. Capacity
planning considerations for video telephony center mainly on network bandwidth, available video ports,
and MCU sessions. In most cases additional capacity can be added by increasing the number of
application servers and MCUs or by upgrading server or MCU hardware with higher-scale models,
assuming the underlying network infrastructure is capable of handling the additional load.
For a complete discussion of system sizing, capacity planning, and deployment considerations related to
sizing, refer to the chapter on Collaboration Solution Sizing Guidance, page 25-1.
12-4
CH A P T E R
13
Bandwidth Management
Revised: June 15, 2015
Bandwidth management is about ensuring the best possible user experience end-to-end for all voice and
video capable endpoints, clients, and applications in the Collaboration solution. This chapter provides a
holistic approach to bandwidth management that incorporates an end-to-end Quality of Service (QoS)
architecture, call admission control, and video rate adaptation and resiliency mechanisms to ensure the
best possible user experience for deploying pervasive video over managed and unmanaged networks.
This chapter starts with a discussion of collaboration media and the differences between audio and video,
and the impact that this has on the network. Next an end-to-end QoS architecture for collaboration is
discussed, with techniques for how to identify and classify collaboration media and signaling for both
trusted and untrusted endpoints, clients, and applications. WAN queuing and scheduling strategies are
also covered, as well as bandwidth provisioning and admission control.
Note
The chapter on Network Infrastructure, page 3-1, lays the foundation for QoS in the LAN and WAN. It
is important to read that chapter and fully understand the concepts discussed therein. This chapter
assumes an understanding of those concepts.
13-1
Chapter 13
Bandwidth Management
Introduction
Introduction
The collaboration landscape is constantly evolving, and two areas that have changed dramatically are the
applications and the network. When Unified Communications was first introduced, it consisted primarily
of fixed hardware endpoints such as IP phones and room system endpoints connected to a completely
managed network where the administrators were able to implement Quality of Service (QoS) everywhere
throughout the network where media traversed. Over time, usage of the Internet and cloud-based services
such as WebEx have been added, which means that some of the collaboration infrastructure is now
located outside of the managed network and in the cloud. The office connectivity options have also
evolved, and companies are interconnecting remote sites and mobile users over the Internet either
directly connected over Cisco Expressway, for example, or over technologies such as Dynamic
Multipoint VPN (DMVPN). Figure 13-1 illustrates the convergence of a traditional on-premises
Collaboration solution in a managed (capable of QoS) network with cloud services and sites located over
an unmanaged (not capable of QoS) network such as the Internet. On-premises remote sites are
connected over this managed MPLS network where administrators can prioritize collaboration media
and signaling with QoS, while other remote sites and branches connect into the enterprise over the
Internet, where collaboration media and signaling cannot be prioritized or can be prioritized only
outbound from the site. Many different types of mobile and teleworkers also connect over the Internet
into the on-premises solution. So the incorporation of the Internet as a source for connecting the
enterprise with remote sites, home and mobile users, as well as other businesses and consumers, is
becoming pervasive and has an important impact on bandwidth management and user experience.
Figure 13-1
On-premises
UC Services
Central
Site
Cloud Services
QoScapable
B2B
B2C
Managed
WAN
Internet
MP
MPLS
PLS
VPN
VP
PN
N
Remote Sites
Home/Mobile Users
348834
DMVPN
DM
MV
New technologies and trends also mean an evolution of endpoints and user experiences and a plethora
of collaboration devices and options. The enterprise is moving from housing single-purpose,
single-media communications devices to multi-purpose, multi-media options. This is evident in trends
13-2
Chapter 13
Bandwidth Management
Introduction
such as Bring Your Own Device (BYOD), where users are bringing to the enterprise their compact and
powerful mobile devices and incorporating collaboration technologies such as instant messaging, video
collaboration and conferencing, and desktop sharing, to name a few, into their work processes, making
them more collaborative and efficient.
Collaboration media has also greatly evolved from fixed single-stream, fixed bit rate audio and video
streams connected point-to-point or via a multi-point bridge to multi-layer, multi-stream, adaptive bit
rate video sessions cascaded across multi-point bridges interconnecting a variety of devices. Figure 13-2
illustrates this evolution.
Figure 13-2
Temporal
layers
Adaptive
video bitrate
Multi-device
sessions
Multipoint
Bridge
Simulcast
multistreaming
Multipoint
Bridge
Collaboration
data
348835
Active cascading
Other technologies and trends that are currently and actively being adopted in the collaboration solution
include:
Multi-device, multi-stream sessions: voice, video, data sharing, and instant messaging
This evolution of managed versus unmanaged networks, new endpoints, and user experiences as well as
new technologies and trends have brought with them challenges such as:
How to manage the bandwidth and ensure a high-quality user experience over managed and
unmanaged networks
How to deploy video pervasively across the enterprise and optimize bandwidth utilization of the
available network resources
13-3
Chapter 13
Bandwidth Management
Collaboration Media
This chapter presents a strategy of leveraging smart media techniques in Cisco Video endpoints, building
an end-to-end QoS architecture, and using the latest design and deployment recommendations and best
practices for managing bandwidth to achieve the best user experience possible based on the network
resources available and the types of networks collaboration media are now forced to traverse.
Collaboration Media
This section covers the characteristics of audio and video streams in real-time media, as well as the smart
media techniques that Cisco Video endpoints employ to ensure high fidelity video in the face of packet
loss, delay, and jitter.
13-4
Chapter 13
Bandwidth Management
Collaboration Media
and picture complexity are the primary factors in determining the number of packets required for
transmission. Forward error correction can be used to estimate some lost information. However, in many
cases multiple IP packets are dropped in sequence. This makes the frame almost impossible to
decompress. The packets that were successfully sent represent wasted bandwidth. RTCP can be used to
request a new frame. Without a valid initial frame, subsequent frames will not decode properly.
Frame Types
The current generation of video coding is known by three names; H.264, MPEG4 part 10, and Advanced
Video Coding (AVC). As with earlier codecs, H.264 employs spatial and temporal compression. Spatial
compression is used on a single frame of video as described previously. These types of frames are known
as I-frames. An I-frame is the first picture in a GOP. Temporal compression takes advantage of the fact
that little information changes between subsequent frames. Changes are a result of motion, although
changes in zoom or camera movement can result in almost every pixel changing. Vectors are used to
describe this motion and are applied to a block. A global vector is used if the encoder determines all
pixels moved together, as is the case with camera panning. In addition, a difference signal is used to
fine-tune any error that results. H.264 allows variable block sizes and is able to code motion as fine as
pixel. The decoder uses this information to determine how the current frame should look based on the
previous frame. Packets that contain the motion vectors and error signals are known as P-frames. Lost
P-frames usually results in artifacts that are folded into subsequent frames. If an artifact persists over
time, then the likely cause is a lost P-frame.
Figure 13-3 illustrates how this works in a basic manner:
1.
An I-frame (Intra-coded picture) is the entire picture encoded as a static image and sent as a group
of packets. This frame does not reference any other frame, and the decoder requires only this frame
to build the entire image. In this case the image is of a little hiker hiking through the mountains.
2.
Next a P-frame (Predicted picture) is sent, which is a frame based on a previously encoded frame
(in this case the I-frame), and only the differences from that I-frame are encoded. The decoder takes
these differences and applies them to the I-frame that it had. In this case it shows the little hiker
moving up the hill. Because only the little hiker and his movement have changed from the last
I-frame, this P-frame is much smaller and represents fewer packets and thus less bandwidth to be
transmitted.
3.
The next P-frame is sent and is a prediction from the last P-frame sent. As in the P-frame from step
2, this P-frame shows the difference between the last movement of the hiker up the hill and this new
movement of the hiker. This progression continues until there is a larger amount of change from the
previous image, thus requiring a new I-frame.
13-5
Chapter 13
Bandwidth Management
Collaboration Media
Figure 13-3
Encoding Basics
348836
H.264 also implements B-frames. This type of frame fills in information between P-frames. This means
that the B-frame must be held until the next P-frame arrives, before the B-frame information can be used.
B-frames are not used in all modes of H.264. The encoder decides what type of frame is best suited.
There are typically more P-frames than I-frames. Lab analysis has shown TelePresence I-frames to
generally be 64 Kbytes wide (50 packets @ 1,316 bytes), while P-frames average 8 Kbytes wide
(9 packets at 900 bytes). So I-frames are larger and create the spikes in bit rate in comparison to
P-frames.
13-6
Chapter 13
Bandwidth Management
Collaboration Media
Figure 13-4
Video Packets
Audio Packets
1400
1400
1000
1000
P-Frame
I-Frame
P-Frame
Bytes
600
600
Audio
Samples
200
20 ms
348837
200
33 msec
Time
As can be seen from Figure 13-4, the audio packets are the same size, sent at exactly the same time
intervals, and they represent a very smooth stream. Video, on the other hand, sends a larger group of
packets over fixed intervals and can vary greatly from frame to frame. Figure 13-4 shows the difference
in the number of packets and packet sizes for an I-frame as opposed to P-frames. This translates to a
stream of media that is very bursty in nature when compared to audio. Figure 13-5 illustrates the
bandwidth profile over time of an HD video stream. Note the large bursts when I-frames are sent.
Figure 13-5
3500
I-Frames
Bandwidth (kbps)
3000
2500
2000
1500
1000
Time (s)
348838
500
13-7
Chapter 13
Bandwidth Management
Collaboration Media
Figure 13-5 shows an HD video call, 720p30 @ 1,920 kbps (1,792 kbps video + 128 kbps audio). The
graph shows the video bandwidth (including L3 overhead), and the red line indicates average bit rate.
While audio and video are both transported over UDP and sensitive to loss and delay, they are quite
different in their network requirements and profile. Audio is a constant bit rate and has a smaller
footprint compared to video, as well as a narrower operational range of 1:6 ratio when comparing the
lowest bit-rate audio codec to one of the highest bit-rate codecs. Video, on the other hand, has a variable
bit rate (is bursty) and has a medium to large footprint when compared to audio, as well as a wide
operational range of 1:40 (250p at 15 fps vs 1080p at 60 fps). Figure 13-6 illustrates some of these
differences.
Figure 13-6
AUDIO
Bandwidth:
Bandwidth
Delay-sensitive
Operational bandwidth
VIDEO
Loss-sensitive
Bandwidth:
Bandwidth
1080p60 (6 Mbps)
Loss-sensitive
Delay-sensitive
Operational bandwidth
348839
Time
The important point to keep in mind is that audio and video, while similar in transport and sensitivity to
loss and delay, are quite different with regard to managing their bandwidth requirements in the network.
It should also be noted that, while video is pertinent to a full collaboration experience, audio is critical.
If, for example, video is lost during a video call due to a network outage or some other network related
event, communication can continue provided that audio is not lost during this outage. This is a critical
concept when thinking through the network requirements of a collaboration design such as QoS
classification and marking.
Resolution
The sending station determines the video resolution and, consequently, the load on the network. This is
irrespective of the size of the monitor used to display the video. Observing the video is not a reliable
method to estimate load. Common high definition formats are 720i, 1080i, 1080p, and so forth. In
addition to high resolution, there is also a proliferation of lower quality video that is often tunneled in
HTTP (or in some cases HTTPS) and SSL (see Table 13-1). Typical resolutions include CIF (352x288)
and 4CIF (704x576). These numbers were chosen as integers of the 16x16 macro blocks that are used
by the DCT (22x18) and (44x36) macro blocks respectively.
13-8
Chapter 13
Bandwidth Management
Collaboration Media
Table 13-1
Format
Resolution
Typical Bandwidth
176x144
260 kbps
CIF
352x288
512 kbps
4CIF
704x576
1 Mbps
SD NTSC
720x480
720 HD
1280x720
1 to 8 Mbps
1080 HD
1080x1920
5 to 8 Mbps H.264
12+ Mbps MPEG-2
Network Load
The impact of resolution on the network load is generally a squared factor; an image that is twice as big
will require four times the bandwidth. In addition, the color sampling, quantization, and frame rate also
impact the amount of network traffic. Standard rates are 30 frames per second (fps), but this is an
arbitrary value chosen based on the frequency of AC power. In Europe, analog video is traditionally
25 fps. Cineplex movies are shot at 24 fps. As the frame rate decreases, the network load also decreases
and the motion becomes less life-like. Video above 24 fps does not noticeably improve motion.
The sophistication of the encoder also has a large impact on video load. H.264 encoders have great
flexibility in determining how best to encode video, and with this comes complexity in determining the
best method. For example, MPEG4.10 allows the encoder to select the most appropriate block size
depending on the surrounding pixels. Because efficient encoding is more difficult than decoding, and
because the sender determines the load on the network, low-cost encoders usually require more
bandwidth than high-end encoders. H.264 coding of real-time CIF video will drive all but the most
powerful laptops well into 90% CPU usage without dedicated media processors.
Table 13-2 through Table 13-4 show average bandwidth utilization ranges based on endpoint and
resolution. These tables are provide only as an example of the bandwidth ranges based on resolution of
common TelePresence and desktop video endpoints. Refer to the current product documentation for the
latest numbers relevant to the endpoints in question.
Table 13-2
MX200
SX20
EX90
TX9000
Resolution
Lowest
Highest
Lowest
Highest
Lowest
Highest
Lowest
Highest
720p30 (1280x720)
736 kbps
1.2 Mbps
812 kbps
1.2 Mbps
812 kbps
1.2 Mbps
3.1 Mbps
6.4 Mbps
1080p30 (1920x1080)
2.6 Mbps
5.7 Mbps
2.6 Mbps
6.2 Mbps
2.5 Mbps
6.1 Mbps
8.8 Mbps
11.9 Mbps
N/A
2.3 Mbps
N/A
2.3 Mbps
N/A
2.4 Mbps
N/A
N/A
1. For more information on TelePresence endpoints, refer to the bandwidth usage white paper available at
http://www.cisco.com/c/dam/en/us/products/collateral/collaboration-endpoints/tested_bandwidth_whitepaperx.pdf.
13-9
Chapter 13
Bandwidth Management
Collaboration Media
Table 13-3
Resolution
240p30 (432x240)
360p30 (640x360)
480p30 (848x480)
576p30 (1024x576)
720p30 (1280x720)
1080p30 (1920x1080)
2 to 4 Mbps
1. For more information on the DX Series, refer to the latest version of the Cisco DX Series Administration Guide, available at
http://www.cisco.com/c/en/us/support/collaboration-endpoints/desktop-collaboration-experience-dx600-series/products-ma
intenance-guides-list.html.
Table 13-4
Resolution
w144p30 (256x144)
156 kbps
w288p30 (512x288)
320 kbps
w448p30 (768x448)
570 kbps
w576p30 (1024x576)
890 kbps
720p30 (1280x720)
1.3 Mbps
1. For more information on Jabber, refer to the latest version of the Cisco Jabber Deployment and Installation Guide, available
at http://www.cisco.com/c/en/us/support/unified-communications/jabber-windows/products-installation-guides-list.html.
Multicast
Broadcast video lends itself well to taking advantage of the bandwidth savings offered by multicast. This
has been in place in many networks for years. Recent improvements to multicast simplify the
deployment on the network. Multicast will play a role going forward; however, multicast is not used in
all situations. Some applications such as multipoint TelePresence use a dedicated MCU to replicate
video. The MCU can make decisions concerning which participants are viewing each sender. The MCU
can also quench senders that are not being viewed.
Transports
MPEG4 uses the same transport as MPEG2. A PES consists of 188-byte datagrams that are loaded into
IP. The video packets can be loaded into RTP/UDP/IP or HTTP(S)/TCP/IP.
Video over UDP is found with dedicated real-time applications such as multimedia conferencing and
TelePresence. In this case, an RTCP channel can be set up from the receiver toward the sender. This is
used to manage the video session. RTCP can be used to request I-frames or report capabilities to the
sender. UDP and RTP each provide a method to multiplex channels. Audio and video typically use
different UDP ports but also have unique RTP payload types. Deep packet inspection (DPI) can be used
on the network to identify the type of video and audio that is present. Note that H.264 also provides a
mechanism to multiplex layers of the video.
13-10
Chapter 13
Bandwidth Management
Collaboration Media
Buffering
Jitter and delay are present in all IP networks. Jitter is the variation in delay. Delay is generally caused
by interface queuing. Video decoders can employ a play-out buffer to smooth out jitter found in the
network. There are limitations to the depth of this buffer. If it is too small, then drops will result. If it is
too deep, then the video will be delayed, which could be a problem in real-time applications such as
TelePresence. Another limitation is handling dropped packets that often accompany deep play-out
buffers. If RTCP is used to request a new I-frame, then more frames will be skipped at the time of
re-sync. The result is that dropped packets have a slightly greater impact in video degradation than they
would have if the missing packet had been discovered earlier. Most codecs employ a dynamic play-out
buffer.
Summary
Video can dramatically impact the performance of the network if planning does not properly account for
this additional load. This chapter attempts to assist administrators in managing real-time video in
enterprise networks.
Encoder Pacing
The number of packet can increase dependent on the frame type (I or P) as well as the number of packets
required, which means that bursts of packets can show up at the beginning, middle, or end of a 33 ms
time interval. This creates spikes in bandwidth as the packets are put onto the wire. Encoder pacing is a
simple technique used to spread the packets as evenly as possible across the 33 ms interval in order to
smooth out the peaks of the bursts of bandwidth. Figure 13-7 illustrates this technique.
13-11
Chapter 13
Bandwidth Management
Collaboration Media
Figure 13-7
1400
Encoder Pacing
P-Frame
I-Frame
P-Frame
P-Frame
1000
Bytes
600
200
33 ms
1400
P-Frame
Time
I-Frame
P-Frame
1000
Bytes
348840
600
200
33 ms
Time
The top image in Figure 13-7 shows packets being placed on the wire without encoder pacing, and the
bottom image is with encoder pacing. As each frame is packetized onto the wire in a 33 ms interval, an
endpoint packet scheduler disperses packets as evenly as possible across that single interval. Large
I-frames might have to be "spread" over two or three frame intervals, and the encoder might then skip
one or two frames to stay within a bit rate budget. This smooths out the peaks in bandwidth utilization
over the same time frame.
13-12
Chapter 13
Bandwidth Management
Collaboration Media
I-frame
348841
Figure 13-8
Predicted
portion
GDR frames
Intra-macroblock
portion
D
d
Decoder
Encoder
E
d
LTRF1
P1 P3
P2 P4
...
P5
...
LTRF1
Repair P-Frame
Built from last synced LTRF
P3 Dropped
Encoder
ACK LTRF1
P4
...
P5
P4
...
Decoder
P3
P5
P1
P2
P2
P1
348842
Figure 13-9
OOS (P4)
As Figure 13-9 illustrates, LTRFs keep the encoder and decoder in sync with active feedback messages.
The encoder instructs the decoder to store raw frames at specific sync points as Long Term Reference
Frames (part of the H.264 standard), and the decoder uses "back channel" (RTCP) to acknowledge the
LTRFs. When a frame is lost, the encoder creates a Repair P-frame based on the last synchronized LTRF
instead of generating a new I-frame, thus saving bandwidth.
13-13
Chapter 13
Bandwidth Management
Collaboration Media
R2 Dropped
R2
Encoder
Dec
Decoder
coder
R1
Binary XOR
LTRF
Repair-P
...
...
R1
R2
0111010001
1000011001
0001100
1110010101
1011010010
1010010
R1
1001000100
0011001011
1011110
FEC
FEC
Binary
XOR
R2
R2 Rebuilt from
FEC Frame
348843
FEC
As Figure 13-10 illustrates, FEC enables the decoder to recover from a limited amount of packet loss
without losing synchronization. It can be applied at different levels (for example, X FEC packets every
N data packets) to protect "important" frames in lossy environments. The correction code can be basic
(binary XOR) or more advanced (Reed-Solomon). The trade-off is increased bandwidth usage, therefore
it is best suited for non-bursty loss.
13-14
Chapter 13
Bandwidth Management
Collaboration Media
Rate Adaptation
Rate adaptation, or dynamic bit rate adjustments, adapt the call rate to the variable bandwidth available,
down-speeding or up-speeding the video bit rate based on the packet loss condition (see Figure 13-11).
Once the packet loss has decreased, up-speeding will occur. Some endpoints use a proactive
sender-initiated approach by utilizing RTCP. In this case the sender is constantly reviewing the RTCP
receiver reports and adjusting its bit rate accordingly. Other endpoints use a receiver-initiated approach,
adjusting via call signaling (H.323 flow control, TMBRR, SIP Re-invite) or an explicit request in the
RTCP messages.
Figure 13-11
Rate Adaption
Packet
Loss
Video
Bitrate
Bitr
SLOW
OWN
DOWN
t1
t1
t2
Sender
d
t2
Receiver
RR 1
RR 2
RR 3
RR
t1
Time Interval
348844
As illustrated in Figure 13-11, the receiver observes delay and packet loss over periods of time and
signals back using RTCP Receiver Reports (RR). the reports cause the sender to adjust its bit rate to
adapt to network conditions (down-speeding or up-speeding of bit rate).
Two approaches are possible with rate adaption:
Receiver-initiated adjustment via call signaling (H.323 flow control, TMBRR, SIP Re-invite) or
explicit request in RTCP message
Burstiness of traffic and mobility of the endpoints make deterministic provisioning for interactive
video difficult for network administrators.
Media resiliency mechanisms help mitigate the impact of video traffic on the network and the impact
of network impairments on video. (See Table 13-5.)
Dynamic rate adaptation creates an opportunity for more flexible provisioning models for
interactive video in enterprise networks.
Media resiliency and rate adaptation also help preserve the user experience when video traffic
traverses the Internet or non-QoS-enabled networks.
Summary
13-15
Chapter 13
Bandwidth Management
Table 13-5
Endpoint or Bridge
Encoder Pacing
Rate Adaption
FEC
LTRF Repair
8800 Series
Yes
No
No
No
9900 Series
No
No
No
No
DX Series
Yes
Yes
No
No
WebEx
Yes
Yes
Yes
No
TX Series
Yes
Yes
No
Yes
Jabber
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
TelePresence Server
Yes
Yes
Yes
Yes
MCU
Yes
Yes
Yes
Yes
13-16
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
Figure 13-12
Network
Management
EF
WAN
AF41
1
CS3
Classification
Queuing and
Scheduling
Provisioning and
Admission
Control
Monitoring
Troubleshooting
Optimization
348846
Identification
13-17
Chapter 13
Bandwidth Management
service providers re-mark traffic for usage throughout their network (service provider network). Because
of this situation, it is important to re-mark the traffic back to the appropriate values to ensure continuity
through the enterprise network end-to-end.
In a Cisco converged network with Cisco IP phones and video endpoints, switches can be configured to
detect the phones using Cisco Discovery Protocol (CDP), and the switch can then trust the Differentiated
Services Code Point (DSCP) marking of packets that the Cisco IP phones and video endpoints send
without trusting the markings of the PC connected to the switch port of the IP phone or video endpoint.
This is referred to as conditional trust and is commonplace in protected VLANs where only Cisco IP
phones are admitted (referred to as voice VLANs) and where their packet marking is trusted by the
switches and passed through the network unchanged. Administrators generally do not trust the traffic
that comes from VLANs where untrusted clients (such as PCs or Macs) are typically located (referred
to as data VLANs). The packets that come from devices in the data VLAN or equivalent areas of the
network typically get remarked to best effort (IP DSCP 0).
From a trust perspective, there are three main categories of endpoints:
Trusted endpoints Secure PCs and servers, video conferencing endpoints, access points, analog
and video conferencing gateways, and other similar devices where CDP is not available
Conditionally trusted endpoints Cisco IP phones as well as Cisco TelePresence endpoints that
support CDP
Trust Boundaries
Trust Boundary
Untrusted Devices
Trust Boundary
Trusted
Devices
Trust Boundary
348847
Conditionally
Trusted Devices
13-18
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
The trust boundary should be set as close to the endpoints as technically and administratively feasible.
The recommendation is to set trust on the switch and use voice VLANs for collaboration media and
signaling, and use data VLANs for non-collaboration data traffic. See the section on Campus Access
Layer, page 3-4, for more information on Layer 2 access design.
For trusted and conditionally trusted endpoints, the DSCP marking of packets on ingress into the switch
are trusted and rewritten to the same value on egress. Figure 13-14 illustrates marking of audio, video,
and signaling traffic for trusted endpoints, and the switch trusting these markings.
Figure 13-14
Trusted
EF
EF
AF41
AF41
CS3
CS3
348848
SWITCH
For Cisco switches configured with trusted or conditionally trusted ports, the switch either uses CoS to
map to DSCP or it uses the original DSCP and maps it to the outbound packet IP header DSCP.
Figure 13-15 illustrates the inbound packet marking at Layer 2 (CoS) and Layer 3 (DSCP); the type of
trust trusted (CoS Trust or DSCP Trust) or untrusted; and the internal switch packet rewriting process
based on CoS trust or DSCP trust.
13-19
Chapter 13
Bandwidth Management
Figure 13-15
TRUST
INBOUND PACKET
1
CoS = 5
Untrusted
no [mls] qos trust
DSCP = 46
CoS = 5
Trust CoS
[mls] qos trust cos
DSCP = 46
CoS = 0
DSCP = 0
CoS = 5
Trust DSCP
[mls] qos trust dscp
DSCP = 46
L2 Header
CoS = 5
L3 Header
DSCP = 46
CoS = 5
Internal
DSCP = 46
CoS = 5
DSCP = 40
DSCP = 46
348849
Internal
DSCP = 40
Multi-Layer Switching (MLS) commands are used in Figure 13-15 as an example only. MLS platforms
include the Cisco 2960, 3560, and 3750 Series switch platforms. On all other currently shipping switch
platforms (including the Cisco 3650, 3850, 4500, 6500, and 6800 Series switch platforms) trust is
enabled by default.
Figure 13-15 shows three events:
1.
A packet marked CoS 5 and DSCP 46 comes inbound on an untrusted port. An internal DSCP of 0
(BE) is used to rewrite the outbound packet CoS and DSCP to 0.
2.
A packet marked CoS 5 and DSCP 46 comes inbound on a trusted port (CoS trust). A lookup is done
on a CoS-to-DSCP mapping table to map CoS 5 to an internal DSCP of 40. An internal DSCP of 40
is used to rewrite the outbound packet CoS to 5 and DSCP to 40. Note that the CoS-to-DSCP map
table has defaults but can be modified to any static CoS-to-DSCP mapping. For example, CoS 5
could be mapped to DSCP 46.
3.
A packet marked CoS 5 and DSCP 46 comes inbound on a trusted port (DSCP trust). An internal
DSCP of 46 (EF) is used to rewrite the outbound packet CoS to 5 and DSCP to 46 (EF).
13-20
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
For CDP-capable Cisco IP Phones, Cisco CTS, Cisco IP Video Surveillance cameras, and Cisco Digital
Media Players (as opposed to software clients such as Jabber), we recommend using the CDP conditional
trust and passing the marking of the trusted endpoint through the network. When electing to trust Cisco
IP Phones, you must trust CoS because the phones can re-mark only PC traffic at Layer 2. Trusted
endpoints derive their DSCP marking from Unified CM. DSCP for endpoints is configured in the
Unified CM service parameters under Clusterwide Parameters (System - QoS).
Unified CM houses the QoS configuration for endpoints in two places: in the service parameters for the
CallManager service and in the SIP Profile applicable only to SIP devices. The SIP Profile configuration
of QoS settings overrides the service parameter configuration. This allows the Unified CM administrator
to set different QoS policies for groups of endpoints (see Bandwidth Management Design Examples,
page 13-91). During endpoint registration, Unified CM passes this QoS configuration to the endpoints
in a configuration file over TFTP. This configuration file contains the QoS parameters as well as a
number of other endpoint specific parameters. For QoS purposes there are two categories of video
endpoints: TelePresence endpoints (any endpoint with TelePresence in the phone type name) and all
other non-TelePresence video endpoints referred to as "UC Video Endpoints" in this document.
Figure 13-16 illustrates how the two categories of Cisco video endpoints derive DSCP. Keep in mind that
these categories apply only to QoS and call admission control (see the section on Enhanced Location
CAC for TelePresence Immersive Video, page 13-60).
Figure 13-16
SIP Profile
Device
Registration
Config File
Unified CM
TelePresence
Endpoints
TelePresence Applicable
TelePresence
T
Applicab DSCP settings:
DSCP for Audio Calls
DSCP for TelePresence
T Presenc Calls
Tele
DSCP for Audio Portion of TelePresence
Calls1
UC Video
Endpoints
348850
QoS
Service
Parameters
13-21
Chapter 13
Bandwidth Management
The parameters DSCP for Audio Portion of Video Calls and DSCP for Audio Portion of TelePresence
Calls, shown in Figure 13-16, currently are not supported on all video endpoints. See Table 13-7 for
information on which endpoint types support these parameters.
The configuration file is populated with the QoS parameters from the CallManager service parameters
or the SIP Profile, when configured, and sent to the endpoint upon registration. The endpoint then uses
the correct DSCP parameters for each type of media stream, depending on which category of endpoint
it is. Table 13-6 lists the DSCP parameter, the type of endpoint, and the type of call flow determining the
DSCP marking of the stream.
Table 13-6
DSCP Parameter
TelePresence
Endpoint
UC Video
Endpoint
Call Flow
Yes
Yes
Voice only
N/A
Yes
Yes
N/A
N/A
1. The DSCP settings for Multi-Level Priority and Preemption (MLPP) are not discussed here. For more information about
MLPP and QoS settings, refer to the latest version of the System Configuration Guide for Cisco Unified Communications
Manager.
2. This parameter is not currently supported on all video endpoints. See Table 13-7 for information on which endpoint types
support this parameter.
Table 13-7
Endpoint Support for DSCP Parameters for the Audio Portion of Video and
TelePresence Calls
Video Endpoint
8800 Series
No
N/A
8900 Series
No
N/A
9900 Series
No
N/A
DX Series
Yes
Yes1
TX Series
N/A
Yes
IX Series
N/A
No
N/A
Yes
13-22
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
Table 13-7
Endpoint Support for DSCP Parameters for the Audio Portion of Video and
TelePresence Calls (continued)
Video Endpoint
Yes
Yes
N/A
1. To enable the DX Series to use DSCP for TelePresence calls as well as DSCP for the audio portion of TelePresence calls, you
must also enable the DSCP parameters for the video promotion of those calls. (For more information on enabling this feature,
refer to the latest version of the System Configuration Guide for Cisco Unified Communications Manager.)
Due to these new features and system-wide capabilities, the current DSCP defaults are not always the
recommended values. This is discussed in further detail in the Bandwidth Management Design
Examples, page 13-91.
Untrusted Endpoints and Clients
For untrusted endpoints the DSCP marking of packets on ingress into the switch is untrusted and
rewritten to 0 (BE). Figure 13-17 illustrates untrusted endpoints marking audio, video, and signaling
traffic, and the switch rewriting this value on the outbound packet.
Figure 13-17
Untrusted
EF
BE
AF41
BE
CS3
BE
348851
SWITCH
In general, trusting markings that can be set by users on their PCs, Macs, or hand-held mobile devices
is not recommended. Users can abuse provisioned QoS policies if permitted to mark their own traffic
(have administrative control of the OS). For example, if a DSCP of EF has been provisioned over the
network, a PC user can configure all their traffic to be marked to EF, which will hijack network priority
queues to service non-real-time traffic. Such abuse could easily ruin the service quality of real-time
applications throughout the enterprise. On the other hand, if enterprise controls are in place that centrally
administer PC QoS markings, such as Global Policy Objects in Windows environments, then it may be
possible to trust the PC markings. For Macs running OSX and hand-held mobile clients, the question
remains whether to trust the markings from them or not. This method is covered in more detail in the
section "Utilizing the Operating System for QoS Trust, Classification and Marking". The general rule is
not to trust any of these personal computing devices, and a method for re-marking traffic is required.
13-23
Chapter 13
Bandwidth Management
A different method from trust is required to ensure that the media and signaling streams from the
software clients such as Jabber are able to get classified and marked appropriately. One method consists
of mapping identifiable media and signaling streams based on specific protocol ports, such as UDP and
TCP ports, then making use of network access lists to remark QoS of the signaling and media streams
based on those protocol port ranges. This method applies to all Cisco Jabber clients because they all
behave similarly when allocating media and signaling port ranges. This method ranges from using the
network to create policies based on access lists to accomplish packet DSCP remarking, to using the
Windows OS itself (Jabber for Windows clients only apply here) and then trusting the marking from the
PC in the network.
This method is the most widely deployed and recommended method to achieve QoS with Cisco Jabber
clients simply because of the trust issue. The Jabber clients are Cisco Jabber for Windows, Cisco Jabber
for Mac OS, Cisco Jabber for iPhone, Cisco Jabber for iPad and Cisco Jabber for Android.
The concept is simple. As all of the traffic from the PC cannot be trusted, an access list is used in the
network access layer equipment to identify the media and signaling streams based on UDP port ranges
and to re-mark them to appropriate values. Although this technique is easy to implement and can be
widely deployed, it is not a secure method.
Figure 13-18 illustrates using network access control lists (ACLs) to map identifiable media and
signaling streams to DSCP.
Figure 13-18
UDP Ports
Untrusted
EF
3xxx
ACL
4xxx
AF41
5060-5061
CS3
348852
SWITCH
Figure 13-18 illustrates the following example ACL-based QoS policy for Jabber clients:
Note
The following example access control list is based on the Cisco Common Classification Policy Language
(C3PL). Refer to your specific switch or router configuration guides to achieve the same policy on a
Cisco device that does not support C3PL or for any updated commands in C3PL. This configuration is
portable to all currently shipping switches including Modular QoS CLI-MQC, Multi-Layer Switching
(MLS), and C3PL.
13-24
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
! This section
access-list
access-list
access-list
configures
100 permit
101 permit
102 permit
the
udp
udp
tcp
ACLs to
any any
any any
any any
match
range
range
range
13-25
Chapter 13
Bandwidth Management
Upper Half
for Video
4000-4999
Registration
Lower Half
for Audio
Config File
Unified CM
SEP[MAC_addr].cnf.xml
Media Port Range UDP
3000-4999
JJabber
bb Cli
Clients
3000-3999
Jabber splits
media range
in half
SIP
Signaling
TCP 5060
348853
Figure 13-19
Jabber can also use the Media Port Range > Separate Port Range for Audio and Video configuration.
In this configuration the Unified CM administrator can configure a non-contiguous audio and video port
range as illustrated in Figure 13-20.
Figure 13-20
5000-5999
Registration
Lower Half
for Audio
Config File
Unified CM
SEP[MAC_addr].cnf.xml
Media Port Range UDP
3000-3999
Video Port Range UDP
Upper Half
for Video
JJabber
bb Cli
Clients
Jabber splits
media range
in half
3000-3999
SIP
Signaling
TCP 5060
348854
5000-5999
Due to the behavior of Jabber clients regarding UDP port range assignment, it is often not possible to
map Enhanced Locations Call Admission Control (EL-CAC) bandwidth deductions correctly with QoS
markings. CAC deducts bandwidth for audio-only calls out of the voice pool, while both audio and video
bandwidth of a video call is deducted out of the video pool. To be consistent with the admission control
logic, audio streams of voice-only calls would need to be marked as EF while both audio and video
streams of video calls would need to be marked AF41. The differentiation of audio between audio of
voice-only calls and audio of video calls is not possible when using Cisco Jabber clients and UDP port
ranges to map identifiable media streams. As a result, this technique is effective to achieve QoS only.
Therefore, we recommend over-provisioning the priority queue for EF traffic to account for the audio of
video sessions from Jabber clients that will send audio as EF, or using an alternate DSCP. Some
strategies are discussed in the Bandwidth Management Design Examples, page 13-91.
13-26
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
Caution
Security Alert: By utilizing identifiable media streams for QoS classification at the network level, the
trust model does not extend to the application itself. Apart from prioritizing streams from the intended
application, other applications "could" potentially be configured to use the same identification criteria
(media port range), and therefore achieve network prioritization. Because this unintended traffic would
not be accounted for in CAC or in the provisioning of the network, severe overall impacts to real-time
conversations can occur. It is good practice to define restricted port ranges to identify media streams
when possible.
When utilizing this technique, it is important to ensure that the audio portion of these video calls that
will be re-marked to the audio traffic class (EF) and the video portions re-marked to the video traffic
class (AF4) are provisioned in the network accordingly. Figure 13-21 is an example of placing audio
traffic into a Priority Queue (PQ) and video traffic into a Class Based Weighted Fair-Queue (CBWFQ).
Note that, because it is not possible to differentiate the audio from voice-only calls versus the audio from
video calls with port ranges in Cisco Jabber endpoints, all audio using this technique will be re-marked
to EF. It is important to provision the PQ adequately to support voice-only and the audio portion of video
calls. An example of such provisioning is illustrated in Figure 13-21. For more information on the design
and deployment recommendations for provisioning queuing and scheduling in the network, see the
section on WAN Queuing and Scheduling, page 13-32.
Provisioning Jabber QoS in the Network
EF
EF
EF
EF
AF41
AF41
PQ
CBWFQ
other queues
348855
Figure 13-21
According to RFC 3551, when RTCP is enabled on the endpoints, it uses the next higher odd port. For
example, a device that establishes an RTP stream on port 3500 would send RTCP for that same stream
on port 3501. This function of RTCP is also true with all Jabber clients. RTCP is common in most call
flows and is commonly used for statistical information about the streams and to synchronize audio and
video in video calls to ensure proper lip-sync. In most cases, video and RTCP can be enabled or disabled
on the endpoint itself or in the common phone profile settings.
13-27
Chapter 13
Bandwidth Management
Note
When deploying Jabber clients in networks where SCCP voice-only endpoints are deployed, the SCCP
endpoints use a non-configurable hard-coded range of 16384 to 32767 for voice-only calls. Due to this,
SCCP voice-only calls could run over the same range as SIP video-enabled endpoint calls if you do not
change the media port range for SIP devices. If you are deploying a collaboration solution with endpoints
that are configured to use SCCP, then we recommend setting the media port range of Jabber clients
outside of the 16384 to 32767 range. The above examples of 3000 to 4999 for video-enabled Jabber
clients and 3000 to 3999 for voice-only Jabber clients work very well to avoid overlap with SCCP
endpoints.
The recommendation to avoid overlap applies to other SIP-based video endpoints as well. To avoid
overlap with SCCP-based audio endpoint ranges, the SIP-based video endpoints should also be allocated
a port range that does not overlap with SCCP-based audio port range (16384 to 32767) or the Jabber
clients media port range.
13-28
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
routing interfaces. In most campus networks these would be VLAN interfaces, but they could also be
Fast Ethernet or Gigabit Ethernet interfaces. Figure 13-22 illustrates the areas of the network where the
various types of trust are applied in relation to the places in the network access, distribution, core, and
WAN Edge.
Figure 13-22
Conditionally-Trusted Endpoints
Conditionally-Truste
Access
Distribution
Core
WAN Edge
Immersive Endpoints
IOS GW
Mac/PC
PC/GPO
MAC
Handheld
WiFi AP
CUBE
MAC
WiFi
Controller
Handheld
UCM
1UCx
IM&P
Expressway
MCU/TS
Conditionally Trusted
348856
Fully Trusted
Untrusted
Layer 3 Remarking
Utilizing the Operating System for QoS Trust, Classification, and Marking
Another method of QoS trust for Cisco Jabber clients is to allow the operating system on which the
applications run to mark the QoS of the media and signaling at the request of the application. The benefit
of this method is that it allows the network operators to extend the QoS trust model to the operating
system itself, and then they can configure the network to "trust" the QoS markings and pass them through
the network. It is not a common enterprise practice to extend QoS trust to the Windows PCs, Mac OS,
and hand-held devices. The reason for this is that this method trusts all traffic from the device, not just
traffic from authenticated application communication. These applications can be installed and used on
these devices to "hijack" a priority QoS and defeat the original purpose of deploying QoS in the first
place. Through administrative global policies, administrators can manage some operating systems such
as Windows OS or user access controls to ensure that the OS does not accept unwanted applications or
configurations. In these cases, it might be acceptable to use this method of QoS trust.
On Windows 7 and 8 operating systems it is necessary to configure specific policies, while in Mac OS,
Apple iOS, and Android devices the OS natively marks at the request of the application without any
specific configuration necessary.
13-29
Chapter 13
Bandwidth Management
The following sections discuss the Cisco Jabber clients and describe how each operating system
functions with regard to application QoS classification and marking. Everything described in these
sections relates to Layer 3 DSCP marking and not Layer 2 Class of Service (CoS):
Caution
Security Alert: In a pure Windows 7 (and later versions) environment, utilizing only GPOs would allow
an enterprise to unconditionally trust all data sent from those Windows devices. Because it is highly
unlikely for such homogeneous environments to exist in real-world deployments, extra effort has to be
taken to separate the trust model for GPO-based devices from other operating systems and devices in the
same VLANs or on similar ports in the access layer.
GPOs are very similar to network access lists in how they allow the operating system to mark a specific
application's QoS based on protocols, ports, and application executable. Figure 13-23 illustrates the
process of QoS re-marking in Windows 7 and 8 with Jabber for Windows.
13-30
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
Figure 13-23
Voice Stream
(EF)
(Ports:3000-3999)
Voice Stream
Video Call
(Voice and Video
Streams)
(AF41)
(Ports:4000-4999)
Video Stream
SIP Signaling
Voice Stream
(EF)
(Ports:3000-3999)
Voice Stream
Video Call
(AF41)
(Ports:5000-5999)
Video Stream
SIP Signaling
348857
The process illustrated in Figure 13-23 starts with a QoS Group Policy that defines the IP address range
(or any), the protocol (UDP), and port ranges (audio 3000 to 3999 and video 4000 to 4999). Once
configured and applied to the OS, the Jabber for Windows client downloads its configuration from
Unified CM on registration and applies the SIP Profile media port range - common. From there, when a
Jabber for Windows client makes a call, it utilizes the media port ranges provided from Unified CM. The
GPO applied to the Windows OS, however, applies its policy to take the media traffic for audio over UDP
ports 3000 to 3999 and re-mark them to EF, and over UDP ports 4000 to 4999 and re-mark them to AF41.
As the traffic leaves the OS, the packets will contain the applied markings. It will be up to the network
to trust these markings and allow them to progress through the network. Figure 13-23 also illustrates a
similar GPO when using non-contiguous port ranges in the SIP Profile for media port range - separated
ports.
Classification in Mac OS
Cisco Jabber for Mac natively requests DSCP QoS marking to the operating system, which then marks
traffic without the need to configure any specific policies.
13-31
Chapter 13
Bandwidth Management
Classification in Android
Cisco Jabber for Android natively requests DSCP QoS marking to the operating system, which then
marks traffic without the need to configure any specific policies.
Use DSCP markings whenever possible because they apply to the IP layer end-to-end and are more
granular and more extensible than Layer 2 markings.
Mark as close to the endpoint as possible, preferably at the LAN switch level.
When deploying Jabber for voice and video in an environment where SCCP-based audio endpoints
are deployed, change the media port range of the Cisco Jabber endpoints to use a range outside of
16384 to 32767 (which is a hard-coded range for SCCP devices). This is to avoid any potential
overlap when creating network policies to re-mark DSCP based on the UDP port range. For example,
use ports 3000 to 3999 for voice-only (video disabled) Jabber clients and 3000 to 4999 for
video-enabled Cisco Jabber endpoints.
When trying to minimize the number of media ports used by the Cisco Jabber client, use a minimum
range of 100 ports. This is to ensure that there are enough ports for all of the streams, such as RTCP,
RTP for audio and video, BFCP, and RTP for secondary video for desktop sharing sessions, as well
as to avoid any overlap with other applications on the same computer.
When deploying Enhanced Locations CAC, over-provision the audio class (EF) to account for the
audio of video from Jabber clients that will be marked EF and not AF41.
Deploying QoS for Cisco Jabber clients can be achieved by mapping identifiable media and signaling
streams with Layer 4 port ranges to Layer 3 DSCP values. Mapping identifiable media and signaling
streams can be done for any Jabber client by using network access control lists (ACLs) or by using the
operating system and then allowing the PC, Mac, or hand-held devices QoS markings to pass through
the network by trusting the QoS markings. Combining both methods is not advisable because the
network ACL method will simply override the OS trust method and force the re-marking of all audio,
thus rendering useless the goal of using the trust method.
13-32
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
Note
This section discusses different Cisco IOS queuing and scheduling technologies that are covered in more
detail in the section on WAN Quality of Service (QoS), page 3-36. This section discusses some of these
technologies with the assumption that they are well understood technologies, and the discussion herein
focuses on the best practices and recommendations for using these various Cisco IOS queuing and
scheduling mechanisms.
Audio of IP Phone
EF
Audio of TelePresence
CS4
Video of TelePresence
CS4
AF41
AF41
PQ
CBWFQ
CBWFQ
other queues
EF
348858
Figure 13-24
Complex provisioning Requires managing multiple video queues and separating bandwidth
allocations for each type of video rather than for video as a whole
Sub-optimal bandwidth usage When video for one class is not using all of its bandwidth, the
remainder of the bandwidth becomes available to all of the other queues on the interface and not just
the other video queue. Thus, it is not optimal for two different classes of video to share the total
video bandwidth allocation effectively.
13-33
Chapter 13
Bandwidth Management
Other considerations of this approach with regard to the audio portion of a video call:
Audio of a video call can be impacted by packet loss in the video queue.
Same DSCP for audio and video streams of a video call
By default both audio and video of a video call are marked with the same DSCP value. As a
result, both audio and video streams are equally impacted during congestion of the video queue.
When video experiences packet loss, some video quality degradation can take place during the
time that it takes for the video endpoints to rate adapt down to an acceptable level until packet
loss is no longer experienced. Audio is a constant bit rate medium and does not have the same
abilities for rate adaptation as video does. Thus, for audio this degradation can mean that the
users are no longer able to communicate until the packet loss in the video queue is under control.
Impacting audio has a greater effect on user experience than does impacting video. When video
is impacted, users can still carry on a meeting or conversation while video is experiencing
packet loss. See the section on Audio versus Video, page 13-6, for more information on the
characteristics of both media.
Audio and video streams of a video call were traditionally marked with the same DSCP value
in order to ensure that there was not a large delay variance between the two streams, otherwise
video endpoints would not be able to sync audio and video correctly. With the implementation
of RTCP in all Cisco endpoints, this is no longer a concern because RTCP can ensure the proper
sync between audio and video of a video call. Of course, this requires RTCP to be enabled on
the video endpoints.
Audio stream classification for untrusted devices cannot be distinguished between voice-only calls
and video calls.
Media stream identification is difficult for untrusted endpoints and clients. As discussed earlier,
when the endpoint or client is not trusted, alternative methods for identification are required.
With alternative methods such as access lists, it is difficult if not impossible in most cases to
differentiate the audio of a voice-only call from the audio of a video call to classify those two
types of audio differently. Therefore, all audio from both types of calls would have to be marked
with a single DSCP value. This makes creating a holistic approach to uniform marking more
difficult.
13-34
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
Figure 13-25
EF
Audio of TelePresence
AF41
Video of TelePresence
AF41
AF42
AF42
Video
CBWFQ
other queues
PQ
EF
348859
Audio of IP Phone
In Figure 13-25 the audio of a voice call is marked as EF and placed into a Priority Queue (PQ) with a
strict policer on how much bandwidth the PQ can allocate to this traffic. Video calls are separated into
two classes, AF41 for TelePresence video and AF42 for Desktop video. Using a CBWFQ with Weighted
Random Early Detection (WRED), the administrator can adjust the drop precedence of AF42 over AF41,
thus ensuring that during times of congestion when the queue is filling up, AF42 packets are dropped
from the queue at a higher probability than AF41. See the section on WAN Quality of Service (QoS),
page 3-36, for more detail on the function of WRED.
This example illustrates how an administrator using a single CBWFQ with DSCP-based WRED for all
video can protect one type of video (TelePresence video) from packet loss over another type of video
(Desktop) during periods of congestion. With this "single video queue approach," unlike the dual video
queue approach, when one type of video is not using bandwidth in the queue, the other type of video
gains full access to the entire queue bandwidth if and when needed. This is a significant point when
looking to deploy pervasive video.
Audio of a video call can be impacted by packet loss in the video queue.
Audio stream classification for untrusted devices cannot be distinguished between voice-only calls
and video calls.
13-35
Chapter 13
Bandwidth Management
A strategy to address these deficiencies is to ensure that all audio is marked with a single value of
Expedited Forwarding (EF) across the solution. In this way, whether the audio stream is associated to a
voice-only call or a video call, it is always marked to the same single value. In this way, audio of a video
call will be prioritized above the video and not subject to any packet loss in the video queue. It also
solves the identification issue with untrusted devices such as Jabber clients. Because the marking of the
client is not trusted by the network access layer, there is no effective way of distinguishing the audio
stream of a voice-only call from the audio of a video call in the network. Thus, moving to this new model
where all audio is marked with the same single value simplifies the network prioritization and treatment
of the traffic.
Note
See the section on Trusted Endpoints, page 13-19, for information on how trusted endpoints acquire
DSCP and how to set the DSCP for the audio portion of a video or TelePresence endpoint, and for
information on which endpoints support this differentiation. Also, the section on Untrusted Endpoints
and Clients, page 13-23, shows how to set DSCP for Jabber clients.
Achieving this holistically across the entire solution depends on a number of conditions that are required
to achieve marking all audio to a DSCP of EF:
The endpoint must support the DSCP for Audio Portion of Video/TelePresence Call QoS setting
in Unified CM to be able to mark all audio as EF. See Table 13-7 for details on endpoint support.
Jabber clients can support marking all audio as EF in a trusted or untrusted implementation.
Enhanced Locations CAC can be implemented in conjunction with marking all audio as EF. ELCAC
relies on the correct DSCP setting to ensure protection of the queues that voice and video CAC pools
represent. Changing the DSCP of audio streams of the video calls requires updating how ELCAC
deducts bandwidth for video calls. This can be done by setting the service parameter under the Call
Admission Control section of the CallManager service, called Deduct Audio Bandwidth from
Audio Pool for Video Call. This parameter can be set to true or false:
True: Cisco Unified CM splits the audio and video bandwidth allocations for video calls into
separate pools. The bandwidth allocation for the audio portion of a video call is deducted from
the audio pool, while the video portion of a video call is deducted from the video pool.
False: Cisco Unified CM applies the legacy behavior, which is to deduct the audio and video
bandwidth allocations of a video call from the video pool. This is the default setting.
For more information on the admission control aspects of marking all audio of video to EF, see the
ELCAC section on Deducting all Audio from the Voice Pool, page 13-50.
Opportunistic Video
When attempting to deploy video pervasively across the organization, bandwidth constraints typically
determine the level of video resolution can be achieved during the busiest hour of the day based on the
bandwidth available and the number of video calls during that busy hour. To address this challenge, a
type of video can be targeted as opportunistic video using a single video queue with DSCP-based WRED
coupled with a strategy for identification and classification of collaboration media.
13-36
Chapter 13
Bandwidth Management
QoS Architecture for Collaboration
Opportunistic video means achieving the best video quality based on the WAN bandwidth resources
available at any given time. To achieve this, a number of requirements must be met:
Ensuring the WAN is configured with a single video queue using DSCP-based WRED with AF4
DSCP class servicing with drop precedence of AF41, AF42, and AF43 (only two DSCPs are
required)
Unified CM Regions
Unified CM locations (see Enhanced Locations Call Admission Control, page 13-39) work in
conjunction with regions to define the characteristics of a call flow. Regions define the type of
compression or bit rate (8 kbps or G.729, 64 kbps or G.722/G.711, and so forth) that is used between
any two devices. Location links define the amount of available bandwidth for the path between devices.
Each device and trunk in the system is assigned to both a region (by means of a device pool) and a
location (by means of a device pool or by direct configuration on the device itself):
Regions allow the bandwidth of video calls to be set. The audio limit on the region can result in
filtering out codecs with higher bit rates. However, for video calls, the video limit constrains the
quality (resolution and transmission rate) of the video.
Locations define the amount of total bandwidth available for all calls on that link. When a call is
made on a link, the regional value for that call must be subtracted from the total bandwidth allowed
for that link.
Building a region matrix to manage maximum video bit rate (video resolution) for groups of devices can
assist in ensuring that certain groups of devices do not over-saturate the network bandwidth. Some
guidelines for creating a region matrix include:
The smaller the number of groups, the easier it is to calculate bandwidth requirements.
Consider the default region settings to simplify the matrix and provide intra-region and inter-region
defaults.
For more about region settings, see the section on Enhanced Locations Call Admission Control,
page 13-39.
Table 13-8 shows an example of a maximum video bit rate region matrix for four groups of devices.
Note
Table 13-8 is only an example of how to group devices and what maximum bit rate might be suggested
for a general resolution between the groups of devices.
13-37
Chapter 13
Bandwidth Management
Table 13-8
Endpoint
Groupings
Legacy (Small
Screen)
Jabber
Room System +
Smart Desktop
Immersive + MCU
Legacy (Small
Screen)
800 kbps
800 kbps
800 kbps
800 kbps
Jabber
800 kbps
1,500 kbps
1,500 kbps
1,500 kbps
Room System +
Smart Desktop
800 kbps
1,500 kbps
2,500 kbps
2,500 kbps
Immersive + MCU
800 kbps
1,500 kbps
2,500 kbps
12,000 kbps
Legacy (Small Screen) These could be legacy endpoints with smaller low-resolution screens or
other devices to be capped at 800 kbps bit rate.
Jabber These would typically make up the largest group of deployed video-capable endpoints and
thus benefit from the opportunistic video approach. When classified as opportunistic video, they can
go up to a maximum of 1,500 kbps (720p@30fps) and would rate adapt downward based on packet
loss.
Room System + Smart Desktop These would be room systems such as Cisco MX, SX, C, or
Profile Series. Also, smart desktop endpoints would be Cisco DX and EX Series. At 2,500 kbps
maximum video bit rate, these endpoints would typically be capable of 720p@30fps
Immersive + MCU These would be the larger Cisco TX or IX Series endpoints as well as MCUs
set to a maximum of 12 Mbps, which roughly translates to 1080p@30fps with other TelePresence
devices and MCUs.
The first consideration is whether to have different intra-region settings versus inter-region settings.
This will determine whether per-site regions are required or not. The concept here is that if intraand inter-regional audio or video bit rates are to be different, then per-site regions will be required.
This augments the configuration of regions to the number of sites (N) multiplied by the number of
video groups (X): NX = number of regions required on average. If intra- and inter-region audio and
video bit rates will be the same, then only the regions for the video groups are required (X).
need to configure new regions. For example, if voice-only devices use the G.729 audio codec
over the WAN and G.711 or G.722 on the LAN, while video devices always use G7.11 or G.722,
then the voice-only and video endpoints cannot share a region. Thus, each site would require a
region per group of devices. Sites = N, and video region groups = 4 + voice-only region group;
then N*4 is the number of regions required. Use the Prime Collaboration Provisioning tool or
the Bulk Administration tool as configuration aids.
Per-site regions might not be needed if a single audio codec is used for both intra-region and
inter-region calls as well as voice-only calls. If both audio and video endpoints use G.711 or
G.722 over the WAN and LAN for voice-only or video calls, then voice-only IP phones and
video endpoints could use the same region.
13-38
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Consider the default region settings to simplify the matrix. The following example illustrates
possible default settings based on the region groupings in Figure 13-25. If it is desired to have the
intra-region bit rate be larger than the inter-region bit rate, then per-site regions are required.
Default Intraregion Max Video Call Bit Rate (Includes Audio): Set to 768, sets the maximum
video bit rate capability of devices for calls within a region to 768 kbps.
Default Interregion Max Video Call Bit Rate (Includes Audio): Set to 768, sets the maximum
video bit rate capability of devices for calls between regions to 768 kbps.
Default Intraregion Max Immersive Video Call Bit Rate (Includes Audio): Set to 12000, sets the
maximum video bit rate capability of devices for calls within a region to 12,000 kbps.
Default Interregion Max Immersive Video Call Bit Rate (Includes Audio): Set to 12000, sets the
maximum video bit rate capability of devices for calls between regions to 12,000 kbps.
In addition to the defaults, 4 regions should be set up, one for each group of video endpoints.
Circuit-Switched
Networks
Packet-Switched
Networks
PSTN
PSTN
Physical
Trunks
PBX
Third call
rejected
IP WAN
Link
STOP
V
Router/
Gateway
IP
Cisco Unified CM
IP
IP
126670
Figure 13-26
As shown on the left side of Figure 13-26, traditional TDM-based PBXs operate within circuit-switched
networks, where a circuit is established each time a call is set up. As a consequence, when a legacy PBX
is connected to the PSTN or to another PBX, a certain number of physical trunks must be provisioned.
When calls have to be set up to the PSTN or to another PBX, the PBX selects a trunk from those that are
available. If no trunks are available, the call is rejected by the PBX and the caller hears a network-busy
signal.
13-39
Chapter 13
Bandwidth Management
Now consider the IP connected Unified Communications system shown on the right side of
Figure 13-26. Because it is based on a packet-switched network (the IP network), no circuits are
established to set up an IP telephony call. Instead, the IP packets containing the voice samples are simply
routed across the IP network together with other types of data packets. Quality of Service (QoS) is used
to differentiate the voice packets from the data packets, but bandwidth resources, especially on IP WAN
links, are not infinite. Therefore, network administrators dedicate a certain amount of "priority"
bandwidth to voice traffic on each IP WAN link. However, once the provisioned bandwidth has been
fully utilized, the IP telephony system must reject subsequent calls to avoid oversubscription of the
priority queue on the IP WAN link, which would cause quality degradation for all voice calls. This
function is known as call admission control, and it is essential to guarantee good voice and video quality
in a multisite deployment involving an IP WAN.
To preserve a satisfactory end-user experience, the call admission control function should always be
performed during the call setup phase so that, if network resources are not available, a message can be
presented to the end-user or the call can be rerouted across a different network (such as the PSTN).
This chapter discusses the following main topics:
13-40
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
There are also tools to administer and troubleshoot Enhanced Location CAC. The CAC enhancements
and design are discussed in detail in this chapter, but the troubleshooting and serviceability tools are
discussed in separate product documentation.
Locations A Location represents a LAN. It could contain endpoints or simply serve as a transit
location between links for WAN network modeling. For example, an MPLS provider could be
represented by a Location.
Links Links interconnect locations and are used to define bandwidth available between locations.
Links logically represent the WAN link and are configured in the Location user interface (UI).
Weights A weight provides the relative priority of a link in forming the effective path between
any pair of locations. The effective path is the path used by Unified CM for the bandwidth
calculations, and it has the least cumulative weight of all possible paths. Weights are used on links
to provide a "cost" for the "effective path" and are pertinent only when there is more than one path
between any two locations.
Path A path is a sequence of links and intermediate locations connecting a pair of locations.
Unified CM calculates least-cost paths (lowest cumulative weight) from each location to all other
locations and builds a map of the various paths. Only one "effective path" is used between any pair
of locations.
Effective Path The effective path is the path with the least cumulative weight.
Bandwidth Allocation The amount of bandwidth allocated in the model for each type of traffic:
audio, video, and immersive video (TelePresence).
Location Bandwidth Manager (LBM) The active service in Unified CM that assembles a network
model from configured location and link data in one or more clusters, determines the effective paths
between pairs of locations, determines whether to admit calls between a pair of locations based on
the availability of bandwidth for each type of call, and deducts (reserves) bandwidth for the duration
of each call that is admitted.
Location Bandwidth Manager Hub A Location Bandwidth Manager (LBM) service that has been
designated to participate directly in intercluster replication of fixed locations, links data, and
dynamic bandwidth allocation data. LBMs assigned to an LBM hub group discover each other
13-41
Chapter 13
Bandwidth Management
through their common connections and form a fully-meshed intercluster replication network. Other
LBM services in a cluster with an LBM hub participate indirectly in intercluster replication through
the LBM hubs in their cluster.
13-42
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-27
Location
Intra-Location
Audio
Video
Immersive
San Jose
Unlimited
Unlimited
Unlimited
SIP
SIP
Link
San Jose
Intra-Location
Audio
Video
Immersive
Richardson
Unlimited
Unlimited
Unlimited
SIP
SIP
Boulder
Boulder
Unlimited
Unlimited
Unlimited
Deduct Bandwidth!
SIP
SIP
Richardson
Deduct Bandwidth!
Intra-Location
Audio
Video
Immersive
RTP
Unlimited
Unlimited
Unlimited
SIP
SIP
RTP
292542
Effective Path
When a call is made between San Jose and RTP, Unified CM calculates the bandwidth of the requested
call, which is determined by the region pair between the two devices (see Locations, Links, and Region
Settings, page 13-46) and verifies the effective path between the two locations. That is to say,
Unified CM verifies the locations and links that make up the path between the two locations and
accordingly deducts bandwidth from each link and (if applicable) from each location in the path. The
intra-location bandwidth also is deducted along the path if any of the locations has configured a
bandwidth value other than unlimited.
Weight is configurable on the link only and provides the ability to force a specific path choice when
multiple paths between two locations are available. When multiple paths are configured, only one will
be selected based on the cumulative weight, and this path is referred to as the effective path. This weight
is static and the effective path does not change dynamically. Figure 13-28 illustrates weight configured
on links between three locations: San Jose, Boulder, and Seattle.
13-43
Chapter 13
Bandwidth Management
Figure 13-28
San Jose
IP
LBM
Effective Path
IP
IP
LBM
Link Weight
50
30
30
IP
IP
IP
IP
IP
Boulder
Seattle
292543
IP
San Jose to Seattle has two paths, one direct link between the locations and another path through the
Boulder location (link San Jose/Boulder and link Boulder/Seattle). The weight configured on the direct
link between San Jose and Seattle is 50 and is less than the cumulative weight of links San Jose/Boulder
and Boulder/Seattle which is 60 (30+30). Thus, the direct link is chosen as the effective path because the
cumulative link weight is 50.
When you configure a device in Unified CM, the device can be assigned to a location. A location can be
configured with links to other locations in order to build a topology. The locations configured in
Unified CM are virtual locations and not real, physical locations. As mentioned, Unified CM has no
knowledge of the actual physical topology of the network. Therefore, any changes to the physical
network must be made manually in Unified CM to map the real underlying network topology with the
Unified CM locations model. If a device is moved from one physical location to another, the system
administrator must either perform a manual update on its location configuration or else implement the
device mobility feature so that Unified CM can correctly calculate bandwidth allocations for calls to and
from that device. Each device is in location Hub_None by default. Location Hub_None is an example
location that typically serves as a hub linking two or more locations, and it is configured by default with
unlimited intra-location bandwidth allocations for audio, video, and immersive bandwidth.
Unified CM allows you to define separate voice, video, and immersive video bandwidth pools for each
location and link between locations. Typically the locations intra-location bandwidth configuration is
left as a default of Unlimited while the link between locations is set to a finite number of kilobits per
second (kbps) to match the capacity of a WAN links between physical sites. If the location's
intra-location audio, video, and immersive bandwidths are configured as Unlimited, there will be
unlimited bandwidth available for all calls (audio, video, and immersive) within that location and
13-44
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
transiting that location. On the other hand, if the bandwidth values are set to a finite number of kilobits
per second (kbps), Unified CM will track all calls within the location and all calls that use the location
as a transit location (a location that is in the calculation path but is not the originating or terminating
location in the path).
For video calls, the video location bandwidth takes into account both the audio and the video portions
of the video call. Therefore, for a video call, no bandwidth is deducted from the audio bandwidth pool.
The same applies to immersive video calls.
The devices that can specify membership in a location include:
IP phones
CTI ports
H.323 clients
Conference bridges
Gateways
Trunks
The Enhanced Location Call Admission Control mechanism also takes into account the mid-call changes
in call type. For example, if an inter-site video call is established, Unified CM will subtract the
appropriate amount of video bandwidth from the respective locations and links in the path. If this video
call changes to an audio-only call as the result of a transfer to a device that is not capable of video,
Unified CM will return the allocated bandwidth to the video pool and allocate the appropriate amount
of bandwidth from the audio pool along the same path. Calls that change from audio to video will cause
the opposite change of bandwidth allocation.
Table 13-9 lists the amount of bandwidth requested by the static locations algorithm for various call
speeds. For an audio call, Unified CM counts the media bit rates plus the IP and UDP overhead. For
example, a G.711 audio call consumes 80 kbps (64k bit rate + 16k IP/UDP headers) deducted from the
location's and link's audio bandwidth allocation. For a video call, Unified CM counts only the media bit
rates for both the audio and video streams. For example, for a video call at a bit rate of 384 kbps,
Unified CM will allocate 384 kbps from the video bandwidth allocation.
Table 13-9
Call Speed
80 kbps
24 kbps
128 kbps
384 kbps
512 kbps
768 kbps
13-45
Chapter 13
Bandwidth Management
For a complete list of codecs and location and link bandwidth values, refer to the bandwidth calculations
information in the Call Admission Control section of the Cisco Unified Communications Manager
System Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
For example, assume that the link configuration for the location Branch 1 to Hub_None allocates
256 kbps of available audio bandwidth and 384 kbps of available video bandwidth. In this case the path
from Branch 1 to Hub_None can support up to three G.711 audio calls (at 80 kbps per call) or ten G.729
audio calls (at 24 kbps per call), or any combination of both that does not exceed 256 kbps. The link
between locations can also support different numbers of video calls depending on the video and audio
codecs being used (for example, one video call requesting 384 kbps of bandwidth or three video calls
with each requesting 128 kbps of bandwidth).
When a call is placed from one location to the other, Unified CM deducts the appropriate amount of
bandwidth from the effective path of locations and links from one location to another. Using
Figure 13-27 as an example, a G.729 call between San Jose and RTP locations causes Unified CM to
deduct 24 kbps from the available bandwidth at the links between San Jose and Boulder, between
Boulder and Richardson, and between Richardson and RTP. When the call has completed, Unified CM
returns the bandwidth to those same links over the effective path. If there is not enough bandwidth at any
one of the links over the path, the call is denied by Unified CM and the caller receives the network busy
tone. If the calling device is an IP phone with a display, that device also displays the message "Not
Enough Bandwidth."
When an inter-location call is denied by call admission control, Unified CM can automatically reroute
the call to the destination through the PSTN connection by means of the Automated Alternate Routing
(AAR) feature. For detailed information on the AAR feature, see Automated Alternate Routing,
page 14-78.
Note
AAR is invoked only when Enhanced Location Call Admission Control denies the call due to a lack of
network bandwidth along the effective path. AAR is not invoked when the IP WAN is unavailable or
other connectivity issues cause the called device to become unregistered with Unified CM. In such cases,
the calls are redirected to the target specified in the Call Forward No Answer field of the called device.
It is also worth noting that video devices can be enabled to Retry Video Call as Audio if a video call
between devices fails CAC. This option is configured on the video endpoint configuration page in
Unified CM and is applicable to video endpoints or trunks receiving calls. It should also be noted that
for some video endpoints Retry Video Call as Audio is enabled by default and not configurable on the
endpoint.
13-46
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Physical sites (for example, a branch office) or transit sites (for example, an MPLS cloud) A
location represents a LAN. It could contain endpoints or simply serve as a transit location between
links for WAN network modeling.
Link bandwidth between adjacent locations Links interconnect locations and are used to define
bandwidth available between locations. Links logically represent the WAN link between physical
sites.
Audio Bandwidth The amount of bandwidth that is available in the WAN link for voice and
fax calls being made from devices in the location to the configured adjacent location.
Unified CM uses this bandwidth value for Enhanced Location Call Admission Control.
Video Bandwidth The amount of video bandwidth that is available in the WAN link for video
calls being made from devices in the location to the configured adjacent location. Unified CM
uses this bandwidth value for Enhanced Location Call Admission Control.
Immersive Video Bandwidth The amount of immersive bandwidth that is available in the
WAN link for TelePresence calls being made from devices in the location to the configured
adjacent location. Unified CM uses this bandwidth value for Enhanced Location Call
Admission Control.
Intra-location bandwidth
Audio Bandwidth The amount of bandwidth that is available in the LAN for voice and fax
calls being made from devices within the location. Unified CM uses this bandwidth value for
Enhanced Location Call Admission Control.
Video Bandwidth The amount of video bandwidth that is available in the LAN for video calls
being made from devices within the location. Unified CM uses this bandwidth value for
Enhanced Location Call Admission Control.
Immersive Video Bandwidth The amount of immersive bandwidth that is available in the
LAN for TelePresence calls being made from devices within the location. Unified CM uses this
bandwidth value for Enhanced Location Call Admission Control.
You can configure regions in Unified CM to define:
The Maximum Audio Bit Rate used for intraregion and interregion calls
The Maximum Session Bit Rate for Video Calls (Includes Audio) used for intraregion and
interregion calls
The Maximum Session Bit Rate for Immersive Video Calls (Includes Audio) used for intraregion
and interregion calls
13-47
Chapter 13
Bandwidth Management
Default Interregion Max Immersive Video Call Bit Rate (Includes Audio)
When adding regions, you should select Use System Default for the Maximum Audio Bit Rate and
Maximum Session Bit Rate for Video Call values.
Changing these values for individual regions and locations from the default has an impact on server
initialization and publisher upgrade times. Hence, with a total of 2,000 regions and 2,000 locations, you
can modify up to 200 of them to use non-default values. With a total of 1,000 or fewer regions and
locations, you can modify up to 500 of them to use non-default values. Table 13-10 summarizes these
limits.
Table 13-10
Note
0 to 200
2,000
2,000
200 to 500
1,000
1,000
The Maximum Audio Bit Rate is used by both voice calls and fax calls. If you plan to use G.729 as the
interregion codec, use T.38 Fax Relay for fax calls. If you plan to use fax pass-through over the WAN,
use audio preference lists to prefer G.729 for audio-only calls and G.711 for fax calls.
Services bandwidth requests from the Cisco CallManager service (Unified CM call control)
The LBM Service is enabled by default when upgrading Cisco Unified CM from earlier releases that
support only traditional Location CAC. For new installations, the LBM service must be activated
manually.
13-48
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
During initialization, the LBM reads local locations information from the database, such as: locations
audio, video, and immersive bandwidth values; intra-location bandwidth data; and location-to-location
link audio, video, and immersive bandwidth values and weight. Using the link data, each LBM in a
cluster creates a local assembly of the paths from one location to every other location. This is referred
to as the assembled topology. In a cluster, each LBM accesses the same data and thus creates the same
local copy of the assembled topology during initialization.
At runtime, the LBM applies reservations along the computed paths in the local assembled topology of
locations and links, and it replicates the reservations to other LBMs in the cluster. If intercluster
Enhanced Location CAC is configured and activated, the LBM can be configured to replicate the
assembled topology to other clusters (see Intercluster Enhanced Location CAC, page 13-51, for more
details).
By default the Cisco CallManager service communicates with the local LBM service; however, LBM
groups can be used to manage this communication. LBM groups provide an active and standby LBM in
order to create redundancy for Unified CM call control. Figure 13-29 illustrates LBM redundancy.
Figure 13-29
UCM2
LBM
Replication
D
ILE
FALBM
UCM
UCM3
UCM
UCM4
LBM Group 1
UCM1
UCM2
LBM
UCM
UCM5
292544
UCM1
Figure 13-29 shows five Unified CM servers: UCM1 and UCM2 are dedicated LBM servers (only LBM
service enabled); UCM3, UCM4, and UCM5 are Unified CM subscribers (Cisco CallManager service
enabled). An LBM Group has been configured with UCM1 as active and UCM2 as standby, and it is
applied to subscribers UCM3, UCM4, and UCM5. This configuration allows for UCM3, UCM4, and
UCM5 to query UCM1 for all bandwidth requests. If UCM1 fails for any reason, the subscribers will
fail-over to the standby UCM2. This example is used to illustrate how the LBM Group configuration
functions and is not a recommended configuration (see recommendations below).
Because LBM is directly involved in processing requests for every call that is processed by the
CallManager service that it is serving, it is important to follow these simple design recommendations in
order to ensure optimal functionality of the LBM.
The recommended configuration is to run LBM co-resident with the Cisco CallManager service (call
processing). If redundancy of the LBM service is required, it is important not to over-subscribe any given
LBM. It is also important to make sure that LBM is no more than a primary and secondary in any given
deployment. This means that LBM should not have the load of more than 2 CallManager services during
failure scenarios, and the load of only one CallManager service during normal operation. The LBM
group can be used to configure a co-resident LBM as the primary, another local (on the same LAN) LBM
13-49
Chapter 13
Bandwidth Management
as secondary, and lastly the service parameter as a failsafe mechanism to ensure that all calls processed
by that CallManager service do not fail. There are many reasons for these recommendations. It is
difficult, at best, to determine the load of any LBM because it is directly related to the call-processing
load of the CallManager service that it is serving. There are also considerations for delay. As soon as an
LBM is off-box from a CallManager service, there is an added delay caused by packetization and
processing for every call serviced by the CallManager service. Compounding call-processing delay can
bring the overall delay budget to an unacceptable level for any given call flow to a ringing state, and
result in a poor user experience. Following these design recommendations will reduce the overall
call-processing delay.
The order in which the Unified CM Cisco CallManager service uses the LBM is as follows:
Service parameter Call Treatment when no LBM available (Default = allow calls)
Note
The Location Bandwidth Manager (LBM) is a Unified CM Feature Service. It is responsible for
modeling the topology and servicing Unified CM bandwidth requests.
LBMs within the cluster create a fully meshed communications network via XML over TCP for the
replication of bandwidth change notifications between LBMs.
Cisco recommends deploying the LBM service co-resident with a Unified CM subscriber running
the Cisco CallManager call processing service.
If redundancy is required for the LBM service, create an LBM Group for each Unified CM
subscriber running the Cisco CallManager call processing service. Add the co-resident LBM service
as the primary LBM, and the LBM from another Unified CM subscriber on the same LAN as a
secondary LBM. This will ensure that the Cisco CallManager call processing service uses the
co-resident LBM as primary, the LBM on another local (same LAN) Unified CM subscriber as
secondary, and the service parameter Call Treatment when no LBM available as tertiary source
for LBM requests.
Cisco recommends having LBM back up more than one Cisco CallManager service. Assuming that the
LBM is handling the load of the co-resident CallManager service, and during failure of another
CallManager service, this would equate to the load of 2 call processing servers on a single LBM.
For deployments with cluster over the WAN and local failover, intracluster LBM traffic is already
calculated into the WAN bandwidth calculations. See the section on clustering over the WAN Local
Failover Deployment Model, page 10-45, for details on bandwidth calculations.
Unified CM now has a feature that allows the administrator to deduct the audio bandwidth of video and
TelePresence calls from the voice pool. Because ELCAC relies on the correct DSCP setting in order to
ensure the protection of the queues that voice and video CAC pools represent, changing how Unified CM
deducts bandwidth from the video pool requires the DSCP of audio streams of the video calls to be
marked the same as the audio streams of audio-only calls. See the section on Considerations for Audio
of Video Calls, page 13-35, for information about aligning admission control with QoS.
13-50
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
In Unified CM this feature is enabled by setting the service parameter Deduct Audio Bandwidth from
Audio Pool for Video Call to True under the Call Admission Control section of the CallManager
service. False is the default setting, and by default Unified CM deducts both audio and video streams of
video calls from the video pool.
Global Topology
Hub-None
Locations
Links
LBM inter-cluster
replication network
HUB
Cluster 1
Hub-None
Loc_11
Loc_12
Cluster 1
Local Topology
LBM
HUB
Cluster 2
Hub-None
292545
LBM
Figure 13-30 shows two clusters, Cluster 1 and Cluster 2, each with a locally configured hub-and-spoke
network topology. Cluster 1 has configured Hub_None with links to Loc_11 and Loc_12, while Cluster 2
has configured Hub_None with links to Loc_21, Loc_22, and Loc_23. Upon enabling intercluster
Enhanced Location CAC, Cluster 1 sends its local topology to Cluster 2, as does Cluster 2 to Cluster 1.
After each cluster obtains a copy of the remote clusters topology, each cluster overlays the remote
clusters topology over their own. The overlay is accomplished through common locations, which are
13-51
Chapter 13
Bandwidth Management
locations that are configured with the same name. Because both Cluster 1 and Cluster 2 have the
common location Hub_None with the same name, each cluster will overlay the other's network topology
with Hub_None as a common location, thus creating a global topology where Hub_None is the hub and
Loc_11, Loc_12, Loc_21, Loc_22 and Loc_23 are all spoke locations. This is an example of a simple
network topology, but more complex topologies would be processed in the same way.
13-52
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-31
LBM Hub
LBM Hubs
LBM Spoke
SME_1
LBM
SME_2
Bootstrap
LBM
SME_1
SME_2
Member3
HUB
HUB
Leaf Cluster 1
Leaf Cluster 2
LBM Hubs
UCM_3
SME_1
SME_2
UCM_1
UCM_A
UCM_A
SME_2
UCM_1
UCM_C
LBM
HUB
HUB
LBM
LBM
HUB
HUB
LBM
UCM_4
UCM_2
UCM_B
UCM_D
348704
SME_1
Bootstrap
LBM
In Figure 13-31, two LBMs from each cluster have been designated as the LBM hubs for their cluster.
These LBM hubs form the intercluster LBM replication network. The bootstrap LBMs configured in
each LBM intercluster replication group are designated as SME_1 and SME_2. These two LBM hubs
from the SME cluster serve as points of contact or bootstrap LBMs for the entire intercluster LBM
replication network. This means that each LBM in each cluster connects to SME_1, replicates its local
topology to SME_1, and gets the remote topology from SME_1. They also get the connectivity
information for the other leaf clusters from SME_1, connect to the other remote clusters, and replicate
their topologies. This creates the full-mesh replication network. If SME_1 is unavailable, the LBM hubs
will connect to SME_2. If SME_2 is unavailable, Leaf Cluster 1 LBMs will connect to UCM_A and Leaf
Cluster 2 LBMs will connect to UCM_1 as a backup measure in case the SME cluster is unavailable.
This is just an example configuration to illustrate the components of the intercluster LBM replication
network.
The LBM has the following roles with respect to the LBM intercluster replication network:
Bootstrap LBMs
Remote LBM hubs responsible for interconnecting all LBM hubs in the replication network
Can be any hub in the network
Can indicate up to three bootstrap LBM hubs per LBM intercluster replication group
13-53
Chapter 13
Bandwidth Management
13-54
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-32
Cluster 3
Cluster 3
Topology
Cluster 1 Locations
Cluster 2 Locations
DC
Cluster 3 Locations
Common Locations
Links
LBM
Loc_31
HUB
Cluster 1
Cluster 2
Global Topology
DC
Loc_31
HUB
Regional 1
Cluster 1
Topology
LBM
DC
DC
Loc_11 Loc_12 Loc_21 Loc_22 Loc_24
Regional 1
Loc_11 Loc_12
HUB
Regional 2
Cluster 2
Topology
Regional 2
292547
LBM
In Figure 13-32, Cluster 1 has devices in locations Regional 1, Loc_11, and Loc_12, but it requires
configuring DC and a link from Regional 1 to DC in order to link to the rest of the global topology.
Cluster 2 is similar, with devices in Regional 2 and Loc_21, Loc_22, and Loc_23, and it requires
configuring DC and a link from DC to Regional 2 to map into the global topology. Cluster 3 has devices
in Loc_31 only, and it requires configuring DC and a link to DC from Loc_31 to map into Cluster 1 and
Cluster 2 topologies. Alternatively, Regional 1 and Regional 2 could be the common locations
configured on all clusters instead of DC, as is illustrated in Figure 13-33.
13-55
Chapter 13
Bandwidth Management
Figure 13-33
Cluster 3
Topology
Cluster 3
DC
Cluster 1 Locations
Loc_31
Regional 1
Cluster 2 Locations
LBM
Cluster 3 Locations
Common Locations
Links
Regional 2
HUB
Cluster 1
Cluster 2
Global Topology
DC
LBM
LBM
Loc_31
HUB
Regional 1
Cluster 2
Topology
Loc_11 Loc_12 Loc_21 Loc_22 Loc_24
Loc_11 Loc_12
Regional 2
292548
Cluster 1
Topology
Regional 1
HUB
Regional 2
The key to topology mapping from cluster to cluster is to ensure that at least one cluster has a common
location with another cluster so that the topologies interconnect accordingly.
Shadow Location
The shadow location is used to enable a SIP trunk to pass Enhanced Location CAC information such as
location name and Video-Traffic-Class (discussed below), among other things, required for Enhanced
Location CAC to function between clusters. In order to pass this location information across clusters,
the SIP intercluster trunk (ICT) must be assigned to the "shadow" location. The shadow location cannot
have a link to other locations, and therefore no bandwidth can be reserved between the shadow location
and other locations. Any device other than a SIP ICT that is assigned to the shadow location will be
treated as if it was associated to Hub_None. That is important to know because if a device other than a
SIP ICT ends up in the shadow location, bandwidth deductions will be made from that device as if it
were in Hub_None, and that could have varying effects depending on the location and links
configuration.
When the SIP ICT is enabled for Enhanced Location CAC, it passes information in the SIP Call-Info
header that allows the originating and terminating clusters to process the location bandwidth deductions
end-to-end. Figure 13-34 illustrates an example of a call between two clusters and some details about
the information passed. This is only to illustrate how location information is passed from cluster to
cluster and how bandwidth deductions are made.
13-56
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-34
Hub-none
Loc_11 Loc_12
Loc_11 Loc_12
Loc_21
Loc_22 Loc_24
Loc_21
LBM
Loc_22 Loc_24
LBM
HUB
HUB
Cluster 1
Cluster 2
LOCATION
INVITE
LOCATION
SIP Call-Info Header
Call-Info: <urn:x-cisco-remotecc:callinfo>;x-ciscoloc-id=4f3fc6c3-f902-eb3c-6fb8-b177eb03192a;xcisco-loc-name=LOC_25;x-cisco-fateshareid=AFL1:72851723 x-cisco-video-trafficclass=desktop
292549
In Figure 13-34, Cluster 1 sends an invite to Cluster 2 and populates the call-info header with the calling
parties location name and Video-Traffic-Class, among other pertinent information such as unique
call-ID. When Cluster 2 receives the invite with the information, it looks up the terminating party and
performs a CAC request on the path between the calling partys and called partys locations from the
global topology that it has in memory from LBM replication. If it is successful, Cluster 2 will replicate
the reservation and extend the call to the terminating device and return a 180 ringing with the location
information of the called party back to Cluster 1. When Cluster 1 receives the 180 ringing, it gets the
terminating devices location name and goes through the same bandwidth lookup process using the same
unique call-ID that it calculates from the information passed in the call-info header. If it is successful, it
too continues with the call flow. Because both clusters use the same information in the call-info header,
they will deduct bandwidth for the same call using the same call-ID, thus avoiding any double bandwidth
deductions.
13-57
Chapter 13
Bandwidth Management
locations, links, and bandwidth values; and once those values are replicated, all clusters use those values
because they are the most restrictive. This design alleviates configuration overhead in deployments
where a large number of common locations are required across multiple clusters.
Recommendations
and links).
The management cluster should be configured with the following:
to devices but should not configure the links between locations. This link information will come
from the management cluster when intercluster Enhanced Location CAC is enabled.
When intercluster Enhanced Location CAC is enabled, all of the locations and links will be
replicated from the management cluster and LBM will use the lowest, most restrictive
bandwidth and weight value.
LBM will always use the lowest most restrictive bandwidth and lowest weight value after
replication.
Benefits
Alleviates location and link configuration overhead when clusters share a large number of common
locations.
Other clusters in the enterprise require the configuration only of locations needed for
location-to-device and endpoint association.
Figure 13-35 illustrates Cisco Unified Communications Manager Session Management Edition (SME)
as a Location and Link Management Cluster for three leaf clusters.
Note
As mentioned, any cluster can act as the Location and Link management cluster. In this example SME
is the Location and Link management cluster.
13-58
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-35
SME Topology
DC
SME
Loc_31
Regional 1
Leaf 3
Regional 2
LBM
Leaf 3
Configured
Topology
Loc_31
LBM
HUB
HUB
Leaf 1
Leaf 2
LBM
LBM
Loc_31
HUB
Regional 1
HUB
Regional 2
Leaf 2
Configured Topology
Leaf 1
Configured Topology
Leaf 1
Regional 2
Leaf 1 Locations
Leaf 3 Locations
Transit Locations
Links
348705
Leaf 2 Locations
Loc_11 Loc_12
In Figure 13-35 there are three leaf clusters, each with devices in only a regional and remote locations.
SME has the entire global topology configured with locations and links, and intercluster LBM
replication is enabled between all four clusters. None of the clusters in this example share locations,
although all of the locations are common locations because SME has configured the entire location and
link topology. Note that Leaf 1, Leaf 2, and Leaf 3 configure only locations that they require to associate
to devices and endpoints, while SME has the entire global topology configured. After intercluster
replication, all clusters will have the global topology.
Intercluster Enhanced Location CAC Design and Deployment Recommendations and Considerations
Each cluster should be configured with the immediately neighboring locations so that each clusters
topology can inter-connect. This does not apply to Location and Link Management Cluster
deployments.
Links need to be configured to establish points of interconnect between remote topologies. This does
not apply to Location and Link Management Cluster deployments.
13-59
Chapter 13
Bandwidth Management
Discrepancies of bandwidth limits and weights on common locations and links are resolved by using
the lowest bandwidth and weight values.
Naming locations consistently across clusters is critical. Follow the practice, "Same location, same
name; different location, different name."
The Hub_None location should be renamed to be unique in each cluster or else it will be a common
(shared) location by other clusters.
An LBM spoke does not directly communicate with other remote clusters. LBM spokes receive and
send messages to remote clusters through the Local LBM Hub.
If a cluster has no LBM hub, or if the LBM hub is not running, the cluster will be isolated and will
not participate in the intercluster LBM replication network.
Performance Guidelines
Maximum of 2,000 locally configured locations. This limit of 2,000 locations also applies to the
Location and Link Management Cluster.
13-60
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
TelePresence immersive endpoints mark their media with a DSCP value of CS4 by default, and desktop
video endpoints mark their media with AF41 by default, as per recommended QoS settings. For Cisco
endpoints this is accomplished through the configurable Unified CM QoS service parameters DSCP for
Video calls and DSCP for TelePresence calls. When a Cisco TelePresence endpoint registers with
Unified CM, it downloads a configuration file and applies the QoS setting of DSCP for TelePresence
calls. When a Unified Communications video-capable endpoint registers with Unified CM, it downloads
a configuration file and applies the QoS setting of DSCP for Video calls. All third-party video endpoints
require manual configuration of the endpoints themselves and are statically configured, meaning they do
not change QoS marking depending on the call type; therefore, it is important to match the Enhanced
Location CAC bandwidth allocation to the correct DSCP. Unified CM achieves this by deducting
desktop video calls from the Video Bandwidth location and link allocation for devices that have a Video
Call Traffic Class of desktop. End-to-end TelePresence immersive video calls are deducted from the
Immersive Video Bandwidth location and link allocation for devices or trunks with the Video Call Traffic
Class of immersive. This ensures that end-to-end desktop video and immersive video calls are marked
correctly and counted correctly for call admission control. For calls between desktop devices and
TelePresence immersive devices, bandwidth is deducted from both the Video Bandwidth and the
Immersive Video Bandwidth location and link allocations.
Endpoint Classification
Cisco TelePresence endpoints have a fixed non-configurable Video Call Traffic Class of immersive and
are identified by Unified CM as immersive. Telepresence endpoints are defined in Unified CM by the
device type. When a device is added in Unified CM, any device with TelePresence in the name of the
device type is classified as immersive, as are the generic single-screen and multi-screen room systems.
Another way to check the capabilities of the endpoints in the Unified CM is to go to the Cisco Unified
Reporting Tool > System Reports > Unified CM Phone Feature List. In the feature drop down list,
select Immersive Video Support for TelePresence Devices; in the product drop down list, select All.
This will display all of the device types that are classified as immersive. All other endpoints have a fixed
Video Call Traffic Class of desktop due to their lack of the non-configurable immersive attribute.
Bandwidth reservations are determined by the classification of endpoints in a video call, and they deduct
bandwidth from the locations and links bandwidth pools as listed in Table 13-11.
Table 13-11
Endpoint A
Endpoint B
Immersive video
Immersive video
Immersive bandwidth
Immersive video
Desktop video
Desktop video
Desktop video
Video bandwidth
Audio-only call
Any
Audio bandwidth
13-61
Chapter 13
Bandwidth Management
A SIP trunk can be classified with any of these three classifications and is used primarily to classify
Video or TelePresence Multipoint Control Units (MCUs), a video device at a fixed location, a
Unified CM cluster supporting traditional Location CAC, or a Cisco TelePresence System Video
Communications Server (VCS).
Bandwidth reservations are determined by the classification of an endpoint and a SIP trunk in a video
call, and they deduct bandwidth from the locations and links bandwidth pools as listed in Table 13-12.
Table 13-12
Endpoint
SIP Trunk
TelePresence endpoint
Immersive
Immersive bandwidth
TelePresence endpoint
Desktop
TelePresence endpoint
Mixed
Desktop endpoint
Immersive
Desktop endpoint
Desktop
Video bandwidth
Desktop endpoint
Mixed
Non-video endpoint
Any
Audio bandwidth
By default, all video calls from either immersive or desktop endpoints are deducted from the locations
and links video bandwidth pool. To change this behavior, set Unified CMs CallManager service
parameter Use Video BandwidthPool for Immersive Video Calls to False, and this will enable the
immersive video bandwidth deductions. After this is enabled, immersive and desktop video calls will be
deducted out of their respective pools.
As described earlier, a video call between a Unified Communications video endpoint (desktop Video
Call Traffic Class) and a TelePresence endpoint (immersive Video Call Traffic Class) will mark their
media asymmetrically and, when immersive video CAC is enabled, will deduct bandwidth from both
video and immersive locations and links bandwidth pools. Figure 13-36 illustrates this.
13-62
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-36
Enhanced Location CAC Bandwidth Deductions and Media Marking for a Multi-Site Deployment
Service Param:
Use Video BandwidthPool
for Immersive Video Calls*
= False
(Default = True)
Cisco Unified CM
IP
P/S
SIP
SC
Location A
Location B
IP WAN
TelePresence
UC Video
Phone
Link Location A to B
Configured
Used
Audio
Unlimited
Video
Unlimited
N/A
Immersive
Unlimited
N/A
Configured
Audio
800k
Video
1536k
Immersive
5000k
Configured
Used
Used
Audio
Unlimited
768k
Video
Unlimited
N/A
768k
Immersive
Unlimited
N/A
292551
SIP
Examples of Various Call Flows and Location and Link Bandwidth Pool
Deductions
The following call flows depict the expected behavior of locations and links bandwidth deductions when
the Unified CM service parameter Use Video BandwidthPool for Immersive Video Calls is set to
False.
Figure 13-37 illustrates an end-to-end TelePresence immersive video call between TP-A in Location L1
and TP-B in Location L2. End-to-end immersive video endpoint calls deduct bandwidth from the
immersive bandwidth pool of the locations and the links along the effective path.
Figure 13-37
Location L1
Link L1 L2
BW
BW
Deducted
Audio
Audio
Video
Immersive
Deducted
BW
Deducted
Audio
Video
-4MB
Location L2
Video
Immersive
-4MB
Immersive
-4MB
TelePresence
TP-A
BW Req/Res
UCM
TelePresence
TP-B
292552
LBM
13-63
Chapter 13
Bandwidth Management
Figure 13-38 illustrates an end-to-end desktop video call between DP-A in Location L1 and DP-B in
Location L2. End-to-end desktop video endpoint calls deduct bandwidth from the video bandwidth pool
of the locations and the links along the effective path.
Call Flow for End-to-End Desktop Video
Location L1
Link L1 L2
BW
BW
Deducted
Audio
Deducted
Audio
Video
-1MB
Immersive
Location L2
BW
Video
-1MB
Immersive
Video
SIP
BW Req/Res
Desktop
DP-A
-1MB
Immersive
LBM
SIP
Deducted
Audio
UCM
Desktop
DP-B
292553
Figure 13-38
Figure 13-39 illustrates a video call between desktop video endpoint DP-A in Location L1 and
TelePresence video endpoint TP-B in Location L2. Interoperating calls between desktop video endpoints
and TelePresence video endpoints deduct bandwidth from both video and immersive locations and the
links bandwidth pools along the effective path.
Call Flow for Desktop-to-TelePresence Video
Location L1
Link L1 L2
BW
BW
Deducted
Audio
Deducted
Audio
Location L2
BW
Deducted
Audio
Video
-1MB
Video
-1MB
Video
-1MB
Immersive
-1MB
Immersive
-1MB
Immersive
-1MB
SIP
Desktop
DP-A
LBM
BW Req/Res
UCM
TelePresence
TP-B
292554
Figure 13-39
In Figure 13-40, a desktop video endpoint and two TelePresence endpoints call a SIP trunk configured
with a Video Traffic Class of immersive that points to a TelePresence MCU. Bandwidth is deducted
along the effective path from the immersive locations and the links bandwidth pools for the calls that are
end-to-end immersive and from both video and immersive locations and the links bandwidth pools for
the call that is desktop-to-immersive.
13-64
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Location L1
BW
Location L2
Link L1 L2
BW
Deducted
Deducted
BW
Audio
Audio
Deducted
Audio
Video
-1MB
Video
-1MB
Video
-1MB
Immersive
-1MB
-4MB
-4MB
Immersive
-1MB
-4MB
-4MB
Immersive
-1MB
-4MB
-4MB
TelePresence
TP-A
SIP Trunk
LBM
SIP
BW Req/Res
Desktop
DP-A
UCM
TelePresence
TP-B
TP MCU
292555
Figure 13-40
Figure 13-41 illustrates an end-to-end immersive video call across clusters, which deducts bandwidth
from the immersive bandwidth pool of the locations and links along the effective path.
Call Flow for End-to-End TelePresence Immersive Video Across Clusters
Location L1
Link L1 L2
Location L2
Link L1 L2
BW
BW
BW
BW
Deducted
Audio
Audio
Video
Immersive
Deducted
Audio
Video
-4MB
TelePresence
TP-A
Deducted
Audio
Video
Immersive
-4MB
LBM
BW Req/Res
Immersive
SIP
Deducted
SIP
SHADOW
BW
Deducted
Audio
Video
-4MB
Location L3
Video
Immersive
-4MB
LBM
BW Req/Res
Immersive
-4MB
TelePresence
TP-B
292556
Figure 13-41
1INVITE(L1, immersive)
UCM
180/200OK(L3, immersive)
UCM
Figure 13-42 illustrates an end-to-end desktop video call across clusters, which deducts bandwidth from
the video bandwidth pool of the locations and links along the effective path.
13-65
Chapter 13
Bandwidth Management
Location L1
Link L1 L2
Location L2
Link L2 L3
BW
BW
BW
BW
Deducted
Audio
Deducted
Audio
Video
-1MB
Immersive
Deducted
Audio
Video
-1MB
Immersive
Video
Deducted
Audio
-1MB
Immersive
Location L3
BW
Deducted
Audio
Video
-1MB
Immersive
Video
-1MB
Immersive
SIP
SIP
Desktop
DP-A
SIP
LBM
BW Req/Res
SIP
LBM
SHADOW
BW Req/Res
Desktop
DP-B
292557
Figure 13-42
1INVITE(L1, desktop)
UCM
UCM
180/200OK(L3, desktop)
Figure 13-43 illustrates a desktop video endpoint calling a TelePresence endpoint across clusters. the
call deducts bandwidth from both video and immersive bandwidth pools of the locations and links along
the effective path.
Figure 13-43
Location L1
Link L1 L2
Location L2
Link L1 L2
BW
BW
BW
BW
Deducted
Audio
Deducted
Audio
Deducted
Audio
Deducted
Audio
Location L3
BW
Deducted
Audio
Video
-1MB
Video
-1MB
Video
-1MB
Video
-1MB
Video
-1MB
Immersive
-1MB
Immersive
-1MB
Immersive
-1MB
Immersive
-1MB
Immersive
-1MB
Desktop
DP-A
LBM
BW Req/Res
SIP
SIP
SHADOW
LBM
BW Req/Res
TelePresence
TP-B
INVITE(L1, desktop)
UCM
180/200OK(L3, immersive)
UCM
13-66
292581
SIP
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
The figures in this section use a bidirectional arrow (<-->) to represent two unidirectional streams.
The following points summarize how Unified CM Enhanced Location CAC deducts bandwidth from
the configured audio, video, and immersive allocations. For more information, see the section on
Locations and Links, page 13-42:
Audio (audio-only calls): RTP bit rate + IP and UDP header overhead
Video (video calls): RTP bit rate only
Immersive (video calls by Cisco TelePresence endpoints): RTP bit rate only
symmetrically routed (both streams routed over the same path). For example, a G.711 audio call
of 80 kbps is 80 kbps in each direction over a full-duplex network; that is 80 kbps on the
transmit pair of wires and 80 kbps on the receive pair of wires, equating to 80 kbps full-duplex.
(See Figure 13-44.) Note that traffic is not always routed symmetrically in the WAN. Check
with your network administrator when necessary to ensure that admission control is correctly
accounting for the media as it is routed in the network over the WAN.
Real-Time Transport Control Protocol (RTCP) bandwidth overhead is not part of Unified CM
bandwidth allocation and should be part of network provisioning. RTCP is quite common in
most call flows and is commonly used for statistical information about the streams. It is also
used to synchronize audio in video calls to ensure proper lip-sync. In some cases it can be
enabled or disabled on the endpoint. RFC 3550 recommends that the fraction of the session
bandwidth added for RTCP should be fixed at 5%. What this means is that it is common practice
for the RTCP session to be up to 5% of the associated RTP session. So when calculating
bandwidth consumption on the network, you should add the RTCP overhead for each RTP
session. For example, if you have a G.711 audio call of 80 kbps with RTCP enabled, you will
be using up to 84 kbps per session (4 kbps RTCP overhead) for both RTP and RTCP. This
calculation is not part of Enhanced Location CAC deductions but should be part of network
provisioning.
Note
There are, however, methods to re-mark this traffic to another Differentiated Services Code Point
(DSCP). For example, RTCP uses odd-numbered UDP ports while RTP uses even-numbered
UDP ports. Therefore, classification based on UDP port ranges is possible. Network-Based
Application Recognition (NBAR) is another option as it allows for classification and re-marking
based on the RTP header Payload Type field. For more information on NBAR, see
http://www.cisco.com. However, if the endpoint marking is trusted in the network, then RTCP
overhead should be provisioned in the network within the same QoS class as audio RTP (default
marking is EF). It should also be noted that RTCP is marked by the endpoint with the same
marking as RTP; by default this is EF (verify that RTCP is also marked as EF).
13-67
Chapter 13
Bandwidth Management
Figure 13-44
Location L1
BW
Deducted
Audio
-80 kbps
Link L1 L2
BW
Deducted
Audio
-80 kbps
Location L2
BW
Audio
Video
Video
Video
Immersive
Immersive
Immersive
Deducted
-80 kbps
LBM
BW
Req/Res
UCM
Audio
Desktop
DP-A
Desktop
DP-B
346332
Audio RTCP
In Figure 13-44 two desktop video phones have established an audio-only call. In this call flow four
streams are negotiated: two audio streams illustrated by a single bidirectional arrow and two RTCP
streams also illustrated by a bidirectional arrow. For this call, the Location Bandwidth Manager (LBM)
deducts 80 kbps (bit rate + IP/UDP overhead) between location L1 and location L2 for a call established
between desktop phones DP-A and DP-B. The actual bandwidth consumed at Layer 3 in the network
with RTCP enabled would be between 80 kbps and 84 kbps, as discussed previously in this section.
In Figure 13-45 two desktop video phones have established a video call. In this call flow eight streams
are negotiated: two audio streams, two audio-associated RTCP streams, two video streams, and two
video-associated RTCP streams. Again for this illustration one bidirectional arrow is used to depict two
unidirectional streams. This particular call is 1024 kbps, with 64 kbps of G.711 audio and 960 kbps of
video (bit rate only for audio and video allocations of video calls). So in this case the LBM deducts
1024 kbps between locations L1 and L2 for a video call established between desktop phones DP-A and
DP-B. RTCP is overhead that should be accounted for in provisioning, depending on how it is marked
or re-marked by the network.
13-68
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-45
Location L1
BW
Deducted
Audio
Link L1 L2
BW
Location L2
Deducted
Audio
Video
-1024 kbps
Immersive
BW
Deducted
Audio
Video
-1024 kbps
Immersive
Video
-1024 kbps
Immersive
LBM
BW
Req/Res
UCM
Audio
Audio RTCP
Video
Video RTCP
Desktop
DP-B
346333
Desktop
DP-A
The example in Figure 13-46 is of a video call with presentation sharing. This is a more complex call
with regard to the number of associated streams and bandwidth allocation when compared to bandwidth
used on the network, and therefore it must be provisioned in the network. Figure 13-46 illustrates a video
call with RTCP enabled and Binary Floor Control Protocol (BFCP) enabled for presentation sharing. All
SIP-enabled telepresence multipurpose or personal endpoints such as a the Cisco TelePresence System
EX, MX, SX, C, and Profile Series function in the same manner.
13-69
Chapter 13
Bandwidth Management
Figure 13-46
Video Call with RTCP and BFCP Enabled and Presentation Sharing
Location L1
BW
Deducted
Link L1 L2
BW
Audio
Audio
Video
Video
Immersive
-6.5 Mbps
Service Parameter:
Use Video BandwidthPool
for Immersive Video Calls
= False
Enables Immersive Pool
Deducted
Location L2
BW
Deducted
Audio
Video
Immersive
-6.5 Mbps
Immersive
-6.5 Mbps
LBM
BW
Req/Res
UCM
Audio
Audio RTCP
Presentation
Video Included in
Total Allocated
Bandwidth
Video
Video RTCP
Video Endpoint
TP-B
Presentation Video
346334
Video Endpoint
TP-A
BFCP Signaling
When a video call is established between two video endpoints, audio and video streams are established
and bandwidth is deducted for the negotiated rate. Unified CM uses regions to determine the maximum
bit rate for the call. For example, with a Cisco TelePresence System EX90 at the highest detail of 1080p
at 30 frames per second (fps), the negotiated rate between regions would have to be set at 6.5 Mbps.
EX90s used in this scenario would average around 6.1 Mbps for this session. When the endpoints start
presentation sharing during the session, BFCP is negotiated between the endpoints and a new video
stream is enabled at either 5 fps or 30 fps, depending on endpoint configuration. When this occurs, the
endpoints will throttle down their main video stream to include the presentation video so that the entire
session does not use more than the allocated bandwidth of 6.5 Mbps. Thus, the average bandwidth
consumption remains the same with or without presentation sharing.
Note
The presentation video stream is typically unidirectional in the direction of the person or persons
viewing the presentation.
Telepresence immersive and office endpoints such as the Cisco TelePresence System 500, 1000, 3000,
and TX9000 Series that negotiate a call between one another function a little differently in the sense that
the video for presentation sharing is an additional bandwidth above and beyond what is allocated for the
main video session, and thus it is not deducted from Enhanced Location CAC. Figure 13-47 illustrates
this.
13-70
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-47
Location L1
BW
Deducted
Audio
BW
Deducted
Audio
Video
Immersive
Link L1 L2
Service Parameter:
Use Video BandwidthPool
for Immersive Video Calls
= False
Enables Immersive Pool
BW
Deducted
Audio
Video
-4 Mbps
Location L2
Video
Immersive
-4 Mbps
Immersive
-4 Mbps
LBM
BW
Req/Res
UCM
Audio
Audio RTCP
Video
Video RTCP
Presentation
Video in Addition
to Total Allocated
Bandwidth
Presentation Video
Video Endpoint
TP-B
346335
Video Endpoint
TP-A
BFCP Signaling
In Figure 13-47 the telepresence immersive video endpoints establish a video call and enable
presentation sharing. The LBM deducts 4 Mbps for the main audio and video session from the immersive
pool for the call, and video is established between the endpoints. When presentation sharing is activated,
the two endpoints exchange BFCP and negotiate a presentation video stream at 5 fps or 30 fps in one
direction, depending on the endpoint configuration. At 5 fps the average bandwidth used is
approximately 500 kbps of additional bandwidth overhead. This bandwidth is above and beyond the
4 Mbps that was allocated for the video call and should be provisioned in the network. At 30 fps the
average bit rate of the presentation video is approximately 1.5 Mbps.
Note
The Cisco TelePresence System endpoints use Telepresence Interoperability Protocol (TIP) to multiplex
multiple screens and audio into two audio and video RTP streams in each direction. Therefore the actual
streams on the wire may be different than what is expressed in the illustration, but the concept of
additional bandwidth overhead for the presentation video is the same.
13-71
Chapter 13
Bandwidth Management
UCM/SUB
UCM/SUB
BW
Upgrade
Unlimited
Video
Unlimited
Shadow
LBM
BW
Allocated
BW
Allocated
Audio
Unlimited
Audio
Unlimited
Video
Unlimited
Video
Unlimited
Hub_none
BW
1200
Video
3840
Immersive
Hub_none
Link to Hub_none
Allocated
Audio
Unlimited
Location 1
BW
Allocated
Audio
1200
Video
3840
Immersive
Allocated
Audio
width
and
rate B
Value
Mig
Unlimited
Location N...
BW
Link to Hub_none
Allocated
BW
Audio
1200
Audio
Video
3840
Video
Immersive
Unlimited
BW
Allocated
3840
Unlimited
BW
Allocated
1200
Audio
Video
3840
Video
Unlimited
1200
Immersive
Audio
Immersive
Allocated
Immersive
Location 1
1200
3840
Unlimited
Location N...
348706
Figure 13-48
Settings after an upgrade to a Cisco Unified CM release that supports Enhanced Location CAC:
The LBM is activated on each Unified CM subscriber running the Cisco CallManager service.
The Cisco CallManager service communicates directly with the local LBM.
Bandwidth values assigned to locations are migrated to a link connecting the user-defined location
and Hub_None.
13-72
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
SC
23
CP
H3
IP/
/S
CP
SC
/SI
P/H
32
SCCP
SIP T
H323
Media
G.711 (64 kbps)
SIP
SIP T
H323
Media
XCODE
G.729 (8 kbps)
SIP
XCODE
292566
Figure 13-49
For dual stack MTP insertion, the bit rate is different on each connection but the bandwidth is different
due to IP header overhead. Figure 13-50 illustrates the difference in bandwidth used for IPv4 and IPv6
networks with dual stack MTP insertion.
Bandwidth Differences for Dual Stack MTP Insertion
SC
CP
2
/H3
/SI
P
/SI
C
SC
SIP T
H323
SIP
P/H
32
SCCP
IPv4 Media
G.711 (64 kbps)
+ L3 Overhead = 80K
IPv6 Media
XCODE
XCODE
SIP T
H323
SIP
292567
Figure 13-50
Enhanced Location CAC does not account for these differences in bandwidth between MTPs and
transcoders. The service parameter Locations Media Resource Audio Bit Rate Policy determines
whether the largest or smallest bandwidths should be used along the locations and links path. Lowest Bit
Rate (default) or Highest Bit Rate can be used to manage these differences in bandwidth consumption.
13-73
Chapter 13
Bandwidth Management
Cisco Unified CM 10.0 or a later release required on both home and visiting clusters
The visiting and home clusters must be in the same intercluster LBM replication network
Both Enhanced Location CAC and EMCC can be designed and deployed according to the guidelines in
the product documentation and this SRND. There are no other requirements or any specific configuration
aspects to employ.
13-74
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
L1
DC2
L2
L3
L4
292568
DC1
Typically these WAN uplinks from the remote sites to the data centers are load-balanced or in a
primary/backup configuration, and there are limited ways for a static CAC mechanism to handle these
scenarios. Although you could configure this multi-path topology in Enhanced Location CAC, only one
path would be calculated as the effective path and would remain statically so until the weight metric was
changed. A better way to support this type of network topology is to configure the two data centers as
one data center or hub location in Enhanced Location CAC and configure a single link to each remote
site location. Figure 13-52 illustrates an Enhanced Location (E-L) CAC locations and links overlay.
Figure 13-52
DC1
DC2
L1
L2
L3
L4
Location 1
Location 2
Location 3
Location 4
348707
13-75
Chapter 13
Bandwidth Management
Design Recommendations
The following design recommendations for dual data centers with remote dual or more links to remote
locations apply to both load-balanced and primary/backup WAN designs:
A single link between the remote locations and Hub_None protects the remote site uplinks from
over-subscription during normal conditions or failure of the highest bandwidth capacity links.
The capacity of link bandwidth allocation between the remote site and Hub_None should be equal
to the lowest bandwidth capacity for the applicable Unified Communications media for a single link.
For example, if each WAN uplink can support 2 Mbps of audio traffic marked EF, then the link audio
bandwidth value should be no more than 2 Mbps to support a failure condition or equal-cost path
routing.
MPLS Clouds
When designing for Multiprotocol Label Switching (MPLS) any-to-any connectivity type clouds in the
Enhanced Location CAC network model, a single location can serve as the MPLS cloud. This location
will not have any devices associated to it, but all of the sites that have uplinks to this cloud will have
links configured to the location. In this way the MPLS cloud serves as a transit location for
interconnecting multiple variable-sized bandwidth WAN uplinks to other remote locations. The
illustrations in this section depict a number of different MPLS networks and their equivalent locations
and links model.
In Figure 13-53, Hub_None represents the MPLS cloud serving as a transit location interconnecting the
campus location where servers, endpoints, and devices are located, with remote locations where only
endpoints and devices are located. Each link to Hub_None from the remote location may be sized
according to the WAN uplink bandwidth allocated for audio, video, and immersive media.
13-76
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-53
Network Topology
Campus
IP
IP
IP
Hub_None
V
MPLS
Any/Any
Campus
Loc_1
Loc_2 Loc_N
Links
V
IP
IP
V
IP
IP
IP
Loc_1
IP
IP
IP
Loc_2
IP
348708
Loc_N
Figure 13-54 shows two MPLS clouds that serve as transit locations interconnecting the campus location
where servers, endpoints, and devices are located, with remote locations where only endpoints and
devices are located. The campus also connects to both clouds. Each link to the MPLS cloud from the
remote location may be sized according to the WAN uplink bandwidth allocated for audio, vide, and
immersive media. This design is typical in enterprises that span continents, with a separate MPLS cloud
from different providers in each geographical location.
13-77
Chapter 13
Bandwidth Management
Figure 13-54
Campus
IP
Campus
IP
IP
MPLS_1
MPLS_2
V
MPLS
Any/Any
MPLS
Any/Any
Links
IP
IP
IP
IP
IP
Loc_1
IP
IP
Loc_3
IP
IP
IP
IP
Loc_4
IP
Loc_6
IP
IP
Loc_2
IP
348709
IP
IP
IP
Loc_5
Figure 13-55 shows multiple MPLS clouds from different providers, where each site has one connection
to each cloud and uses the MPLS clouds in either an equal-cost load-balanced manner or in a
primary/backup scenario. In any case, this design is equivalent to the dual data center design where a
single location represents both clouds and a single link represents the lowest capacity link of the two.
13-78
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-55
Campus
IP
Campus
IP
IP
V
MPLS
Any/Any
MPLS
Any/Any
MPLS
Links
V
IP
IP
IP
IP
IP
Loc_1
IP
IP
IP
Loc_N
348710
IP
Loc_2
Design Recommendations
The MPLS cloud should be configured as a location that does not contain any endpoints but is used
as a hub to interconnect locations.
The MPLS cloud serves as a transit location for interconnecting multiple variable-sized bandwidth
WAN uplinks to other remote locations.
Remote sites with connectivity to dual MPLS clouds should treat those connections as a single link
and size to the lowest capacity of the links in order to avoid oversubscription during network failure
conditions.
13-79
Chapter 13
Bandwidth Management
Group Policy Objects (GPO) using application, IP addresses, and UDP/TCP port ranges to mark traffic
with DSCP. Group Policy Objects are very similar in function to network access lists in their ability to
mark traffic.
QoS is critical to admission control because without it the network has no way of prioritizing the media
to ensure that admitted traffic gets the network resources that it requires above that of non-admitted or
other traffic classifications. In Unified CM's CallManager service parameters for QoS, there are five
main QoS settings that are applicable to endpoint media classification and that also allow immersive and
desktop classified endpoints (see the section on Enhanced Location CAC for TelePresence Immersive
Video, page 13-60) to have different QoS markings for their media based on their video classification of
immersive or desktop. Table 13-13 shows the five main DSCP settings along with their default settings
and Per Hop Behavior (PHB) equivalents.
Table 13-13
Default Value
PHB Equivalent
46
EF
34
AF41
34
AF41
32
CS4
32
CS4
The DSCP for Audio Calls setting is used for any device that makes an audio-only call. The DSCP for
Video Calls setting is used for the audio and video traffic of any device that is classified as "desktop."
DSCP for TelePresence Calls is used for the audio and video traffic of any device that is classified as
"immersive." The DSCP for Audio Portion of Video Calls and DSCP for Audio Portion of
TelePresence Calls are currently applicable to a subset of video endpoints and differentiate only the
audio portion of video calls dependent on the video call type based on classification. See the section on
Trusted Endpoints, page 13-93, for more information.
As mentioned in the section on Enhanced Location CAC for TelePresence Immersive Video, page 13-60,
Cisco Unified CM E-LCAC has the ability to perform admission control for TelePresence calls
separately from other video calls. E-LCAC does this through a classification of endpoints and SIP trunks
as "immersive" or "desktop." This classification gives Unified CM the ability to deduct bandwidth from
a separate immersive bandwidth pool for those devices and trunks classified as immersive. By default
LBM deducts ALL video, no matter the classification, from the video bandwidth pool (Unified CM's
CallManager service parameter Use Video BandwidthPool for Immersive Video Calls set to True).
Also by default, all immersive classified endpoints have a DSCP set to CS4 (DSCP 32; DSCP for
TelePresence Calls), while desktop endpoints have a DSCP set to AF41 (DSCP 34; DSCP for Video
Calls). The default settings for QoS and E-LCAC differentiate DSCP but deduct all video from the same
E-LCAC bandwidth pool. Figure 13-56 illustrates the QoS and E-LCAC bandwidth pool associations
and defaults for immersive and desktop classified devices.
13-80
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
Figure 13-56
DSCP Value
PHB Value
CAC Pool
Audio Only
46
EF
Voice
Audio Poron of
*DSCP forVideo
Calls
Audio of Video
34
AF41
Video
Video of Video
34
AF41
Video
Audio of TP
32
CS4
Video
Video of TP
32
CS4
Video
348977
DSCP for TelePresence Calls is the immersive classification, and DSCP for Video Calls is the desktop
classification.
Design Recommendations
The following design recommendations apply to video solutions that employ Enhanced Location CAC:
If you are deploying Unified Communications video (desktop classification) and TelePresence video
(immersive classification) where differentiation between desktop video and TelePresence video is a
requirement, then ensure that the Unified CM service parameter Use Video Bandwidth Pool for
Immersive Video Calls is set to false. This enables the immersive bandwidth pool for TelePresence
calls.
In Enhanced Location CAC, TelePresence endpoints can be managed in the same location as Unified
Communications video endpoints. If TelePresence calls are not to be tracked through Enhanced
Location CAC, then set the immersive location and links bandwidth pool to unlimited. This will
ensure that CAC will not be performed on TelePresence or SIP trunks classified as immersive. If
TelePresence calls are to be tracked through Enhanced Location CAC, then set immersive location
and links bandwidth pool to a value according to the bit rate used and the number of calls to be
allowed over the locations and link paths.
Cisco Unified CM uses two different cluster-wide QoS service parameter to differentiate between
the Differentiated Services Code Point (DSCP) settings of UC video endpoints and TelePresence
endpoints. TelePresence endpoints use the DSCP for Telepresence calls QoS parameter while the
Cisco UC video endpoints use the DSCP for video calls QoS service parameter.
13-81
Chapter 13
Bandwidth Management
When marking video with the default QoS markings, the following recommendations apply:
For sites that deploy only UC endpoints and no TelePresence endpoints, ensure that the CS4
DSCP class is added to the AF41 QoS traffic class on inbound WAN QoS configurations to
account for the inbound CS4 marked traffic, thus ensuring QoS treatment of CS4 marked media.
For sites that deploy only UC TelePresence endpoints and no UC endpoints, ensure that the
AF41 DSCP class is added to the CS4 QoS traffic class on inbound WAN QoS configurations
to account for the inbound AF41 marked traffic, thus ensuring QoS treatment of AF41 marked
media.
Design Considerations
When deploying Enhanced Location CAC for immersive video calls, consider the affects of DSCP
marking for both QoS classes, as the interoperable calls where an immersive classified endpoint is
connected with a desktop classified endpoint are by default asymmetrically marked.
Bandwidth Deductions and Media Marking in a Multi-Site Deployment with Enhanced Location CAC
Service Param:
Use Video BandwidthPool
for Immersive Video Calls*
= False
(Default = True)
Cisco Unified CM
IP
P/S
SIP
SC
Location A
Location B
IP WAN
TelePresence
UC Video
Phone
Link Location A to B
Configured
Used
Audio
Unlimited
Video
Unlimited
N/A
Immersive
Unlimited
N/A
Configured
Audio
800k
Video
1536k
Immersive
5000k
Configured
Used
Used
Audio
Unlimited
768k
Video
Unlimited
N/A
768k
Immersive
Unlimited
N/A
292559
SIP
13-82
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
All leaf clusters that support Enhanced Location CAC should be enabled for intercluster Enhanced
Location CAC with SME.
SME can be used as a centralized bootstrap hub for the Enhanced Location CAC intercluster hub
replication network. See LBM Hub Replication Network, page 13-52, for more information.
All trunks to leaf clusters supporting Enhanced Location CAC should be SIP trunks placed in the
shadow location to enable Enhanced Location CAC on the trunk between SME and the leaf clusters
supporting Enhanced Location CAC.
For TelePresence video interoperability, see the section on Call Admission Control Design
Recommendations for Video Deployments, page 13-79.
Connectivity from SME to any trunk or device other than a Unified CM that supports Enhanced
Location CAC (some examples are third-party PBXs, gateways, Unified CM clusters that support
only traditional Location CAC, voice messaging ports or trunks to conference bridges, Cisco Video
Communications Server, and so forth) should be configured in a location other than a phantom or
shadow location. The reason for this is that both phantom and shadow locations are non-terminating
13-83
Chapter 13
Bandwidth Management
locations; that is, they relay information about locations and are effectively placeholders for
user-defined locations on other clusters. Phantom locations are legacy locations that allow for the
transmission of location information in versions of Unified CM that support only traditional
Location CAC, but they are not supported with Unified CM Enhanced Location CAC. Shadow
locations are special locations that enable trunks between Unified CM clusters that support
Enhanced Location CAC to accomplish it end-to-end.
Figure 13-58
SME can be used as a locations and link management cluster. See Figure 13-58 as an example of
this.
Global Topology
LBM Replication
Hub-none
SME
Global Topology
Hub-none
Loc_31
Loc_31
Hub_none
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
LBM
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
HUB
Loc_21 Loc_22
LBM
HUB
LBM
HUB
Leaf 1
LBM
HUB
Leaf 2
LBM
HUB
Leaf 3
HUB
Leaf 4
Global Topology
Hub-none
Loc_31
Hub_none
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
Global Topology
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
Loc_21 Loc_22
Global Topology
Hub-none
Hub-none
Loc_31
Hub-none
Hub_none
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
Loc_12
Loc_21
Loc_31
Leaf 6
Global Topology
Loc_31
Hub-none
HUB
Leaf 5
Hub-none
Global Topology
LBM
Hub_none
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
Loc_31
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
Hub-none
Loc_31
Loc_11 Loc_12
Loc_21 Loc_22 Loc_24
Figure 13-58 illustrates SME as a location and link management cluster. The entire location and link
global topology is configured and managed in SME, and the leaf clusters configure locally only the
locations that they require to associate with the end devices. When intercluster Enhanced Location CAC
is enabled and locations and links are replicated, each leaf cluster will receive the global topology from
SME and overlay this on their configured topology and use the global topology for call admission
control. This simplifies configuration and location and link management across multiple clusters, and it
13-84
292561
LBM
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
diminishes the potential for misconfiguration across clusters. For more information and details on the
design and deployment see the section on Location and Link Management Cluster, page 13-57.
Figure 13-59 illustrates an SME design where intercluster Enhanced Location CAC has been enabled on
one or more leaf clusters (right) and where one or more leaf clusters are running a version of Unified CM
that supports only traditional Location CAC (left). In this type of a deployment the locations managed
by traditional Location CAC cannot be common or shared locations between clusters enabled for
Enhanced Location CAC. Leaf 1 has been configured in a traditional hub and spoke, where devices are
managed at various remote sites. SME and the other leaf clusters that are enabled for intercluster
Enhanced Location CAC share a global topology, as illustrated in the E-L CAC Modeled Topology.
Leaf1_Hub is a user-defined location in SME assigned to the SIP or H.323 intercluster trunk that
represents the hub of the Leaf 1 topology. This allows SME to deduct bandwidth for calls to and from
Leaf 1 up to the Leaf1_Hub. In this way SME and Leaf 2 manage the Enhanced Location CAC locations
and links while Leaf 1 manages its remote locations with traditional Location CAC.
Figure 13-59
SME Design with Enhanced Location CAC and Traditional Location CAC in Leaf Clusters
Location CAC
LBM
HUB
SIP T
H323
SIP T
H323
Hub_none
Leaf1_Hub
Leaf 1
Unified CM 8.x
SIP T
SIP T
Shadow
Shadow
LBM
HUB
LBM
HUB
Leaf 2
SME
IP
IP
IP
IP
IP
IP
IP
Leaf 1
Leaf 1
Remote1 Remote2
Site N
Remote1
Remote2
IP
Site 1
IP
IP
IP
IP
IP
Site 1
Site 2
Site N
Site 2
348711
Leaf 1
Location CAC Hub
and Spoke Topology
Hub_none
IP
13-85
Chapter 13
Bandwidth Management
For Internet-based endpoints calling one another, the media is routed through Cisco Expressway E
and Expressway C back out to the Internet, as is illustrated between endpoints B and C in
Figure 13-60.
2.
For Internet-based endpoints calling internal endpoints, the media flows through the Expressway E
and Expressway C, as is illustrated between endpoints A and C in Figure 13-60.
Figure 13-60
B
Inside firewall
(Intranet)
DMZ
Collaboration
Services
Outside firewall
(Internet)
Internet
Unified
CM
Expressway Expressway
C
E
MEDIA
348712
SIGNALING
13-86
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
For multiple deployments of Cisco Expressway for VPN-less access in the same enterprise, with the
Internet-based endpoints registered through one Expressway pair calling Internet-based endpoints
registered through another Expressway pair, the media will be routed through the enterprise. This is
illustrated in Figure 13-61 with a call between endpoint D and endpoint C, both registered from the
Internet but through two different Expressway pairs. The media flow will be the same whether the
endpoints are registered to the same Unified CM cluster or to different Unified CM clusters.
Figure 13-61
Inside firewall
(Intranet)
DMZ
Outside firewall
(Internet)
Unified
CM
Collaboration
Services
Internet
Inside firewall
(Intranet)
DMZ
Outside firewall
(Internet)
SIGNALING
MEDIA
348713
Unified
Unifie
CM Expressway Expressway
C
E
Figure 13-62 illustrates an example configuration for locations and links that integrate bandwidth
tracking for media flows that traverse the enterprise, while still allowing media flows over the Internet
without admission control.
13-87
Chapter 13
Bandwidth Management
Figure 13-62
MPLS
Location
Link
RTP - MPLS
Link
BLD - MPLS
Link
RTP - MPLS
Audio=1MB
Video=3MB
Audio=1MB
Video=3MB
Audio=1MB
Video=3MB
RTP
BLD
SJC
Link
RTP - RTP_INET
Link
BLD - BLD_INET
Link
SJC - SJC_INET
UNLIMITED
UNLIMITED
UNLIMITED
RTP_INET
BLD_INET
SJC_INET
INTERNET LOCATIONS
348714
Enterprise Locations
Figure 13-62 illustrates an example deployment of ELCAC consisting of three main sites: RTP, BLD,
and SJC. These sites are all connected to an MPLS provider and thus each has a separate WAN
connection to the MPLS cloud. Locations and links are created accordingly so that the enterprise
locations are linked directly to a location called MPLS, with bandwidth links limited for audio and video
calls mapping to the network topology. Devices are located in one of the three sites when in the
enterprise and thus have a location associated to them. Each of these sites has a Cisco Expressway
solution for VPN-less remote and mobile access for Internet-based endpoints registering to Unified CM.
Three new locations are configured for the Internet-based devices, one for each Expressway solution
site, named RTP_INET, BLD_INET, and SJC_INET. These three locations represent "Internet locations"
because they are locations for devices registering from the Internet to Unified CM through an
Expressway pair. These locations are not interconnected with direct links. This is because calls between
Expressways are routed through the enterprise and thus flow through the MPLS cloud. These Internet
locations, instead, have a link to their associated enterprise location. For example, RTP_INET has a link
to RTP, BLD_INET has a link to BLD, and so forth. These links between the Internet locations and the
enterprise locations should be set to unlimited bandwidth.
As mentioned, Enhanced Location CAC for Cisco Expressway deployments requires the use of a feature
in Unified CM called Device Mobility. (For details about this feature, see the section on Device Mobility,
page 21-13.) Enabling device mobility on the endpoints allows Unified CM to know when the device is
13-88
Chapter 13
Bandwidth Management
Enhanced Locations Call Admission Control
registered through the Cisco Expressway or when it is registered from within the enterprise. Device
mobility also enables Unified CM to provide admission control for the device as it roams between the
enterprise and the Internet. Device mobility is able to do this by knowing that, when the endpoints
register to Unified CM with the IP address of Expressway C, Unified CM will associate the applicable
Internet location. However, when the endpoint is registered with any other IP address, Unified CM will
use the enterprise location that is configured directly on the device (or from the device pool directly
configured on the device). It is important to note that device mobility does not have to be deployed across
the entire enterprise for this function to work. Configuration of Device Mobility in Unified CM is
required only for the Expressway IP addresses, and the feature is enabled only on the devices that require
the function (that is to say, those devices registering through the Internet). Figure 13-63 illustrates an
overview of the device mobility configuration. Although this is a minimum configuration requirement
for Device Mobility for ELCAC to function for internet-based devices, Device Mobility can be
configured to support mobility for these same endpoints within the enterprise. (See the section on Device
Mobility, page 21-13, for more information.)
Figure 13-63
Device
Mobility Info
Location
RTP_EXP_DP
RTP INT
BLD_EXP_DP
BLD INT
SJC_EXP_DP
SJC INT
RTP_EXP1_DMI
10.10.20.50/32
RTP_EXP2_DMI
10.10.20.51/32
BLD_EXP1_DMI
10.10.30.50/32
BLD_EXP2_DMI
10.10.30.51/32
SJC_EXP2_DMI
10.10.40.51/32
348715
SJC_EXP1_DMI
10.10.40.50/32
Figure 13-63 shows a simplified version of device mobility for the example deployment of ELCAC as
described in Figure 13-62. The IP addresses of the Expressway C servers are configured in the device
mobility information. In this example there is a redundant pair of Expressway C servers for each of the
three sites, RTP, BLD, and SJC. RTP_EXP1_DMI and RTP_EXP2_DMI are configured respectively
with the server IP addresses of the RTP Expressway C servers. These two are associated to a new device
pool called RTP_EXP_DP, which has the location RTP_INET configured on it. Each site is configured
similarly. With this configuration, when any device enabled for device mobility registers to Unified CM
with the IP Address that corresponds to the device mobility information in RTP_EXP1_DMI or
RTP_EXP2_DMI, it will be associated with the RTP_EXP_DP device pool and thus with the RTP_INET
location.
With the above configuration, when an Internet-based device registers through the Expressway to
Unified CM, it will register with the IP address of Expressway C. Unified CM then uses the IP address
configured in the device mobility information and associates the device pool and thus the Internet
13-89
Chapter 13
Bandwidth Management
location associated to this device pool. This process is illustrated in Figure 13-64.
Figure 13-64
2. Device in RTP
IP Subnet
Device Mobility
Info
Device Pool
Location
10.10.20.50
RTP_EXP1_DMI
RTP_EXP_DP
RTP_INET
10.10.30.50
BLD_EXP1_DMI
BLD_EXP_DP
BLD_INET
10.10.40.50
SJC_EXP1_DMI
SJC_EXP_DP
SJC_INET
348716
3. Here is your
configuration
In Figure 13-64 the client registers with Unified CM through the Expressway in RTP. Because the
signaling is translated at the Expressway C in RTP, the device registers with the IP address of the
Expressway C. The device pool RTP_EXP_DP is associated to the device based on this IP address. The
RTP_EXP_DP pool is configured with the RTP_INET location, and therefore that location is associated
to the device. Thus, when devices register to the Expressway, they get the correct location association
through device mobility. When the endpoint relocates to the enterprise, it will return to its static location
configuration. Also, if the endpoint relocates to another Expressway in SJC, for example, it will get the
correct location association through device mobility.
Design and Deployment Best Practices for Cisco Expressway VPN-less Access with Enhanced
Location CAC
Each site with Internet access, where a Cisco Expressway solution resides, requires an Internet
location and an enterprise location. Each Cisco Expressway deployment requires these location
pairs. The enterprise location is associated to devices when they are in the enterprise (see locations
RTP, BLD, and SJC in Figure 13-62). The Internet location is associated to the endpoints through
the Device Mobility feature when the endpoints are registering from the Internet (see locations
RTP_INET, BLD_INET, and SJC_INET in Figure 13-62). For example, in Figure 13-62, RTP and
RTP_INET form a location pair for the physical site RTP.
13-90
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Internet locations will always have a single link to the enterprise location that they are paired with.
For example, in Figure 13-62, RTP and RTP_INET form an enterprise location and internet location
pair.
Links from Internet locations to enterprise locations are set to unlimited bandwidth. Unlimited
bandwidth between these location pairs ensures that bandwidth is not counted for calls from the
Internet location to the local enterprise location, and vice versa (for example, calls from RTP to
RTP_INET in Figure 13-62).
In a Cisco Expressway solution where more than one Cisco Expressway site is deployed, and
requiring multiple Internet locations, ensure that Internet locations do not have direct links between
one another. Direct links between Internet locations will create multiple paths in ELCAC, and for
that reason they are not recommended.
Example Enterprise #1
Example Enterprise #1 is a large enterprise with users across a large geographic area, with a data center
(DC) at the headquarters site as well as multiple large, small, and micro-sized branches with roughly
500, 50, 15, and 5 users in each branch type, respectively. To simplify the illustration of the network,
these categories of sites (HQ, large, small and micro) are used as a template to size bandwidth
considerations for each site that has a similar user base and endpoint density. Figure 13-65 illustrates
each type of site. The enterprise has deployed Jabber with video to ensure that users have access to a
video terminal for conferencing. The TelePresence video conferencing resources are located in the DC
at HQ. IP phones are for voice-only communications; video endpoints are Jabber clients, Collaboration
desktop endpoints (DX Series), and room endpoints (MX, Profile, and SX Series); and the HQ and large
sites have immersive TelePresence units such as the IX Series.
13-91
Chapter 13
Bandwidth Management
Figure 13-65
Example Enterprise #1
Unified CM
Multipoint
Video (TS)
500 users:
Central
Site
Expressway
B2B
Cloud
Services
S
erv
r ices
s
Mobile
Users
Internet
Int
MPLS VPN
15 users:
5 users:
50 Jabber clients
30 IP phones
6 video endpoints
(room + desktop)
15 Jabber clients
10 IP phones
2 video endpoints
(room)
5 Jabber clients
3 IP phones
1 video endpoint
Large Branch
Small
S
Branch
Micro Branch
348861
50 users:
The IT department is tasked with determining the bandwidth requirements for the WAN edge for each
type of site in Example Enterprise #1. The following sections list the requirements and illustrate a
methodology for applying QoS and for determining bandwidth and queuing requirements as well as
admission control requirements.
13-92
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Jabber clients
(mobile and desktop):
DSCP
UDP Ports
3xxx
ACL
AF41
4xxx
AF42
5060
CS3
348862
Figure 13-66
The administrator creates an ACL for the access switches for the data VLAN to re-mark UDP ports to
the following DSCP values:
Audio streams of all Jabber calls (voice-only and video calls) are marked AF41.
For the Jabber endpoints, we also recommend changing the default QoS values in the Jabber SIP profile.
This is to ensure that, if for any reason the QoS of a Jabber client is trusted via a wireless route or any
other wired route, the correct trusted values will be coherent between the trusted QoS and the QoS that
is re-marked with the ACLs. Therefore, the QoS parameters in the SIP Profile for Jabber clients need to
be set as shown in Table 13-14.
Table 13-14
Changed Value
EF
AF41
AF41
AF42
AF41
CS4
N/A
CS4
N/A
The configuration settings in Table 13-14 ensure that video of Jabber clients will be set to AF42 if for
any reason the traffic follows a trusted network path and is not re-marked via UDP port ranges as in the
untrusted network path. The DSCP for Audio Portion of Video Calls is left at the default setting of AF41.
This is simply to ensure a consistent configuration across Jabber endpoints, whether trusted or re-marked
via the network using UDP port ranges.
Trusted Endpoints
For the trusted endpoints, Cisco Discovery Protocol (CDP) is used and the QoS of the IP phones and
video endpoints is trusted using the conditional trust mechanism configured at the access switch. The
configuration uses the Unified CM default system settings of audio for voice-only calls as EF, audio and
13-93
Chapter 13
Bandwidth Management
video for video calls as AF41, audio and video for TelePresence as CS4, and signaling as CS3. Therefore,
the administrator must change the QoS defaults in Unified CM for the trusted endpoints with a SIP
Profile to ensure that the QoS of the TelePresence endpoints is adjusted accordingly.
Figure 13-67 illustrates the conditional trust (CDP based) and packet marking at the access switch.
Figure 13-67
DSCP
T t
Trust
AF41
CS3
Trusted TelePresence
and Desktop Video Endpoints
DSCP
EF
348863
T t
Trust
CS3
Trusted IP Phones
The administrator configures all access switches with a conditional QoS trust for IP phones and video
and TelePresence endpoints, classified as follows:
The administrator must also change the QoS defaults in Unified CM for the trusted endpoints with a SIP
Profile using the values in Table 13-15.
Table 13-15
Changed Value
EF
AF41
AF41
CS4
AF41
CS4
AF41
On ingress at the WAN edge, it is expected that the packets arriving with a specific DSCP value have
been trusted at the access layer or re-marked accordingly if they were not trusted at the access switch.
As a failsafe practice, on ingress it is important to re-mark any untrusted traffic at the WAN edge that
13-94
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
could not be re-marked at the access layer. While QoS is important in the LAN, it is paramount in the
WAN; and as routers assume a trust on ingress traffic, it is important to configure the correct QoS policy
that aligns with the business requirements and user experience. The WAN edge re-marking is always
done on the ingress interface into the router, while the queuing and scheduling is done on the egress
interface. The following example walks through the WAN ingress QoS policy as well as the egress
queuing policy. Figure 13-68 illustrates the configuration and the re-marking process.
In Figure 13-68 the packets from both the trusted and untrusted areas of the network are identified and
classified with the appropriate DSCP marking via the trust methods discussed or via a simple ACL
matching on UDP port ranges. Keep in mind that this ACL could also match more granularly on IP
addresses or some other attributes that would further limit the scope of the marking.
Figure 13-68
UDP Ports
DSCP
3xxx
AF41
4xxx
5060
AF42
ACL
Trust
Ingress
Policy
CS3
EF
348864
AF41
CS3
Figure 13-69 through Figure 13-73 illustrate the ingress QoS policy matching criteria and DSCP
re-marking. The process involves the following steps shown in the figures:
1.
Packets arrive into the router ingress interface, which is configured with an input service policy
(Figure 13-69).
2.
The policy-map is configured with 4 classes of traffic setting the appropriate DSCP: Voice setting a
DSCP of EF, Prioritized-Video (includes Jabber audio) setting a DSCP of AF41, Jabber-Video
setting a DSCP of AF42, and Signaling setting a DSCP of CS3 (Figure 13-69).
3.
Each one of these classes matches a class-map of the same name configured with match-any criteria
and a DSCP match as well as an ACL match (Figure 13-69). This match-any criteria means that the
process will start top-down, and the first matching criteria will be executed and thus set the DSCP
according to each class in the policy-map statements. Another option is match-all, which would
require all criteria to be matched and thus would match DSCP and ACL. This, however, would not
provide the intended functionality of re-marking either marked or unmarked traffic.
13-95
Chapter 13
Bandwidth Management
Figure 13-69
Ingress
Policy
4.
policy-map INGRESS-MARKING
class VOICE
set dscp ef
class PRIORITIZED-VIDEO
set dscp af41
class JABBER-VIDEO
set dscp af42
class SIGNALING-SIP
set dscp cs3
class class-default
348865
The first match statement in the class-map is match dscp, If the traffic matches the DSCP, then
DSCP is set again to the same value that was matched, as is configured in the policy-map statements
(Figure 13-70).
13-96
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Figure 13-70
Ingress
Policy
5.
policy-map INGRESS-MARKING
class VOICE
set dscp ef
class PRIORITIZED-VIDEO
set dscp af41
class JABBER-VIDEO
set dscp af42
class SIGNALING-SIP
set dscp cs3
348866
class class-default
If DSCP was not matched, then the next line in the class-map statement is parsed, which is the ACL
that matches the UDP ports set in Unified CM for the Jabber clients (see Identification and
Classification, page 13-17). When the ACL criteria (protocol and port range) are met, then the traffic
is set as is configured in the corresponding policy-map statements (Figure 13-71).
13-97
Chapter 13
Bandwidth Management
Figure 13-71
Ingress
Policy
Note
6.
policy-map INGRESS-MARKING
class VOICE
set dscp ef
class PRIORITIZED-VIDEO
set dscp af41
Egress
Policy
class JABBER-VIDEO
set dscp af42
class SIGNALING-SIP
set dscp cs3
class class-default
348867
This is an example QoS ingress marking policy based on the Modular QoS CLI (MQC). Refer
to your specific router configuration guide for information on how to achieve a similar policy on
a Cisco router supporting MQC and for any updated commands.
The traffic goes to an outbound interface to be queued and scheduled by an output service policy
that has 3 queues created: a Priority Queue called VOICE, a CBWFQ called VIDEO, and another
CBWFQ called SIGNALING (Figure 13-72). This highlights the fact that this egress queuing policy
is based only on DSCP as network marking occurring at the access switch and/or on ingress into the
WAN router ingress interface. This is an example simply to illustrate the matching criteria and
queues, and it does yet not contain the WRED functionality (covered in the next subsection). For
more information on WRED, see the next section on WAN Queuing and Scheduling, page 13-100.
13-98
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Figure 13-72
UDP Ports
DSCP
3xxx
AF41
4xxx
AF42
ACL
CS3
WAN
5060
7
Egress
Policy
EF
Trust
348868
AF41
CS3
7.
Figure 13-73
The traffic is matched against the class-map match statements, and all traffic marked EF goes to the
VOICE PQ, AF41 and AF42 traffic goes to the VIDEO CBWFQ, and CS3 traffic goes to the
SIGNALING CBWFQ (Figure 13-73).
Egress
Policy
policy-map EGRESS-QUEUING
Note
This is an example egress queuing policy based on the Cisco Common Classification Policy
Language (C3PL). Refer to your specific router configuration guide for information on how to
achieve a similar policy on a Cisco router supporting C3PL and for any updated commands.
13-99
Chapter 13
Bandwidth Management
Voice-only calls from trusted endpoints (EF) are mapped to the PQ.
EF
AF41
Video of Video
AF41
Audio of Jabber
AF41
Video of Jabber
AF42
Video
CBWFQ
other queues
Audio of Video
PQ
348870
EF
Audio of IP
Phone
Weighted Random Early Detection (WRED) minimum and maximum thresholds are also configured in
the Video CBWFQ. To illustrate how the WRED thresholds are configured, assume that the interface had
been configured with a queue depth of 256 packets. Then following the guidelines listed above, the
WRED minimum and maximum thresholds for AF42 and AF41 would be configured as illustrated in
Figure 13-75.
13-100
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Figure 13-75
Drop Probability
100%
policy-map WAN-EDGE
class VIDEO
bandwidth percent 40
random-detect dscp-based
random-detect dscp af41 120 256 50
random-detect dscp af42 30 80 20
mark-probability-denominator
5%
AF42
WRED
mark
probability
2%
AF41
(100/denom)
30
80
256
120
WRED
min threshold
WRED
max threshold
13-101
Chapter 13
Bandwidth Management
Figure 13-76
Admission Control
Voice
V
Vo
oiice
Queue
Queue
Video
Queue
WAN
Link
other
queues
EF
AF42
348872
AF41
Priority queue is provisioned for voice calls from trusted endpoints and is protected by admission
control (ELCAC voice bandwidth pool).
13-102
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Figure 13-77
WAN Link Speed
155 Mbps
(OC3)
34-44 Mbps
(E3/DS3)
10 Mbps
5 Mbps
<2 Mbps
(T1/E1)
Control (%)
10
Voice (%)
10
10
10
20
25
30
Video (%)
43
53
55
40
30
--
Signalling (%)
Scavenger (%)
Default (%)
43
33
31
36
40
54
348873
Class
The following sections cover each site (Central, Large Branch, Small Branch, Micro Branch) and the
link bandwidth provisioned for each class based on the number of users and available bandwidth for each
class. Keep in mind that these values are based on bandwidth calculated for Layer 3 and above.
Therefore, they do not include the Layer 2 overhead, which is dependent on the link type (Ethernet,
Frame-relay, MPLS, and so forth). See the chapter on Network Infrastructure, page 3-1, for more
information on Layer 2 overhead.
Central Site Link (100 Mbps) Bandwidth Calculation
As illustrated in Figure 13-78, the Central Site has the following bandwidth requirements:
Immersive endpoints are sized for the busy hour. One endpoint is expected to be in a call across the
WAN. This would be for a point-to-point call, since any conference call would terminate locally at the
TelePresence server. It is important to take into account the worst-case scenario for the busy hour.
13-103
Chapter 13
Bandwidth Management
Video endpoints are sized for 20% WAN utilization (0.2). A possible total of 30 calls at 1.2 Mbps is
based on the number of endpoints. But assuming only 20% WAN utilization in active calls over the
WAN, compared to active local calls, gives the WAN utilization rate of above 7.2 Mbps.
TelePresence Servers are sized at an average bit rate of 1.5 Mbps to account for the average of various
endpoint resolutions from remote sites. The TelePresence Server would then be able to support up to 40
calls total (local and remote), and this is multiplied by 50% (0.5) to account for the possibility of half of
the TelePresence calls going over the WAN while the other half might be serving local endpoints.
In addition there is 15.8 Mbps for Jabber calls, which could be 18 calls at 576p, or 50 calls at 288p, or
variations thereof. This gives an idea of what the Jabber video calls have available for bandwidth. When
more Jabber video calls occur past the 15.8 Mbps, packet loss will occur and will force all Jabber clients
to adjust their bit rates down. This can be either a very subtle process with no visible user experience
implications if the loss rate is low as new calls are added, or it can be very disruptive to the Jabber video
if there is an immediate and sudden loss of packets. The expected packet loss rate as new video calls are
added is helpful in determining the level of disruption in the user experience for this opportunistic class
of video.
Figure 13-78
Central Site
500 users:
500 Jabber clients
300 IP phones
30 video endpoints
(room + desktop)
1 immersive telepresence
2 Telepresence Servers
Voice
10 Mbps
Video
55 Mbps
100
Mbps
voice
10%
Default
31 Mbps
348874
video
55%
WAN
Link
13-104
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
As illustrated in Figure 13-79, the Large Branch site has the following bandwidth requirements:
Large Branch
50 users:
50 Jabber clients
30 IP phones
6 video endpoints
(room + desktop)
Voice
3.4 Mbps
Video
18.7 Mbps
34
Mbps
voice
10%
Default
10.5 Mbps
348875
video
55%
WAN
Link
As illustrated in Figure 13-80, the Small Branch site has the following bandwidth requirements:
13-105
Chapter 13
Bandwidth Management
Figure 13-80
Small Branch
15 users:
15 Jabber clients
10 IP phones
2 video endpoints
(room)
Voice
2 Mbps
Video
4 Mbps
WAN
Link
10
Mbps
voice
20%
Default
3.6 Mbps
348876
video
40%
As illustrated in Figure 13-81, the Micro Branch site has the following bandwidth requirements:
Asymmetric download/upload broadband: consider limiting transmit bit rate on video endpoint
Micro Branch
5 users:
5 Jabber clients
3 IP phones
1 video endpoint
Inte
Internet
e
DMVPN
Tunnel
QoS-enabled
Router
Voice
1.25 Mbps
Video
1.5 Mbps
WAN
Link
5
Mbps
Default
2 Mbps
Broadband
Modem/Router
voice
25%
video
30%
348877
Figure 13-81
13-106
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Large Branch with Constrained WAN Link (Enhanced Locations CAC Enabled for Video)
In specific branch sites with lower-speed WAN links, over-provisioning the video queue is not feasible
(see Figure 13-82). ELCAC can be applied to these Location links for video to ensure that video calls
do not over-subscribe the link bandwidth. This template requires using site-specific region configuration
to limit maximum bandwidth used by video endpoints and Jabber clients. Also keep in mind that device
mobility is required if Jabber users roam across sites.
Note
Bandwidth for voice-only Jabber calls is subtracted from "voice" ELCAC, but it impacts the video queue
(since it is marked AF41). Adjust the delta between video ELCAC bandwidth and video queue size.
Figure 13-82
Large Branch with Constrained WAN Link (Enhanced Locations CAC Enabled for
Video)
50 users:
50 Jabber clients
30 IP phones
4 video endpoints
(room + desktop)
Voice
2 Mbps
Video
4 Mbps
WAN
Link
10
Mbps
Default
3.6 Mbps
video
40%
348878
voice
20%
As illustrated in Figure 13-82, a Large Branch site with a constrained WAN link (10 Mbps) has the
following bandwidth requirements:
13-107
Chapter 13
Bandwidth Management
Example Enterprise #2
Example Enterprise #2 is a large enterprise with users across a large geographic area, with a data center
(DC) at the headquarters site as well as multiple large, small, and micro-sized branches with roughly
500, 50, 15, and 5 users in each branch type, respectively. To simplify the illustration of the network,
these categories of sites (HQ, large, small, and micro) are used as a template to size bandwidth
considerations for each site that has a similar user base and endpoint density. Figure 13-83 illustrates
each type of site. The enterprise has deployed Jabber with video to ensure that users have access to a
video terminal for conferencing. The TelePresence video conferencing resources are located in the DC
at HQ. IP phones are for voice-only communications; video endpoints are Jabber clients, Collaboration
desktop endpoints (DX Series), and room endpoints (MX, Profile, and SX Series); and the HQ and large
sites have immersive TelePresence units such as the IX Series.
Note
Figure 13-83
Example Enterprise #2 is markedly different from Example Enterprise #1 in the sense that all endpoints
(trusted and untrusted) in Example Enterprise #2 are configured to mark EF for all audio (voice-only and
video calls) and mark video AF41 or AF42 for Jabber video. Also, Example Enterprise #2 uses Enhanced
Locations CAC to protect the voice queue for the audio portion. Cisco Collaboration System Release
(CSR) 11.x provides a new feature whereby all audio can be deducted from the video pool. See the
section on Enhanced Locations Call Admission Control, page 13-39, for more information.
Example Enterprise #2
Unified CM
Multipoint
Video (TS)
500 users:
Central
Site
Expressway
B2B
Cloud
Services
S
erv
r ices
s
Mobile
Users
Internet
Int
MPLS VPN
15 users:
5 users:
50 Jabber clients
30 IP phones
6 video endpoints
(room + desktop)
15 Jabber clients
10 IP phones
2 video endpoints
(room)
5 Jabber clients
3 IP phones
1 video endpoint
Large Branch
Small
S
Branch
Micro Branch
348861
50 users:
The IT department is tasked with determining the bandwidth requirements for the WAN edge for each
type of site in Example Enterprise #2. The following sections list the requirements and illustrate a
methodology for applying QoS, determining bandwidth and queuing requirements, and determining
admission control requirements.
13-108
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Jabber clients
and desktop):
((mobile
ob e a
DSCP
UDP Ports
3xxx
ACL
EF
5xxx
AF42
5061
CS3
348879
Figure 13-84
The administrator creates an ACL for the access switches for the data VLAN to re-mark UDP ports to
the following DSCP values:
Audio streams of all Jabber calls (voice-only and video) are marked EF.
For the Jabber endpoints we also recommend changing the default QoS values in the Jabber SIP profile.
This is to ensure that, if for any reason the QoS is "trusted" via a wireless route or any other way, the
correct "trusted" values will be the same as they would be for the re-marked value. Therefore, the QoS
parameters in the SIP Profile need to be set as shown in Table 13-16.
Table 13-16
Changed Value
EF
AF41
AF42
AF41
EF
CS4
AF41
CS4
EF
13-109
Chapter 13
Bandwidth Management
The configuration in Table 13-16 ensures that audio of Jabber clients will be set to EF and the video will
be set to AF42, if for any reason they are trusted and not re-marked via UDP port range at the access
switch. This is simply to ensure a consistent configuration across Jabber endpoints.
Trusted Endpoints
For the trusted endpoints, Cisco Discovery Protocol (CDP) is used and the QoS of the IP phones and
video endpoint is trusted using the conditional trust mechanism configured at the access switch. The
defaults need to be changed to ensure that all audio is set to EF for all endpoints. In this case Unified CM
is configured with a SIP Profile that changes the audio of video and TelePresence calls to EF
respectively.
Figure 13-85 illustrates the conditional trust (CDP based) and packet marking at the access switch.
Figure 13-85
DSCP
Trusted endpoints:
EF
T t
Trust
AF41
CS3
DSCP
EF
CS3
348880
T t
Trust
The administrator configures all access switches with a conditional QoS trust for IP phones and video
and TelePresence endpoints, classified as follows:
The administrator configures the Unified CM SIP Profile for trusted endpoints with the DSCP values
listed in Table 13-17.
13-110
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Table 13-17
Changed Value
EF
AF41
AF41
EF
CS4
AF41
CS4
EF
At the WAN edge, on ingress it is expected that the packets arriving with a specific DSCP value have
been trusted at the access layer or re-marked accordingly if they were not trusted at the access switch.
As a failsafe practice, on ingress it is important to re-mark any untrusted traffic at the WAN edge that
could not be re-marked at the access layer. While QoS is important in the LAN, it is paramount in the
WAN; and as routers assume a trust on ingress traffic, it is important to configure the correct QoS policy
that aligns with the business requirements and user experience. The WAN edge re-marking is always
done on the ingress interface into the router, while the queuing and scheduling is done on the egress
interface. The following example walks through the WAN ingress QoS policy as well as the egress
queuing policy. Figure 13-86 illustrates the configuration and the re-marking process.
In Figure 13-86 the packets from both the trusted and untrusted areas of the network are identified and
classified with the appropriate DSCP marking via the trust methods discussed or via a simple ACL
matching on UDP port ranges. Keep in mind that this ACL could also match more granularly on IP
address or some other attributes that would further limit the scope of the marking.
Figure 13-86
UDP Ports
DSCP
3xxx
EF
5xxx
AF42
5061
ACL
Trust
Ingress
Policy
CS3
EF
CS3
348881
AF41
Figure 13-87 through illustrate the policy matching criteria and DSCP re-marking. The process involves
the following steps shown in the figures:
1.
Packets arrive into the router ingress interface, which is configured with an input service policy
(Figure 13-87).
13-111
Chapter 13
Bandwidth Management
Figure 13-87
2.
The policy-map is configured with 4 classes of traffic setting the appropriate DSCP: Voice setting a
DSCP of EF, Prioritized-Video setting a DSCP of AF41, Jabber-Video setting a DSCP of AF42, and
Signaling setting a DSCP of CS3 (Figure 13-87).
3.
Each one of these classes matches a class-map of the same name configured with match-any criteria
and a DSCP match as well as an ACL match. This match-any criteria means that the process will
start top-down, and the first matching criteria will be executed and thus set the DSCP according to
each class in the policy-map statements. Another option is match-all, which would require all
criteria to be matched and thus would match DSCP and ACL. This, however, would not provide the
intended functionality of re-marking either marked or unmarked traffic.
Ingress
Policy
4.
policy-map INGRESS-MARKING
class VOICE
set dscp ef
class PRIORITIZED-VIDEO
set dscp af41
class JABBER-VIDEO
set dscp af42
class SIGNALING-SIP
set dscp cs3
class class-default
348882
The first match statement in the class-map is match dscp. If the traffic matches the DSCP, then
DSCP is set again to the same value that was matched, as is configured in the policy map statements
(Figure 13-88).
13-112
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Figure 13-88
Ingress
Policy
5.
policy-map INGRESS-MARKING
class VOICE
set dscp ef
class PRIORITIZED-VIDEO
set dscp af41
class JABBER-VIDEO
set dscp af42
class SIGNALING-SIP
set dscp cs3
348883
class class-default
If DSCP was not matched, then the next line in the class-map statement is parsed, which is the ACL
that matches the UDP ports set in Unified CM for the Jabber clients (see Identification and
Classification, page 13-17). When the ACL criteria (protocol and port range) are met, then the traffic
is set as is configured in the corresponding policy map statements (Figure 13-89).
13-113
Chapter 13
Bandwidth Management
Figure 13-89
Ingress
Policy
Note
6.
policy-map INGRESS-MARKING
class VOICE
set dscp ef
class PRIORITIZED-VIDEO
set dscp af41
Egress
Policy
class JABBER-VIDEO
set dscp af42
class SIGNALING-SIP
set dscp cs3
class class-default
348884
This is an example QoS ingress marking policy based on the Cisco Common Classification
Policy Language (C3PL). Refer to your specific router configuration guide for information on
how to achieve a similar policy on a Cisco router supporting C3PL and for any updated
commands.
The traffic goes to an outbound interface to be queued and scheduled by an output service policy
that has 3 queues created: a Priority Queue called VOICE, a CBWFQ called VIDEO, and another
CBWFQ called SIGNALING (Figure 13-90. This highlights the fact that this egress queuing policy
is based only on DSCP as network marking occurring at the access switch and/or on ingress into the
WAN router ingress interface. This is an example simply to illustrate the matching criteria and
queues, and it does not yet contain the WRED functionality (covered in the next subsection). For
more information on WRED, see the next section on WAN Queuing and Scheduling, page 13-116.
13-114
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Figure 13-90
UDP Ports
DSCP
3xxx
EF
5xxx
AF42
Egress
Policy
CS3
WAN
5061
ACL
EF
Trust
348885
AF41
CS3
7.
Figure 13-91
The traffic is matched against the class-map match statements, and all traffic marked EF goes to the
VOICE PQ, AF41 and AF42 traffic goes to the VIDEO CBWFQ, and CS3 traffic goes to the
SIGNALING CBWFQ (Figure 13-91).
Egress
Policy
13-115
Chapter 13
Bandwidth Management
Note
This is an example egress queuing policy based on the Cisco Common Classification Policy
Language (C3PL). Refer to your specific router configuration guide for information on how to
achieve a similar policy on a Cisco router supporting C3PL and for any updated commands.
All audio from all endpoints (trusted and untrusted) marked EF is mapped to the PQ.
EF
EF
Audio of Jabber
EF
Video of Video
AF41
Video of Jabber
AF42
Video
CBWFQ
other queues
EF
Audio of Video
PQ
348887
Audio of IP Phone
13-116
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Weighted Random Early Detection (WRED) minimum and maximum thresholds are also configured in
the Video CBWFQ. To illustrate how the WRED thresholds are configured, assume that the interface had
been configured with a queue depth of 256 packets. Then following the guidelines listed above, the
WRED minimum and maximum thresholds for AF42 and AF41 would be configured as illustrated in
Figure 13-93.
Figure 13-93
Drop Probability
100%
policy-map WAN-EDGE
class VIDEO
bandwidth percent 40
random-detect dscp-based
random-detect dscp af41 120 256 50
random-detect dscp af42 30 80 20
mark-probability-denominator
5%
AF42
WRED
mark
probability
2%
AF41
(100/denom)
30
80
256
120
WRED
min threshold
WRED
max threshold
13-117
Chapter 13
Bandwidth Management
Figure 13-94
Admission Control
Voice
Voi
Vo
ice
Queue
Queue
Video
Queue
WAN
Link
other
queues
voice-only call
EF
AF42
348888
AF41
Priority queue is provisioned for all calls from both trusted and untrusted endpoints, and it is
protected by admission control (E-LCAC voice BW pool).
13-118
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Figure 13-95
WAN Link Speed
155 Mbps
(OC3)
34-44 Mbps
(E3/DS3)
10 Mbps
5 Mbps
<2 Mbps
(T1/E1)
Control (%)
10
Voice (%)
10
10
10
20
25
30
Video (%)
43
53
55
40
30
--
Signalling (%)
Scavenger (%)
Default (%)
43
33
31
36
40
54
348873
Class
The following sections describe each site (Central, Large Branch, Small Branch, and Micro Branch) and
the link bandwidth provisioned for each class based on the number of users and available bandwidth for
each class. Keep in mind that these values are based on bandwidth calculated for Layer 3 and above.
Therefore, they do not include the Layer 2 overhead, which is dependent on the link type (Ethernet,
Frame-relay, MPLS, and so forth). See the chapter on Network Infrastructure, page 3-1, for more
information on Layer 2 overhead.
Also, note that the audio portion of bandwidth for video calls is now deducted from the voice pool. So
when provisioning the voice queue, this will include the audio bandwidth for both voice-only and video
calls. These examples are the same as those for Example Enterprise #1, page 13-91. The only difference
is that for Example Enterprise #2 the audio portion of bandwidth for video calls is deducted from the
voice admission control pool, and the audio streams go into the voice queue.
Central Site Link (100 Mbps) Bandwidth Calculation
As illustrated in Figure 13-96, the Central Site has the following bandwidth requirements:
13-119
Chapter 13
Bandwidth Management
Calculation Notes
Immersive endpoints are sized for the busy hour. One endpoint is expected to be in a call across the
WAN. This would be for a point-to-point call, since any conference call would terminate locally at the
TelePresence server. It is important to take into account the worst-case scenario for the busy hour.
Video endpoints are sized for 20% WAN utilization (0.2). A possible total of 30 calls at 1.2 Mbps is
based on the number of endpoints. But assuming only 20% WAN utilization in active calls over the
WAN, compared to active local calls, gives the WAN utilization rate of above 7.2 Mbps.
TelePresence Servers are sized at an average bit rate of 1.5 Mbps to account for the average of various
endpoint resolutions from remote sites. The TelePresence Server would then be able to support up to 40
calls total (local and remote), and this is multiplied by 50% (0.5) to account for the possibility of half of
the TelePresence calls going over the WAN while the other half might be serving local endpoints.
In addition there is 15.8 Mbps for Jabber calls, which could be 18 calls at 576p, or 50 calls at 288p, or
variations thereof. This gives an idea of what the Jabber video calls have available for bandwidth. When
more Jabber video calls occur past the 15.8 Mbps, packet loss will occur and will force all Jabber clients
to adjust their bit rates down. This can be either a very subtle process with no visible user experience
implications if the loss rate is low as new calls are added, or it can be very disruptive to the Jabber video
if there is an immediate and sudden loss of packets. The expected packet loss rate as new video calls are
added is helpful in determining the level of disruption in the user experience for this opportunistic class
of video.
Figure 13-96
Central Site
500 users:
500 Jabber clients
300 IP phones
30 video endpoints
(room + desktop)
1 immersive telepresence
2 Telepresence Servers
Voice
10 Mbps
Video
55 Mbps
100
Mbps
voice
10%
Default
31 Mbps
348874
video
55%
WAN
Link
13-120
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
As illustrated in Figure 13-97, the Large Branch site has the following bandwidth requirements:
Large Branch
50 users:
50 Jabber clients
30 IP phones
6 video endpoints
(room + desktop)
Voice
3.4 Mbps
Video
18.7 Mbps
34
Mbps
voice
10%
Default
10.5 Mbps
348875
video
55%
WAN
Link
As illustrated in Figure 13-98, the Small Branch site has the following bandwidth requirements:
13-121
Chapter 13
Bandwidth Management
Figure 13-98
Small Branch
15 users:
15 Jabber clients
10 IP phones
2 video endpoints
(room)
Voice
2 Mbps
Video
4 Mbps
WAN
Link
10
Mbps
voice
20%
Default
3.6 Mbps
348876
video
40%
As illustrated in Figure 13-99, the Micro Branch site has the following bandwidth requirements:
Asymmetric download/upload broadband: consider limiting transmit bit rate on video endpoint
Micro Branch
5 users:
5 Jabber clients
3 IP phones
1 video endpoint
Inte
Internet
e
DMVPN
Tunnel
QoS-enabled
Router
Voice
1.25 Mbps
Video
1.5 Mbps
WAN
Link
5
Mbps
Default
2 Mbps
Broadband
Modem/Router
voice
25%
video
30%
348877
Figure 13-99
13-122
Chapter 13
Bandwidth Management
Bandwidth Management Design Examples
Large Branch with Constrained WAN Link (Enhanced Locations CAC Enabled for Video)
In specific branch sites with lower-speed WAN links, over-provisioning the video queue is not feasible
(see Figure 13-100). ELCAC can be applied to these Location links for video to ensure that video calls
do not over-subscribe the link bandwidth. This template requires using site-specific region configuration
to limit maximum bandwidth used by video endpoints and Jabber clients. Also keep in mind that device
mobility is required if Jabber users roam across sites.
Note
Because audio bandwidth for both voice-only and video calls is deducted from the voice CAC pool, there
is no need for any queue bandwidth adjustment as is the case in Example Enterprise #1.
Figure 13-100
Large Branch with Constrained WAN Link (Enhanced Locations CAC Enabled for
Video)
50 users:
50 Jabber clients
30 IP phones
4 video endpoints
(room + desktop)
Voice
2 Mbps
Video
4 Mbps
WAN
Link
10
Mbps
Default
3.6 Mbps
video
40%
348878
voice
20%
As illustrated in Figure 13-100, a Large Branch site with a constrained WAN link (10 Mbps) has the
following bandwidth requirements:
13-123
Chapter 13
Bandwidth Management
13-124
CH A P T E R
14
Dial Plan
Revised: June 15, 2015
The dial plan is one of the key elements of a Unified Communications and Collaboration system, and an
integral part of all call processing agents. Generally, the dial plan is responsible for instructing the call
processing agent on how to route calls. Specifically, the dial plan performs the following main functions:
Endpoint addressing
For destinations registered with the call processing agent, addresses are assigned to provide
reachability. These internal destinations include all endpoints (such as IP phones, video endpoints,
soft clients and analog endpoints) and applications (such as voicemail systems, auto attendants, and
conferencing systems).
Path selection
Depending on the calling device and the destination dialed, a path to the dialed destination is
selected. If a secondary path is available, this path will also be considered if the primary path fails.
Calling privileges
Different groups of devices can be assigned to different classes of service, by granting or denying
access to certain destinations. For example, lobby phones might be allowed to reach only internal
and local PSTN destinations, while executive phones could have unrestricted PSTN access.
14-1
Chapter 14
Dial Plan
This chapter presents information intended to guide the system designer toward a dial plan that
accommodates the legacy dialing habits of telephony and video users, while also taking advantage of
new functionality afforded by the increasing integration between computing technology and telephony,
such as dialing from contacts, click-to-call actions from computers and smart phones, and adoption of
mobility-related features. The chapter is structured to offer information about the following main areas:
For more details, refer to the System Configuration Guide for Cisco Unified Communications Manager,
the Feature Configuration Guide for Cisco Unified Communications Manager, the Cisco IOS Voice and
Video Configuration guides, and other product documentation available at
http://www.cisco.com
Described in
Revision Date
14-2
Chapter 14
Dial Plan
Dial Plan Fundamentals
Endpoint Addressing
Reachability of endpoints registered to a call processing agent, users, and applications is provided by
addresses assigned to these addressable entities. In enterprise collaboration networks we differentiate
between numeric addresses and alphanumeric uniform resource identifiers (URIs).
NSN
Maximum of 15-n digits (n = number of CC digits)
CC
1 to 3 digits
NDC
Defined by National Numbering Plan
SN
Defined by National Numbering Plan
Maximum of 15 digits
The following definitions apply to Figure 14-1:
CC = Country Code
SN = Subscriber Number
According to ITU recommendation E.164, the maximum length of any phone number is 15 digits. The
first part of a geographic E.164 number is the country code. Country codes are between one and three
digits long (country code 1 and 7 are the only single-digit country codes). The remainder of a geographic
E.164 number is the national significant number (NSN). The general structure of a NSN is that the first
few digits represent a national destination code (NDC), or area code, and the last digits represent the
14-3
Chapter 14
Dial Plan
subscriber number. ITU recommendation E.164 does not define national numbering plans and thus does
not prescribe the schema to be used for NSNs in specific countries. This is left to the national numbering
plan authorities. A collection of documentation on various national numbering plans can be found at
http://www.itu.int/oth/T0202.aspx?parent=T0202
National numbering plans can be very different in structure. As an example, Table 14-2 compares the
numbering plans used in the US and in Germany.
Table 14-2
Country Code
NSN Length
NDC Length
SN Length
1 (US)
10
49 (Germany)
3 to 13
2 to 5
ITU recommendation E.164 also mentions that a leading "+" should be used to indicate if an
international prefix is required. Throughout this design guide we consistently use the term "E.164" to
refer to E.164 numbers and "+E.164" to refer to E.164 numbers with a leading "+".
Using +E.164 numbers as numeric addresses has the benefit that these numbers by definition are unique
and that it is very easy to map between +E.164 and any habitual dialing that might be required to be
supported by an enterprise dial plan.
As an alternative to using unique numeric +E.164 addresses, numeric addresses following an enterprise
numbering plan may also be used. Building an enterprise address plan or numbering scheme in multi-site
deployments involves the definition of a typical hierarchical addressing structure with the following
characteristics:
Provides unique numeric addresses for all endpoints, users, and applications in the enterprise.
Needs to be extensible so that the numbering scheme allows for adding new sites without having to
redesign the whole numbering scheme, which would involve address reassignments for existing
endpoints, users, and applications.
In a typical enterprise numbering plan, numerical addresses would consist of a site code and a site
subscriber number. When designing an enterprise numbering plan, reserve enough digits for both the site
code and the site subscriber number to make sure that additional sites can be added if required and
enough subscriber numbers can be defined per site. Enterprise numbering plans typically are designed
to be fixed length.
If or when dialing of addresses defined by an enterprise numbering plan needs to be supported directly
as a dialing habit, typically a single-digit access code is selected and prefixed to the enterprise number.
In that case a full enterprise numbering address would have three pieces: enterprise address access code
(for example, 8), site code (for example, 496), and site subscriber number (for example, 9123); or for
example, 8-496-9123.
The enterprise address access code in this case needs to be selected so that it does not cause overlap with
any other dialing habit (see Dialing Habits, page 14-6).
To be able to uniquely identify an addressable entity, either all addresses have to be unique or they at
least have to be unique within a defined sub-domain managed by the call processing agent. If two distinct
entities need to be addressed using the same address, then the two identical addresses have to reside in
disjunct addressing domains that are managed independently. With numeric addresses, this situation can
occur if site-significant numeric addresses (for example, a four-digit extension) are used and two
endpoints with the same site-significant address (same four-digit extension) in different sites need to be
addressed by the same call control agent. Table 14-3 shows an example of this situation.
14-4
Chapter 14
Dial Plan
Dial Plan Fundamentals
Table 14-3
+E.164 Number
Site (Sub-Domain)
4-Digit DN (Address)
Frankfurt
3001
San Francisco
3001
In Table 14-3, two E.164 numbers result in the same site-specific four-digit directory number based on
the respective site's DID range. This implies that the four-digit DNs cannot be used as numeric endpoint
addresses directly.
Addresses following an enterprise numbering plan, also known as enterprise significant numbers (ESN),
can be used to address destinations for which no PSTN numbers (E.164 numbers) exist. These
destinations include:
Lobby phones
Services (call pickup numbers, call park numbers, conferences, and so forth)
Alphanumeric Addresses
Alphanumeric addresses based on SIP URIs can also be used to address endpoints, users, and
applications. A commonly used scheme for alphanumeric addresses is simplified SIP URIs of the form
user@host, where the left-hand side (LHS, user portion) can be alphanumeric and the right-hand side
(RHS, host portion) is a domain name. The following examples represent valid alphanumeric addresses
based on SIP URIs:
bob@example.org
bob.home-office@example.org
bob@de.eu.example.org
bob.ex60@example.org
bob.vmbox@example.org
voicemail@de.eu.example.org
All of these URIs can serve as individual alphanumeric addresses for individual endpoints, users, and
applications. From the addressing perspective, any hierarchy implied by using dot-separated identifiers
(bob.ex60, de.eu.example.org) does not have any impact on the decision of whether any two URIs are
considered to be equivalent.
According to RFC 3261, comparison of the LHS of SIP URIs has to be case-sensitive, while the RHS
has to be compared case-insensitive. According to this standardized behavior, Alice@example.com and
alice@example.com are not to be considered equivalent and thus represent distinct individual addresses.
In reality, using addressing schemes for endpoints, users, and applications that rely on case sensitivity
of the user portion is considered bad practice because it leads to increased troubleshooting complexity.
Also keep in mind that RFC 2543 (the RFC specification preceding RFC 3261) explicitly defined that
SIP URIs (host and user portion) are case-insensitive. Different behaviors regarding the case sensitivity
of URI equivalence are common. To avoid problems, Cisco highly recommends always using only all
lowercase URIs as alphanumeric addresses.
The URI lookup policy of Unified CM can be configured to be case-sensitive (default) or
case-insensitive by configuring the enterprise parameter URI Lookup Policy accordingly.
14-5
Chapter 14
Dial Plan
Dialing Habits
Dialing destinations such as users, endpoints, and applications can be done using various formats. The
numeric +E.164 address +496907739001, for example, could be dialed as any of the following:
+496907739001
These examples show that dial strings typically are interpreted in a context, and with the exception of
dialing a +E.164 address directly, only the combination of dialed string and context provides proper
identification of the intended destination address.
The term "dialing habit" is commonly used to refer to a given dialing behavior used to reach a given set
of destinations. Examples of some dialing habits include:
9-0-1-1 plus E.164 for international destinations dialed from within the US
A dialing habit is described by specifying both the format of the string to be dialed (dialing structure)
and the destination address class to be reached. Examples of destination address classes typically used
in enterprise dial plans include:
On-net/intra-site
On-net/inter-site
Off-net/local
Off-net/national
Off-net/international
Off-net/emergency
Identifying the dialing habits that need to be supported by the dial plan is one of the first steps when
starting the design of an enterprise dial plan. It is essential to start the dial plan design with the full view
of all dialing habits to be supported, because the dialing habits need to be defined so that there is no
overlap between any two dialing habits to be supported for any given caller. Violating this rule will lead
to bad user experience because the call control cannot deterministically differentiate between
overlapping dialing habits by analyzing the dialed digits as they are dialed one-by-one. This ultimately
leads to inter-digit timeouts.
With alphanumeric dialing we typically differentiate only between fully qualified addresses and
non-fully qualified addresses. Fully qualified addresses contain the user and the host portion of a SIP
URI, whereas a non-fully qualified alphanumeric address represents only the user portion of the address
and the host portion needs to be derived from the dialing context of the calling party. For example,
dialing "bob" would be equivalent to dialing "bob@example.com" if the dialing context of the calling
party defines that "@example.com" should be appended to all non-fully qualified alphanumeric
addresses.
14-6
Chapter 14
Dial Plan
Dial Plan Fundamentals
Dialing Domains
As described in the previous section, a given destination might be dialed using different strings by
different users. A dialing domain specifies a group of users or devices sharing the same set of dialing
habits (dialing the same strings to reach identical destinations). The concept of dialing domains is
important because an enterprise dial plan has to implement the same treatment for each dialing domain.
All users belonging to any given dialing domain share the same dialing treatment.
To identify dialing domains, it is important to take all dialing habits into consideration. Users in two sites
in the US, even though they share the same PSTN dialing habits, would still belong to different dialing
domains if we also take into account how on-net calls are placed. In a typical environment, an on-net
intra-site call could be supported by dialing four digits, while an on-net inter-site call would be placed
by using a dial string equivalent to the PSTN dialing habit (forced on-net would still keep the call
on-net).
Figure 14-2 shows an example for this. Although endpoint 1234 in site SJC and endpoint 2001 in site
RTP share the same dialing habit for national destinations (dialing 91212555600 to reach PSTN
destination +1 212 555 6000) and international destinations (dialing 901149890123456 to reach PSTN
destination +49 890 123456), the dialing habit to reach endpoint 1001 in site SJC is different for
endpoints in RTP than for those in SJC: endpoint 1234 in site SJC would dial 1001 while endpoint 2001
in site RTP would need to dial 914085551001. In this example, endpoints in site RTP and site SJC would
belong to different dialing domains.
Figure 14-2
Dialing Domains
PSTN
914085551001
1001
2001
Site RTP: +1 919 555 2XXX
348604
1001
1234
14-7
Chapter 14
Dial Plan
Classes of Service
Not all users, applications, and endpoints in an enterprise are allowed to reach the same set of
destinations. Reasons for restricting the set of reachable destinations include cost avoidance, security
considerations, and privacy. As examples, not all users might be able to place international calls, the
voicemail systems might not be able to call any PSTN destination to avoid toll fraud, and only a very
limited set of users might be allowed to place direct calls to the CEO of a company. The term generally
used to refer to any given set of a restrictions or class of allowed destinations is class of service, or CoS.
Requirements for cost-driven classes of service heavily depend on the phone tariffs and the cost structure
associated with them. With voice services becoming cheaper (or being available for free), the trade-off
between the increased complexity associated with maintaining more classes of service and the potential
savings in call costs is changing. In certain cases, for example, it might not make sense any more to
differentiate between local and national calling if both call types are billed exactly the same.
The definition of a class of service might also be based on time schedules. Access to certain destinations
might be allowed only at certain times.
To reduce the complexity of an enterprise dial plan, Cisco recommends minimizing the number of
differentiated classes of service. This can be achieved by either removing classes of service with little
or no value (for example, differentiation between classes of service "national" and "local" even though
national calls are essentially for free) or by combining (almost) equivalent classes of service into a single
class of service.
Independent of restricting the access of certain users, devices, or applications to certain call types that
incur costs, typically access to emergency services (911, 112, and so forth) has to be provided to all users
at all times. Therefore, all classes of service have to allow access to emergency services at all times.
Call Routing
Routing calls involves several aspects:
Identifying the dialing habit based on the structure of the dial string.
Allowing/blocking the call based on the class of service of the calling entity.
Selecting a route to the called destination, establishing the call, and presenting the identity of the
parties involved in the expected format. Route selection also involves selecting alternate routes if
the primary route is not available for some reason.
An end-to-end enterprise dial plan needs to consider all of these aspects and is not limited only to
establishing a route between the calling and called entities.
14-8
Chapter 14
Dial Plan
Dial Plan Fundamentals
The case of numeric dialing needs a little more attention, especially if the dialed digits are entered
digit-by-digit. In this case the call control receives the destination from the calling entity digit-by-digit,
and part of selecting the correct route to the destination is to determine the exact time when enough digits
have been received and the call can be routed without having to wait for expiration of an inter-digit
timeout.
Figure 14-3 shows some typical US dialing habits for PSTN and emergency calls. Although all of these
dialing habits share the identical initial digit 9, international dialing and the first emergency dialing
string can easily be distinguished as soon as the second digit (0 or 9) is dialed. As soon as the third digit
is dialed, dialing 911 and dialing a national destination do not overlap any more because the North
American Numbering Plan (NANP) does not allow any NPA codes (numbering plan area codes) starting
with 1.
Typical US Dialing Habits for PSTN and Emergency Calls
International: 9
<E.164>
Emergency: 9
Emergency: 9
National: 9
<10 digits>
348605
Figure 14-3
Given the PSTN dialing habit in Figure 14-3. four-digit intra-site dialing for extensions starting with 9
must be avoided because this could potentially create partial overlap. For example, extension 9113
would overlap with emergency calling, and after receiving 911 the call control would have to wait to
determine whether the caller is going to continue to dial 3 (extension 9113) or whether dialing actually
was complete after 911. This would delay all emergency calls! Similarly, extensions such as 9140 would
create overlaps with national PSTN calls, and calls to those extensions would be delayed.
To minimize overlaps, the first digit of a dialing habit can be defined as an access code uniquely
identifying a class of destinations. The PSTN or trunk access code is a perfect example for this scheme.
The most commonly used trunk access codes are 9 (US, UK, and others) and 0 (commonly used in most
European countries).
As mentioned earlier, selecting non-overlapping dialing habits is key to avoiding bad user experience
due to inter-digit time-outs. Typical overlaps seen in enterprise dial plans include:
10-digit dialing with abbreviated intra-site dialing (for example, four digits)
NPA codes in the US can start with any digit other than 0 or 1, which means that the first few digits
of 10-digit dialing would overlap with almost any abbreviated intra-site dialing.
PSTN access code (such as 9 in the US) with abbreviated intra-site dialing
A PSTN access code of 9 will overlap with all abbreviated intra-site dialing to extensions starting
with 9.
14-9
Chapter 14
Dial Plan
Access to features such as call park numbers and voicemail also requires mapping into the set of dialing
habits defined. Dialing these features should typically require only a short dialing sequence. To achieve
this, either the feature access codes can be mapped into the abbreviated intra-site dialing or a dedicated
feature access code needs to be defined.
1234
+1 408 555 2345
+1 408 555 1XXX
914085552345
914085551001
1001
PSTN
912125555001
348606
5001
Forced on-net routing is especially important if dialing of +E.164 destinations from directories is
implemented. In a normalized directory, all destinations are defined as +E.164 numbers, regardless of
whether the person that the number is associated with is internal or external. In this case forced on-net
routing is a mandatory requirement to avoid charges caused by internal calls routed through the PSTN.
14-10
Chapter 14
Dial Plan
Dial Plan Fundamentals
Call initiator
Dialed destination
Resource availability
Resource prioritization
To be able to select an external connection based on the dialed destination, the dialed destination must
be classified. As explained earlier, E.164 numbers have a hierarchical structure that implies some
geographic association of numbers so that the egress connection selection could be based on a
prefix-based hierarchical routing scheme that tries to select an egress point "closest" (in the sense of the
geographic semantics of the E.164 number) to the destination. This behavior is called Tail End Hop Off
(TEHO). When implementing Tail End Hop Off, local legal regulations have to be considered.
An interesting special case of TEHO exists if strange phone tariffs allow for cheaper national calls (for
example, interstate) than local calls (in-state). In this case an egress point selection policy might be
implemented that actually tries to avoid selecting an egress connection "too" close to the dialed
destination. Decreasing phone charges make these kinds of routing schemes less and less useful.
In contrast to E.164 numbers, which have a clear hierarchical geographical structure with the most
significant information on the left alphanumeric, SIP URIs allow addressing hierarchy in the host portion
(RHS) of the URI. Depending on the domain name used as RHS, the addressing hierarchy of URIs does
not necessarily allow for geographic mapping a URI to a location in the routing topology, especially if
a flat URI scheme such as user@example.org is used. More interestingly, the most significant piece of
a SIP URI is the right-most piece of the host portion (top level domain).
14-11
Chapter 14
Dial Plan
but are internal connections to other enterprise call control entities. To make sure that only calls to
remote on-net destinations are routed to the remote call control, the call routing decision needs to be
based on the specific address ranges local to the remote call control.
Figure 14-5 shows why the prefix-based routing between independent call controls must be very
specific. In this example, to enable the left call control to decide wether 912125556001 needs to be
treated as an on-net call, the left call control has to have very specific numeric prefix routes for all
numeric addresses served by the right call control entity.
The maintenance of on-net prefix routes provisioned on each call control becomes more complex with
an increasing number of call controls involved and sites and DID ranges to be considered. Dynamic
learning of remote destinations helps to eliminate this complexity. Global Dial Plan Replication (GDPR)
is one example of an architecture that allows call controls to automatically learn about destinations
existing on remote call controls.
Figure 14-5
912125557001
+14085551XXX
+19195557XXX
+12125557XXX
+14045556XXX
PSTN
+1 404 555 6XXX
348607
The equivalence of prefix-based routing of numeric addresses for alphanumeric URIs is to use a domain
hierarchy and implement routing based on the host or domain portion of the URI. Figure 14-6 shows an
example of hierarchical routing with alpha URIs. In this example all three independent call controls use
a dedicated (sub) domain so that the on-net routing can easily be implemented based on this hierarchical
domain structure.
14-12
Chapter 14
Dial Plan
Dial Plan Elements
Figure 14-6
alice@sfo.example.com
bob@nyc.example.com
348608
carol@fra.example.com
carol@fra.example.com
In cases where the URI addressing scheme is not hierarchical, each call control has to have knowledge
of all URIs hosted on remote call controls. Global Dial Plan Replication (GDPR) offers a mechanism for
call controls to exchange information about URIs hosted on each call control to enable deterministic
routing even with a flat URI naming scheme.
14-13
Chapter 14
Dial Plan
Note
Different types of IP telephones accept keypad input and present visual information in different ways.
For purposes of this chapter only, we define the following phone types:
Type-A phones Include the Cisco Unified IP Phone 7905, 7912, 7940, and 7960.
Type-B phones Include the Cisco Unified IP Phone 6901, 6911, 6921, 6941, 6945, 6961, 7906,
7911, 7921, 7925, 7931, 7941, 7942, 7945, 7961, 7962, 7965, 7970, 7971, 7975, 8961, 9951, 9971,
and newer phones.
For calls originating from phones where the configured directory numbers are not in a globalized
(+E.164) form, the inbound calls calling party transformation CSS can be used to define the
appropriate globalization. This CSS can be found on the phone configuration page in the Number
Presentation Transformation section or in the Phone Settings section on the device pool
configuration page under Caller ID For Calls From This Phone.
For calls terminated on the phones, the outbound calls calling party transformation CSS can be used
to define the localization scheme to be applied to calling party numbers. This CSS can be found on
the phone configuration page in the Number Presentation Transformation section under Remote
Number.
For phones, outbound or remote number calling party transformations affect the number displayed while
the phone is ringing.
The outbound calls calling party transformation CSS (also referred to as localization or remote number
calling party transformation CSS) can also be used to localize remote connected party information. To
enable this, the advanced service parameter Apply Transformations On Remote Number must be
enabled.
Being able to provide localized connected party information to phones enables consistent remote party
information display on IP phones even if mid-call features are invoked.
14-14
Chapter 14
Dial Plan
Dial Plan Elements
A Type-B phone in New York receives a call from +1 408 526 4000. Calling party transformation
patterns are placed in the calling party transformation CSS in the phone's device pool. One of the patterns
is configured as: \+1.!, strip pre-dot.
As the call rings in, the called phone displays an incoming calling party number of 4085264000. After
the call is answered and released, the received-calls directory displays the last call as 408 526 4000, but
the number called when the user initiated the callback from the directory entry is +1 408 526 4000.
14-15
Chapter 14
Dial Plan
Dial Plan
(digit analysis)
IP
M
M
Dialing actions:
1000
Signaling
M
M
141878
Figure 14-7
It is neither required nor possible to configure dial plan information on IP phones running SCCP. All dial
plan functionality is contained in the Unified CM cluster, including the recognition of dialing patterns
as user input is collected.
If the user dials a pattern that is denied by Unified CM, reorder tone is played to the user as soon as that
pattern becomes the best match in Unified CM's digit analysis. For instance, if all calls to the
pay-per-minute Numbering Plan Area (or area code) 976 are denied, reorder tone would be sent to the
users phone as soon as the user dials 91976.
14-16
Chapter 14
Dial Plan
Dial Plan Elements
User Input and Feedback for Type-A SIP Phones with No Dial Rules Configured
Dial Plan
(digit analysis)
IP
M
M
Dialing actions:
1 0 0 0 Dial
141879
Figure 14-8
Signaling
If the user dials digits but then does not press the Dial softkey or the # key, the phone will wait for
inter-digit timeout (15 seconds by default) before sending a SIP INVITE message to Unified CM. For
the example in Figure 14-8, dialing 1, 0, 0, 0 and waiting for inter-digit timeout would result in the phone
placing a call to extension 1000 after 15 seconds.
Note
If the user presses the Redial softkey, the action is immediate; the user does not have to press the Dial
key or wait for inter-digit timeout.
If the user dials a pattern that is denied by Unified CM, the user must enter the entire pattern and press
the Dial key, and the INVITE message must be sent to Unified CM, before any indication that the call is
rejected (reorder tone) is sent to the caller. For instance, if all calls to NPA 976 are denied, the user would
have to dial 919765551234 and press Dial before the reorder tone would be played.
Dial Plan
(digit analysis)
IP
M
M
M
Dialing actions:
1000
Signaling
M
M
141880
Figure 14-9
Pattern 1...
Timeout 0
14-17
Chapter 14
Dial Plan
In Figure 14-9, the phone is configured to recognize all four-digit patterns beginning with 1 and has an
associated timeout value of 0. All user input actions matching the pattern will trigger the sending of the
SIP INVITE message to Unified CM immediately, without requiring the user to press the Dial key.
Type-A phones using SIP dial rules offer a way to dial patterns not explicitly configured on the phone.
If a dialed pattern does not match a SIP dial rule, the user can press the Dial key or wait for inter-digit
timeout.
If a particular pattern is recognized by the phone but blocked by Unified CM, the user must dial the entire
dial string before receiving an indication that the call is rejected by the system. For instance, if a SIP dial
rule is configured on the phone to recognize calls dialed in the form 919765551234 but such calls are
blocked by the Unified CM dial plan, the user will receive reorder tone at the end of dialing (after
pressing the final 4 key).
Dial Plan
(digit analysis)
IP
M
M
M
Dialing actions:
1000
Signaling
M
M
141881
Figure 14-10
Every user key press triggers a SIP NOTIFY message to Unified CM to report a KPML event
corresponding to the key pressed by the user. This messaging enables Unified CM's digit analysis to
recognize partial patterns as they are composed by the user and to provide the appropriate feedback, such
as immediate reorder tone if an invalid number is being dialed.
In contrast to Type-A IP phones running SIP without dial rules, Type-B SIP phones have no Dial key to
indicate the end of user input. In Figure 14-10, a user dialing 1000 would be provided call progress
indication (either ringback tone or reorder tone) after dialing the last 0 and without having to press the
Dial key. This behavior is consistent with the user interface on phones running the SCCP protocol.
14-18
Chapter 14
Dial Plan
Dial Plan Elements
Dial Plan
(digit analysis)
IP
M
M
M
Dialing actions:
1000
Signaling
M
M
141882
Figure 14-11
Pattern 1...
Timeout 0
In Figure 14-11, the phone is configured to recognize all four-digit patterns beginning with 1, and it has
an associated timeout value of 0. All user input actions matching these criteria will trigger the sending
of a SIP INVITE message to Unified CM.
Note
As soon as SIP dial rules are implemented on Type-B IP phones, KPML-based dialing is disabled. If a
user dials a string of digits that do not match a SIP dial rule, none of the individual digit events will be
relayed to Unified CM. Instead, the entire dialed string will be sent en-bloc to Unified CM once the
dialing is complete (that is, once inter-digit timeout has occurred).
Type-B phones using SIP dial rules offer only one way to dial patterns not explicitly configured on the
phone. If a dialed pattern does not match a SIP dial rule, the user has to wait for inter-digit timeout before
the SIP NOTIFY message is sent to Unified CM. Unlike Type-A IP phones, Type-B IP phones do not
have a Dial key to indicate the end of dialing, except when on-hook dialing is used. In the latter case, the
user can press the dial key at any time to trigger the sending of all dialed digits to Unified CM.
Note
When a Type-B phone registers with the SRST router, the configured SIP dial rules are disabled.
If a particular pattern is recognized by the phone but blocked by Unified CM, the user must dial the entire
dial string before receiving an indication that the call is rejected by the system. For instance, if a SIP dial
rule is configured on the phone to recognize calls dialed in the form 919765551234 but such calls are
blocked by the Unified CM dial plan, the user will receive reorder tone at the end of dialing (after
pressing the 4 key).
14-19
Chapter 14
Dial Plan
It is important to note that pattern recognition configuration on the phone through the use of SIP dial
rules does not supersede the Class of Service and Route Plan configurations of Unified CM. A phone
might very well be configured to recognize long-distance patterns while Unified CM is configured to
block such calls because the phone is assigned a class of service allowing only local calls.
There are two types of SIP dial rules, based on the phone model on which they will be deployed:
There are four basic Dial Parameters that can be used as part of a dial rule:
Pattern
This parameter is the actual numerical representation of the pattern. It can contain digits, wildcards,
or instructions to play secondary dial tone. The following table provides a list of values and their
effect for the two types of dial rules.
Dial Rule Type
Pattern
7905_7912
7940_7960_OTHER
Digits 0
through 9
Corresponding digit
Corresponding digit
n/a
tn
n/a
Indicates a timeout value of n
seconds. For example, 1t3 would
match 1000 and trigger the sending of
an invite to Unified CM after 3
seconds.
rn
14-20
Chapter 14
Dial Plan
Dial Plan Elements
Pattern
7905_7912
7940_7960_OTHER
n/a
IButton
This parameter specifies the button to which the dial pattern applies. If the user is initiating a call
on line button 1, only the dial patterns specified for Button 1 apply. If this optional parameter is not
configured, the dial pattern applies to all lines on the phone. This parameter applies only to the Cisco
SIP IP Phones 7940, 7941, 7942, 7945, 7960, 7961, 7962, 7965, 7970, 7971, and 7975. The button
number corresponds to the order of the buttons on the side of the screen, from top to bottom, with 1
being on top button.
Timeout
This parameter specifies the time, in seconds, before the system times out and dials the number as
entered by the user. To have the number dialed immediately, specify 0. This parameter is available
only for 7940_7960_OTHER dial rules. If this parameter is omitted, the phone's default inter-digit
timeout value is used (default of 10 seconds).
User
This parameter represents the tag that automatically gets added to the dialed number. Valid values
include IP (when SIP Call Agents other than Unified CM are deployed) and Phone. This parameter
is available only for 7940_7960_OTHER dial rules. This parameter is optional, and it should be
omitted in deployments where Unified CM is the only call agent. Keep in mind that a user=phone
tag in a SIP request sent to Unified CM will force Unified CM to treat the SIP URI as a numeric
URI. A SIP URI in the form of alice@cisco.com;user=phone will never be routed successfully
because the user=phone tag forces numeric treatment and alice will not match any numeric pattern
provisioned in Unified CM.
Note
The Cisco Unified IP Phone 7905 and 7912 choose patterns in the order in which they have been created
in the SIP dial rules, whereas all the other phone models choose the pattern offering the longest match.
The following table shows which pattern would be chosen if a user dialed 95551212.
7905_7912
7940_7960_OTHER
..
9.
14-21
Chapter 14
Dial Plan
Among all the potentially matching patterns, it matches the fewest strings other than the dialed
string.
For example, consider the case shown in Figure 14-12, where the call routing table includes the patterns
1XXX, 12XX, and 1234.
Figure 14-12
1XXX
User A dials
"1200"
User B dials
"1212"
Gateways
12XX
Pool of IP Phones
121X
IP
IP
IP Phone
User C dials
"1234"
1234
IP
126911
IP
When user A dials the string 1200, Unified CM compares it with the patterns in its call routing table. In
this case, there are two potentially matching patterns, 1XXX and 12XX. Both of them match the dialed
string, but 1XXX matches a total of 1000 strings (from 1000 to 1999) while 12XX matches only 100
strings (from 1200 to 1299). Therefore, 12XX is selected as the destination of this call.
When user B dials the string 1212, there are three potentially matching patterns, 1XXX, 12XX and
121X. As mentioned above, 1XXX matches 1000 strings and 12XX matches 100 strings. However, 121X
matches only 10 strings; therefore it is selected as the destination of the call.
When user C dials the string 1234, there are three potentially matching patterns, 1XXX, 12XX, and
1234. As mentioned above, 1XXX matches 1000 strings and 12XX matches 100 strings. However, 1234
matches only a single string (the dialed string); therefore it is selected as the destination of this call.
14-22
Chapter 14
Dial Plan
Dial Plan Elements
When determining the number of matched strings for a variable-length pattern, Unified CM takes into
account only the number of matched strings that are equal in length to the number of digits dialed.
Assuming a user dials 1311 and we have patterns 1XXX, 1[2-3]XX, and 13!, the following table shows
the number of matched strings of these potentially matching patterns.
Pattern
1XXX
1000
1000 to 1999
1[2-3]XX
200
13!
100
In this example the variable-length pattern 13! is selected as the best match.
Note
Whenever a directory number (DN) is configured in Cisco Unified CM, it is placed in the call routing
table, regardless of whether the respective device (for example, an IP phone) is registered or not. An
implication of this behavior is that it is not possible to rely on secondary, identical patterns to provide
failover capabilities to applications when the device (and hence the primary pattern) is unregistered.
Because the primary pattern is permanently in the call routing table, the secondary pattern will never be
matched.
Directory URIs
All endpoints registered with Unified CM are provisioned with one or more numeric (possibly including
a leading +) directory numbers. Up to five directory URIs can be associated with each directory number.
This association can be created by explicitly associating directory URIs to directory numbers. If a
directory URI is configured for an end user, this directory URI will be automatically associated with the
primary extension of that end user as soon as the primary extension gets defined for that end user. All
automatically associated directory URIs are created in the partition Directory URI, while manually
configured directory URIs can be in any partition. Manually configured directory URIs can reside in the
same partition as the directory number they are associated with, but do not have to. Directory URIs have
to be unique per partition.
Exactly one of the directory URIs associated with a directory number has to be marked as the primary
directory URI of that directory number. If a user directory URI gets associated automatically with the
primary extension of that user, then this directory URI will also automatically be the primary directory
URI for that directory number. If no directory URI is associated automatically, then one of the
14-23
Chapter 14
Dial Plan
configured directory URIs has to be selected as the primary directory URI. The purpose of the primary
directory URI is that this directory URI will be used as the calling identity directory URI for calls
originating from the respective directory number.
The possible association of directory URIs with any directory number allows callers to reach any
directory number by dialing the associated directory URI. The called directory number can be on any
device registered to Unified CM using any protocol. Similarly, Unified CM can deliver a directory URI
caller ID for any call from any directory number as long as a directory URI is associated with the calling
directory number.
To enable intercluster routing of directory URIs, Unified CM can be provisioned to exchange directory
URI catalogs with other clusters through the Intercluster Lookup Service (ILS). Each cluster configured
to exchange directory URI catalogs with other clusters advertises all locally configured directory URIs
in a single directory URI catalog together with a location attribute, the SIP route string. This location
attribute in multi-cluster environments is used to direct calls for directory URIs to the correct cluster
when the host portion of the directory URI cannot be used to deterministically route the SIP request.
This, for example, is the case when a flat URI scheme such as <user>@example.com is use. The host
portion "example.com" does not uniquely identify the remote Unified CM cluster that hosts a given URI.
For details of how calls to directory URIs learned from remote clusters are routed, see the section on
Routing of SIP Requests in Unified CM, page 14-49.
Translation Patterns
Translation patterns are one of the most powerful tools in Unified CM to manipulate digits for any type
of call. They follow the same general rules and use the same wildcards as route patterns. As with route
patterns, you assign a translation pattern to a partition. However, when the dialed digits match the
translation pattern, Unified CM does not route the call to an outside entity such as a gateway; instead, it
performs the translation first and then routes the call again, this time using the calling search space
configured within the translation pattern.
Translation patterns can be used for a variety of applications, as shown by the example in Figure 14-13.
Application Example for Translation Patterns
Calling Search
Spaces
Partitions
Translations_pt
Phone_css
IP
Dials "0"
to reach
operator
Translation Pattern
transforms 0 into
2001 and forces
second lookup
Delivers "2001"
AllPhones_pt
Internal_css
348589
Figure 14-13
In this example, the administrator wishes to provide users with an operator service that is reached by
dialing 0, while also maintaining a fixed-length internal numbering plan. The IP phones are configured
with the Phone_css calling search space, which contains the Translations_pt partition (among others). A
translation pattern 0 is defined in this partition, and the configured Called Party Transform Mask
instructs Unified CM to replace the dialed string (0) with the new string 2001, which corresponds to the
14-24
Chapter 14
Dial Plan
Dial Plan Elements
DN of the operator phone. A second lookup (of 2001 this time) is forced through the call routing engine,
using the Internal_css calling search space, and the call can now be extended to the real operator DN of
2001, which resides in the AllPhones_pt partition.
Note
When a dialed number is manipulated using a translation pattern, the translated number is recorded in
the call detail record (CDR). However, when the digit manipulation occurs within a route list, the CDR
will show the originally dialed number, not the translated one. The Placed Calls directory on the IP phone
always shows the string as it was dialed by the user.
The general use case for translation patterns is to create a mapping from a certain dial string format to a
string to be matched by other dial plan elements. This mapping implements overlay dialing habits on top
of the "native" dialing habits created by other patterns such as route patterns and directory numbers.
Typically for the secondary lookup, translation patterns that implement a dialing normalization should
simply use the calling search space that activates the translation pattern. This behavior, referred to as
CSS Inheritance, is selected by the option Use Originator's Calling Search Space on the translation
pattern. Enabling this option allows reuse of dialing normalization translation patterns for different
classes of service, each defined by a different calling search space.
14-25
Chapter 14
Dial Plan
Figure 14-14
Route Pattern
Matches dialed number for external calls
Performs digit manipulation (optional)
Points to a route list for routing
Hunt/Route List
Chooses path for call routing
Per-route group digit manipulation
Points to prioritized route groups
Route
Pattern
Route
List
1st Choice
2nd Choice
Route
Group
Route
Group
Route Group
Distributes calls to
GWs/Trunks
Transformation Patterns
Modifies calling (Cg) party
Modifies called (Cd) party
Transformation
Patterns
Transformation
Patterns
GK
IP WAN
PSTN
271554
Devices
Gateways (H.323,
MGCP, SIP)
Trunk (H.225,
ICT, SIP)
Configuration Order
The following sections describe the individual elements of the external route construct in Unified CM:
Route Patterns
Route patterns are strings of digits and wildcards, such as 9.[2-9]XXXXXX, configured in Unified CM
to route calls to external entities. The route pattern can point directly to a gateway for routing calls or
point to a route list, which in turn points to a route group and finally to a gateway.
Cisco strongly recommends that you use the complete route pattern, route list, and route group construct
because it provides the greatest flexibility for call routing, digit manipulation, route redundancy, and
future dial plan growth.
The @ Wildcard
The @ wildcard is a special macro function that expands into a series of patterns representing the
entire national numbering plan for a certain country. For example, configuring a single unfiltered
route pattern such as 9.@ with the North American Numbering Plan really adds 166 individual route
patterns to the Unified CM internal dial plan database.
14-26
Chapter 14
Dial Plan
Dial Plan Elements
It is possible to configure Unified CM to accept other national numbering plans. Once this is done,
the @ wildcard can be used for different numbering plans even within the same Unified CM cluster,
depending on the value selected in the Numbering Plan field on the Route Pattern configuration
page. For more information, please refer to the Cisco Unified Communications Manager Dial Plan
Deployment Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps5629/prod_maintenance_guides_list.html
The @ wildcard can be practical in several small and medium deployments, but it can become harder
to manage and troubleshoot in large deployments because adopting the @ wildcard forces the
administrator to use route filters to block certain patterns (see Route Filters, page 14-27).
Route Filters
Use route filters only with the @ route pattern to reduce the number of route patterns created by the
@ wildcard. A route filter applied to a pattern not containing the @ wildcard has no effect on the
resulting dial plan.
The logical expression you enter with the route filter can be up to 1024 characters, excluding the
NOT-SELECTED fields.
As you increase the number of logical clauses in a route filter, the refresh time of the configuration
page also increases and can become unacceptably long.
For large-scale deployments, use explicit route patterns rather than the @ wildcard and route filters.
This practice also facilitates management and troubleshooting because all patterns configured in
Unified CM are easily visible from the Route Pattern configuration page.
International destinations are usually configured using the ! wildcard, which represents any quantity
of digits. For example, in North America the route pattern 9.011! is typically configured for
international calls. In most European countries, the same result is accomplished with the 0.00! route
pattern.
The ! wildcard is also used for deployments in countries where the dialed numbers can be of varying
lengths. In such cases, Unified CM does not know when the dialing is complete and will wait for
15 seconds (by default) before sending the call. You can reduce this delay in any of the following
ways:
Reduce the T302 timer (Service Parameter TimerT302_msec) to indicate end of dialing, but do
not set it lower than 4 seconds to prevent premature transmission of the call before the user is
finished dialing.
Configure a second route pattern followed by the # wildcard (for example, 9.011!# for North
America or 0.00!# for Europe), and instruct the users to dial # to indicate end of dialing. This
action is analogous to hitting the "send" button on a cell phone.
Overlap Sending and Overlap Receiving
In countries whose national numbering plan is not easily defined with static route patterns, you can
configure Unified CM for overlap sending and overlap receiving.
Overlap sending means that Unified CM keeps collecting digits as they are dialed by the end users, and
passes them on to the PSTN as they are dialed. To enable overlap sending, check the Allow Overlap
Sending box on the Route Pattern Configuration page. The route pattern needs to include only the PSTN
access code (for example, "9." in North America or "0." in many European countries).
14-27
Chapter 14
Dial Plan
Overlap receiving means that Unified CM receives the dialed digits one-by-one from a PRI PSTN
gateway, and it then waits for completion of the dialed string before attempting to route the call to an
internal destination. To enable overlap receiving, set the OverlapReceivingFlagForPRI service
parameter to True.
Digit Manipulation in Route Patterns
Digit manipulations configured on a route pattern affect the calling and called party number, no
matter what route group the call eventually takes. Digit manipulations configured in the route list's
view of its member route groups have a route-specific effect: only the transformations configured
on the route group used to place the call will be performed.
Digit manipulation in the route list's view of its member route group overrides any digit
manipulation done in the route pattern.
Transformation patterns configured on the device selected to route the call (or on that devices
device pool) take precedence over calling and called party transformations configured in the route
pattern and/or route list. If a transformation calling search space (CSS) is configured on the device
selected to route the call (or on that devices device pool), then transformations configured in the
route pattern or route list are considered only if no match is found using the respective
transformation CSS. The input to the transformation CSS always is the untransformed number
before applying route pattern or route list transformations.
If you configure digit manipulation in the route pattern, the Call Detail Record (CDR) records the
dialed number after the digit manipulation has occurred. If you configure digit manipulation only in
the route group or on the device level, the CDR records the actual dialed number prior to the digit
manipulation.
Similarly, if you configure digit manipulation in the route pattern, the IP phone display of the calling
party will show the manipulated number. If you configure digit manipulation only in the route group,
the manipulations will be transparent to the end user.
Calling Line ID
The calling line ID presentation can be enabled or disabled on the gateway and also can be
manipulated in the route pattern, based on site requirements.
If you select the option Use Calling Party's External Phone Number Mask, then the external call uses
the calling line ID specified for the IP phone placing the call. If you do not select this option, then
the mask placed in the Calling Party Transform Mask field is used to generate the calling party ID.
Call Classification
Calls using this route pattern can be classified as on-net or off-net calls. This route pattern can be
used to prevent toll fraud by prohibiting off-net to off-net call transfers or by tearing down a
conference bridge when no on-net parties are present. (Both of these features are controlled by
Service Parameters within Unified CM Administration.)
When the "Allow device override" box is enabled, the calls are classified based on the call
classification settings on the associated gateway or trunk.
The Forced Authorization Codes (FAC) checkbox is used to restrict the outgoing calls when using
a particular route pattern. If you enable FAC through route patterns, users must enter an
authorization code to reach the intended recipient of the call.
When a user dials a number that is routed through a FAC-enabled route pattern, the system plays a
tone that prompts for the authorization code. To complete the call, the user authorization code must
meet or exceed the level of authorization that is specified to route the dialed number.
14-28
Chapter 14
Dial Plan
Dial Plan Elements
Only the authorization name appears in the call detail records (CDR); the authorization code does
not appear in the CDR.
The FAC feature is not available if the "Allow overlap sending" checkbox is enabled.
The Client Matter Code (CMC) checkbox is used to track calls to certain numbers when using a
particular route pattern. (For example, a company can use it to track calls to certain clients.)
If you enable CMC for a route pattern, users must enter a code to reach the intended destination.
When a user dials a number that is routed through a CMC-enabled route pattern, the system plays a
tone that prompts for the code. The user must enter the correct code in order to complete the call.
Client Matter Codes appear in the call detail records so that they can be used by the CDR analysis
and reporting tool to generate reports for client billing and accounting.
The CMC feature is not available if the "Allow overlap sending" checkbox is enabled.
If both CMC and FAC are enabled, the user dials a number, enters the FAC when prompted to do so,
and then enters the CMC at the next prompt.
Route Lists
A route list is a prioritized list of eligible paths (route groups) for an outbound call. A typical use of a
route list is to specify two paths for a remote destination, where the first-choice path is across the IP
WAN and the second-choice path is through a PSTN gateway.
Route lists have the following characteristics:
A route list is a prioritized list of route groups that function as alternate paths to a given destination.
For example, you can use a route list to provide least-cost routing, where the primary route group in
the list offers a lower cost per call and the secondary route group is used only if the primary is
unavailable due to an "all trunks busy" condition or insufficient IP WAN resources.
Each route group in the route list can have its own digit manipulation. For example, if the route
pattern is 9.@ and a user dials 9 1 408 555 4000, the IP WAN route group can strip off the 9 1 while
the PSTN route group may strip off just the 9.
Multiple route lists can contain the same route group. The digit manipulation for the route group is
associated with the specific route list that points to the route group.
14-29
Chapter 14
Dial Plan
If you are performing several digit manipulations in a route pattern or a route group, the order in
which the transformations are performed can impact the resulting calling and called party numbers
used for the call. Unified CM performs the following major types of digit manipulations in the order
indicated:
1.
Discarding digits
2.
Called and calling party transformations as defined in the route pattern or for the route group
3.
Prefixing digits
Keep in mind that calling and called party transformations defined on the egress device (gateway or
trunk) override calling and called party transformations defined in route patterns and route groups.
Route Groups
Route groups control and point to specific devices, which are typically gateways (MGCP, SIP, or H.323),
H.323 or SIP trunks to a gatekeeper, remote Unified CM cluster, or Cisco Unified Border Element.
Unified CM sends calls to the devices according to the distribution algorithm assigned. Unified CM
supports top-down and circular algorithms.
Note
SIP gateways
H.323 gateways
H.225 trunk, gatekeeper controlled trunk to standard H.323 gateways, via a gatekeeper
Intercluster trunk, not gatekeeper controlled direct trunk to another Unified CM cluster
Intercluster trunk, gatekeeper controlled trunk to other Unified CM clusters and/or H.323
gateways, via a gatekeeper
SIP trunk trunk to another Unified CM cluster, a Cisco Unified Border Element, a Session Border
Controller, or a SIP proxy
Both the H.225 and intercluster trunk (gatekeeper controlled) will automatically discover if the other
endpoint is a standard H.323 gateway or a Unified CM and will select H.225 or Intercluster Trunk
protocol accordingly.
14-30
Chapter 14
Dial Plan
Dial Plan Elements
In Figure 14-15, a route pattern defined as 9.1[2-9]XX[2-9]XXXXXX points to a route list referencing
a non-local route group containing San Francisco gateways. If this route pattern is placed in a partition
contained in the calling search spaces of phones in Dallas, San Francisco, and New York, national calls
from devices in those three cities will egress to the PSTN in San Francisco.
Figure 14-15
SFODevices
DFWDevices
CSSs
Partitions
Route Lists
US_pstn_part
US LOC
RL
Route Groups
DFWDevices
SFO RG
SFODevices
9.1[2-9]XX[2-9]XXXXXX
V
V
JFKDevices
271557
JFKDevices
SFO Gateways
14-31
Chapter 14
Dial Plan
By contrast, if this same route pattern is modified to point to a route list containing the Standard Local
Route Group as in Figure 14-16, then calls made from the Dallas site would egress to the PSTN through
the Dallas gateway, calls made from the New York site would egress to the PSTN through the New York
gateway, and calls made from the San Francisco site would egress to the PSTN through the San Francisco
gateway.
The use of Local Route Group allows for egress gateway selection based on the calling device, which
allows for site-independent route patterns that can be reused by calling search spaces for phones in all
sites.
Figure 14-16
Partitions
Route Lists
Route Groups
DFW RG
DFWDevices
US_pstn_part
SFODevices
9.1[2-9]XX[2-9]XXXXXX
US LOC
RL
Local
Route
group
DFW
Gateways
SFO RG
V
SFO
Gateways
JFKDevices
JFK RG
V
JFK
Gateways
254267
JFK Devices
SFO Devices
DFW Devices
CSSs
The Device Mobility feature allows the device pool to be assigned to an endpoint based on the current
subnet to which it has roamed. This permits assignment of the local route group to be based on the site
where the phone is currently located.
Example 14-3 Device Mobility
A phone is moved from the San Francisco site to the New York site. Based on the phone's new IP address
(part of the IP subnet associated with the New York site), a New York device pool is assigned to the
phone. If the next call placed by the roaming phone matches a route pattern using a route list containing
the Standard Local Route Group, it will be routed through the New York gateway.
If a local route group is used in forwarded call scenarios where, for example, phone A calls phone B and
B is forwarded to a destination in the PSTN, then the route pattern in the call forward calling search
space of phone B determines the class of service for calls forwarded by phone B, whereas by default the
local route group associated with phone A's device pool is used to determine the egress gateways when
hitting Standard Local Route group in the route list selected by the route pattern found using phone B's
call forward calling search space. As a result, typically a gateway local to phone A is used for the
forwarded call. This makes sure that the caller ID of the initial caller (phone A) can be sent to the PSTN
14-32
Chapter 14
Dial Plan
Dial Plan Elements
and that this caller ID will not be screened by the provider. There is a service parameter that allows
administrators to configure the local route group selection policy for forwarded calls. The service
parameter can be set to:
Calling Party's Local Route Group Backward compatible default. The local route group
associated with the initial caller's device pool is selected (phone A in above example).
Original Called Party The local route group associated with the called phone's device pool is
selected (phone B in above example).
Last Redirecting Party The local route group associated with the phone's device pool that is
forwarding the call to the PSTN is selected (phone B in above example). These last two options
differ only in cases where the call is forwarded through multiple hops before it finally gets
forwarded out to the PSTN.
A company negotiates a favorable PSTN interconnection rate for a group of trunks located in Chicago.
If a route list includes a route group containing gateways in Chicago as its first entry and the Standard
Local Route Group as the second choice, then any call it processes will first be sent to the preferred
lower-cost gateways in Chicago. If a Chicago gateway is not available, if no ports are free, or if there is
not enough bandwidth to allow the call between the calling phone and the Chicago gateway, then the next
choice will be to attempt to route the call through the gateway co-located with the calling phone, as
determined by the local route group in the calling phone's device pool configuration. (See Figure 14-17.)
14-33
Chapter 14
Dial Plan
Figure 14-17
CSSs
Partitions
Route Lists
Route Groups
ORD RG
V
ORD
Gateways
DFW RG
First
choice
DFWDevices
DFW
Gateways
US_pstn_part
SFODevices
9.1[2-9]XX[2-9]XXXXXX
US LOC
RL
Local
Route
group
Second
choice
JFKDevices
SFO RG
V
SFO
Gateways
JFK RG
V
JFK
Gateways
254268
JFK Devices
SFO Devices
DFW Devices
LRG_Emergency
SJC (branch)
RG_SFO
RG_SJC
OAK (branch)
RG_SFO
RG_OAK
SFO (hub)
RG_SFO
RG_SFO
TPA (branch)
RG_MCO
RG_TPA
MIA (branch)
RG_MCO
RG_MIA
MCO (hub)
RG_MCO
RG_MCO
14-34
Chapter 14
Dial Plan
Dial Plan Elements
In this example the gateways in the major hubs (SFO and MCO) are used for PSTN calls by users in the
hub sites and in the branch sites associated with the hub (SJC and OAK use SFO; TPA and MIA use
MCO), while emergency calls always use local PSTN resources.
Figure 14-18 shows the call routing and local route group selection for an emergency call. Route list
RL_911 used by the emergency route pattern would have LRG_Emergency as the first route group entry.
The second entry in the route list refers to the Standard Local Route Group to make sure that the default
PSTN resource defined on the device pool is selected as failover. Whenever an emergency call is placed
and the route list entry LRG_Emergency is selected, Unified CM will dereference the placeholder
LRG_Emergency and will instead use the route group configured for LRG_Emergency on the device
pool of the calling device. The example shows how, for phones in sites SFO and SJC, local PSTN
gateways are selected for emergency calls.
Figure 14-18
Route Group:
RG_SJC
LRG_PSTN_1: RG_SFO
LRG_Emergency: RG_SJC
Standard Local Route Group:
up: RG_SJC
911
Route Pattern
911
Route List
RL_911
Route Groups:
911
LRG_Emergency
Standard Local
Route Group
R
Route
Group:
R
RG_SFO
Device Pool SFOPhone
LRG_PSTN_1: RG_SFO
LRG_Emergenca: RG_SFO
Standard Local Route Group: RG_SFO
348787
Using the same concept, a site-independent PSTN route pattern can be defined to point to a route list that
uses LRG_PSTN. LRG_PSTN then is dereferenced to the route group defined on the device pool level
for named local route group LRG_PSTN. Figure 14-19 shows how PSTN calls from sites SJC and SFO
are routed to centralized PSTN gateways in site SFO, based on the device pool local route group settings.
14-35
Chapter 14
Dial Plan
Figure 14-19
Route Group:
RG_SJC
LRG_PSTN_1: RG_SFO
LRG_Emergency: RG_SJC
Standard Local Route Group: RG_SJC
+493100123456
Route Pattern
\+!
Route List
RL_PSTN
Route Groups:
+493100123456
LRG_PSTN
Standard Local
Route Group
Route Group:
R
R
RG_SFO
Phone in Device Pool SFOPhone
348788
LRG_PSTN_1: RG_SFO
LRG_Emergenca: RG_SFO
Standard Local Route Group: RG_SFO
Undefined local route groups are skipped during the egress routing device selection process. If a route
list contains a local route group to which no route group has been assigned on the device pool of the
calling device, then this entry in the route list is skipped and the next route group member of the route
list is considered. When using route lists containing only local route groups, it is important to make sure
that route groups are defined consistently on all device pools of all call originating devices to avoid
dropping egress calls due to route list exhaustion without ever reaching a real route group.
Always using Standard Local Route Group as the last entry in all route lists and making sure that a route
group for Standard Local Route Group is selected on all device pools, can be used as a safeguarding
mechanism to avoid above route list exhaustion problem.
Pattern Urgency
Translation patterns, route patterns, and DNs can be configured as urgent patterns. The default value for
pattern urgency is urgent for translation patterns and non-urgent for route patterns and DNs. Only the
pattern urgency of route patterns, translation patterns, and DNs can be configured. All other patterns are
always non-urgent.
Marking a pattern as urgent is often used to force immediate routing of certain calls as soon as a match
is detected, without waiting for the T302 timer to expire. For example, in North America, if the patterns
9.911 and 9.[2-9]XXXXXX are configured and a user dials 9911, Unified CM has to wait for the T302
timer before routing the call because further digits may cause the 9.[2-9]XXXXXX pattern to match. If
Urgent Priority is enabled for the 9.911 route pattern, Unified CM makes its routing decision as soon as
the user has finished dialing 9911, without waiting for the T302 timer.
14-36
Chapter 14
Dial Plan
Dial Plan Elements
Making a pattern urgent forces the T302 timer to expire as soon as the configured pattern is the best
match for the dialed number. This does not mean that the urgent pattern has a higher priority than other
patterns; the closest-match logic described in the section on Call Routing in Unified CM, page 14-22,
still applies.
For example, assume the route pattern 1XX is configured as urgent and the pattern 12! is configured as
a non-urgent route pattern. If a user dials 123, Unified CM will not make its routing decision as soon as
it receives the third digit because even though 1XX is an urgent pattern, it is not the best match (10 total
patterns matched by 12! versus 100 patterns matched by 1XX). Unified CM will have to wait for
inter-digit timeout before routing the call because the pattern 12! allows for more digits to be input by
the user.
Consider another example where pattern 12[2-5] is marked as urgent and 12! is configured as a
non-urgent pattern. If the user dials 123, the pattern 12[2-5] is the best match (four total patterns matched
by 12[2-5] versus 10 patterns matched by 12!). Because the T302 timer is aborted and because the
urgent-priority pattern is the best match, no further user input is expected. Unified CM routes the call
using pattern 12[2-5].
A variable-length urgent translation pattern like 9011.! in Figure 14-20 will not force inter-digit timeout.
As the dialed digits are received and analyzed digit-by-digit, as soon as an urgent translation pattern is
the only (or best) match, the digit transformations defined on the translation pattern will be executed
immediately and the secondary lookup as defined by the CSS on the translation pattern occurs.
Figure 14-20
CSSs
Partitions
Translations
9011.!/urgent, strip pre-dot, prefix +
someCSS
PSTNInternational
\+!
\+33XXXXXXXXX
348609
E164PSTN
Assuming the configuration in Figure 14-20, when the user dials 901133158405858 the call will be
routed immediately after the last digit is dialed. The call will match translation pattern 9011.!, the dialed
digits will be transformed to +3333158405858 (9011 discarded and + prefixed), which matches the
fixed-length PSTN route pattern \+33XXXXXXXXX (nine-digit NSNs used in France).
14-37
Chapter 14
Dial Plan
On the other hand, if the user dials 9011496907739001, the user will experience inter-digit timeout.
After matching 9011.! the resulting digits +496907739001 match route pattern \+!, and Unified CM
needs to wait for further digits to make sure that the caller did not intent to continue to dial further digits.
Further digits dialed would still match on the same route pattern.
The example in Figure 14-20 also shows how urgent translation patterns can be used to implement some
abbreviated off-net dialing habit. Both translation patterns starting with 8 will accept exactly eight digits,
transform the dialed digits to +E.164, and then execute the secondary lookup.
Dialing 83315858 will be routed immediately without inter-digit timeout. The dialed digits match fixed
length translation pattern 8331.5XXX, and the translated called party number +33158405858 matches
the fixed-length route pattern \+33XXXXXXXXX.
However, dialing 84969001 will not be routed immediately by default. The dialed digits are matched by
the fixed-length translation pattern 8496.9XXX, and the translated called party number +496907739001
matches the variable-length PSTN route pattern \+!. This example shows that neither the pattern urgency
nor the fixed-length characteristic of an intermediate translation pattern match is inherited by the
secondary lookup defined by the CSS configured on the intermediate translation pattern (E164PSTN).
Because the route pattern matched in the secondary lookup is a variable length pattern, Unified CM is
forced to wait for inter-digit timeout. If the intermediate translation pattern is a fixed length translation
pattern, waiting for further digits in the secondary lookup does not make much sense because any further
digits will lead to a situation where the intermediate translation pattern will not be matched any more.
Hence, for fixed length translation patterns it does make sense to change the inter-digit timeout handling
for the secondary lookup. To achieve this, the option Do Not Wait For Interdigit Timeout On
Subsequent Hops on the translation pattern has to be set. If this option is set, then after matching the
translation pattern, Unified CM will not wait for any further digits and will just match the translated
called party number against the patterns identified by the CSS defined on the intermediate translation
pattern. As a general rule, Do Not Wait For Interdigit Timeout On Subsequent Hops should be
enabled on all fixed length translation patterns.
Another typical use case for the Do Not Wait For Interdigit Timeout On Subsequent Hops option is
the secondary lookup of dialing normalization translation patterns using a special key to terminate digit
collection to avoid interdigit timeout. As an example, in a US dial plan a dialing normalization
translation pattern matching international destinations with termination character # (such as 9011.!#) can
match on variable length international dialing and allows users to terminate dialing by pressing #. This
translation pattern's secondary lookup would typically match on a variable length route pattern such as
\+[^1]! and this match in the secondary lookup would again force digit analysis to wait for further digits.
Again the easiest way to avoid this timeout is to set the Do Not Wait For Interdigit Timeout On
Subsequent Hops option on the dialing normalization translation pattern 9011.!#.
Note
Called party transformation patterns do not have any effect on phones. The called party transformation
pattern CSS of the device pool does not impart any effects on the phones to which it is assigned.
14-38
Chapter 14
Dial Plan
Dial Plan Elements
Both pattern types consist of a numerical representation of the calling or called party number to be
matched. The syntax used is the same as that of other patterns such as route patterns, translation patterns,
directory numbers, and so forth. (See Figure 14-21.)
The transformation operators include discard digit instructions (for example, pre-dot), a calling party
transformation mask, prefix digits, and control over the calling party presentation (either Default,
Allowed, or Restricted). Calling party transformation patterns can be configured to use the calling party's
external phone number mask as the calling party number.
Partitions and calling search spaces control which calling party transformation patterns are applied to
which gateways or trunks. Gateways or trunks can use either their associated device pool's calling party
transformation CSS or the device's own calling party transformation CSS, in reverse order of precedence.
The same mechanism is used to control the applicability of called party transformation patterns.
Calling and called party transformation patterns configured on a Gateway Configuration page under Call
Routing Information - Outbound Calls affect the calling or called party number sent to the gateway,
as well as the calling or called party's numbering type and numbering plan. Calling and called party
transformation patterns applied under Incoming Calling Party Settings apply to calls coming from the
gateway.
Figure 14-21
CSSs
Partitions
pattern
France_CdPTP
NANP_called_xforms
discard
prefix
type
\+.1!
pre-dot
\+.!
pre-dot
national
011
national
V
French
HQ Gateways
pattern
YOW_called_xforms
discard
prefix
type
\+1.613[2-9]XXXXXX pre-dot
subscriber
France_CgPTP
pattern
V
V
Nice
Gateways
YOW_CdPTP
France_called_xforms
discard
prefix
type
\+.!
pre-dot
00
international
\+33.!
pre-dot
national
pattern
NANP_calling_xforms
discard
prefix
type
\+1.!
pre-dot
\+.!
pre-dot
national
011
international
V
pattern
NANP_CgPTP
France_calling_xforms
discard
prefix
type
\+.!
pre-dot
00
international
\+33.!
pre-dot
national
271556
Ottawa
Gateways
14-39
Chapter 14
Dial Plan
Figure 14-21 illustrates how calling and called party transformation patterns would be applied to
different groups of gateways connected to the PSTN in different parts of the PSTN (France and NANP
area).
Within the North American Numbering Plan (NANP), gateways located in Ottawa, Canada (airport code
YOW) are assigned to the Calling Party Transformation CSS NANP_CgPTP, which contains partition
NANP_calling_xforms. Any call with a calling party number beginning with +1 (that is, originating from
within the NANP) would match both patterns configured within partition NANP_calling_xforms.
Following the best-match logic, the first pattern will be chosen, and the calling party number will be
stripped of the + sign and NANP country code 1. The remaining part of the calling party number will be
used as the calling party number sent to the PSTN, with numbering type set to National.
For example, if a call from +1 613 555 1234 were sent out the YOW gateways, the calling party number
would be transformed to 613 555 1234 with a numbering type set to National.
If a call from the same caller were to be sent to a French gateway, a different set of calling party
transformation patterns would apply. For example, if a call from +1 613 555 1234 were sent out a
gateway located in Nice, France (airport code NCE), the calling party transformation patterns contained
in partition France_calling_xforms would be applied. In this case, the calling party number would be
transformed to 001 613 555 1234 with the numbering type set to International.
Note
Calling party number transformations may be overridden once the call is sent out the gateway. Many
service providers will not permit calling party numbers outside a given range, as determined by local
service agreements or regulations.
The same process applies to the called party number transformation patterns. For Ottawa gateways, the
assigned called party transformation CSS is YOW_CdPTP, which contains partitions
NANP_Called_xforms and YOW_Called_xforms. A call placed to a destination number within the
Numbering Plan Area 613 would match all patterns contained in these two partitions. However, the best
match process would select pattern \+1.613[2-9]XXXXXX.
For example, on a call placed to +1 613 555 9999 through the Ottawa gateways, the called party number
would be transformed to 516 555 9999 with a numbering type set to Subscriber.
14-40
Chapter 14
Dial Plan
Dial Plan Elements
A partition is a group of directory numbers (DNs) or directory URIs with similar accessibility, and a
calling search space defines which partitions are accessible to a particular device. A device can call only
those DNs and directory URIs located in the partitions that are part of its calling search space.
As illustrated in Figure 14-22, items that can be placed in partitions all have a dialable pattern, and they
include phone lines, route patterns, translation patterns, CTI route group lines, CTI port lines, voicemail
ports, and Meet-Me conference numbers. Conversely, items that have a calling search space are all
devices capable of dialing a call, such as phones, phone lines, gateways, and applications (via their CTI
route groups or voicemail ports).
14-41
Chapter 14
Dial Plan
IP
Phones
PartitionA
2002
2001
2000
CSS1
PartitionA
PartitionB
more
Lines
CSS2
PartitionB
CSS3
V
PartitionB
Gateways
PartitionA
Applications
CSS4
PartitionA
"Dialable" devices
"Dialing" devices
15644
15644
15644
Lines
(Directory Numbers)
Translation
Patterns
911
\+.!
Route Patterns
9.[2-9]XX[2-9]XXXXXX
PartitionB
5000
Application Numbers
(CTI Route Points, CTI Ports)
900X
99XX
Special numbers
(MeetMe, CallPickup...)
8001
8001
9.011!
Route Patterns
271555
Figure 14-22
Partitions
The dial plan entries that you may place in a partition include IP phone directory numbers, directory
URIs, translation patterns, route patterns, CTI route points, and voicemail ports. As described in the
section on Call Routing in Unified CM, page 14-22, if two or more numeric dial plan entries (directory
numbers, route patterns, or so forth) overlap, Unified CM selects the entry with the closest match (most
specific match) to the dialed number. In cases where two dial plan entries match the dialed pattern
equally, Unified CM selects the dial plan entry that appears first in the calling search space of the device
making the call. Directory URIs always have to match completely; there is no concept of partial matches
for directory URIs.
For example, consider Figure 14-23, where route patterns 1XXX and 23XX are part of Partition_A and
route patterns 12XX and 23XX are part of Partition_B. The calling search space of the calling device
lists the partitions in the order Partition_A:Partition_B. If the user of this device dials 2345, Unified CM
selects route pattern 23XX in Partition_A as the matching entry because it appears first in the calling
device's calling search space. However, if the user dials 1234, Unified CM selects route pattern 12XX in
Partition_B as the matching entry because it is a closer match than 1XXX in Partition_A. Remember that
the partition order in a calling search space is used exclusively as a tie-breaker in case of equal matches
based on the closest-match logic.
14-42
Chapter 14
Dial Plan
Dial Plan Elements
Figure 14-23
Device
User dials
"2345"
23XX
IP
Partition_B
User dials
"1234"
12XX
114715
23XX
Note
When multiple equal-precision matches occur in the same partition, Unified CM selects the entry that is
listed first in its local dial plan database. Since you cannot configure the order in which the dial plan
database lists dial plan entries, Cisco strongly recommends that you avoid any possibility of
equal-precision matches coexisting within the same partition because the resulting dial plan logic is not
predictable in such cases.
Partitions can be activated or deactivated based on the time and date. You can activate or deactivate
partitions by first configuring time periods and schedules within Unified CM Administration and then
assigning a specific time schedule to each partition. Outside of the times and days specified by the
schedule, the partition is inactive, and all patterns contained within it are ignored by the Unified CM call
routing engine. For more information on this feature, see Time-of-Day Routing, page 14-89.
14-43
Chapter 14
Dial Plan
Line CSS
Partition L1
Partition L2
Partition L3
15644
15644
15644
more
Line
IP
Device
Note
Device CSS
Partition D1
Partition D2
Partition D3
Resulting CSS
Partition L1
Partition L2
Partition L3
Partition D1
Partition D2
Partition D3
114716
Figure 14-24
When device mobility is not used, the device calling search space is static and remains the same even as
the device is moved to different parts of the network. When device mobility is enabled, the device calling
search space can be determined dynamically based on where in the network the phone is physically
located, as determined by the phone's IP address. See Device Mobility, page 14-82, for more details.
If the same route pattern appears in two partitions, one contained in the line's calling search space and
one contained in the device's calling search space, then according to the rules described in the section
on Partitions, page 14-42, Unified CM selects the route pattern listed first in the concatenated list of
partitions (in this case, the route pattern associated with the line's calling search space).
The maximum length of the combined calling search space (device plus line) is 1024 characters,
including separator characters between each partition name. (For example, the string
partition_1:partition_2:partition_3 contains 35 characters.) Thus, the maximum number of partitions
in a calling search space varies, depending on the length of the partition names. Also, because the calling
search space clause combines the calling search space of the device and that of the line, the maximum
character limit for an individual calling search space is 512 (half of the combined calling search space
clause limit of 1024 characters).
Therefore, when you are creating partitions and calling search spaces, keep the names of partitions short
relative to the number of partitions that you plan to include in a calling search space. For more details
on configuring calling search spaces, refer to the Cisco Unified Communications Manager
Administration Guide, available online at
http://www.cisco.com
Before you configure any partitions or calling search spaces, all DNs reside in a special partition named
<None>, and all devices are assigned a calling search space also named <None>. When you create
custom partitions and calling search spaces, any calling search space you create also contains the
<None> partition, while the <None> calling search space contains only the <None> partition.
Note
Any dial plan entry left in the <None> partition is implicitly reachable by any device making a call.
Therefore, to avoid unexpected results, Cisco strongly recommends that you do not leave dial plan
entries in the <None> partition.
14-44
Chapter 14
Dial Plan
Dial Plan Elements
Note
Cisco strongly recommends that you do not leave any calling search space defined as <None>. Doing so
can introduce dial plan behavior that is difficult to predict.
Note
Call Forward All actions are different than any other call-forward action in that the destination number
is entered by each individual user when the feature is activated from a phone.
The system allows you to decide how call-forward calling search spaces take effect. There are three
possible options, as selected by the Calling Search Space Activation policy:
14-45
Chapter 14
Dial Plan
On Type-A IP phones running SIP, if Call Forward All is invoked from the phone itself, the device's
Rerouting Calling Search Space is used for forwarded calls. If Forward All actions are invoked from the
Unified CM User page or the Unified CM Administrative page, then any Forward All action initiated
from the phone is irrelevant.
For example, assume an Type-A IP phone running SIP is configured with Forward All to extension 3000
from the Unified CM User page. At the same time, the phone itself is configured to Forward All to
extension 2000. All calls made to that phone will be forwarded to extension 3000.
Note
On Type-A IP phones running SIP, invoking Forward All from the Unified CM User or Administrative
pages will not be reflected on the phone. The phone does not display any visual confirmation that calls
are forwarded.
When Forward All is initiated from an IP phone running SCCP or from an Type-B IP phone running SIP,
user input is simultaneously compared to the patterns allowed in the configured Forward All calling
search space(s). If an invalid destination pattern is configured, the user will be presented with reorder
tone. When Forward All is invoked from an Type-A IP phone running SIP, Forward All user input is
stored locally on the phone and is not verified against any calling search space in Unified CM. If user
input corresponds to an invalid destination, no notification is offered to the user. Calls made to that phone
will be presented with reorder tone as the phone tries to initiate a SIP re-route action to an invalid
destination number.
Always provision the call-forward calling search spaces with a value other than <None>. This
practice avoids confusion and facilitates troubleshooting because it enables the network
administrator to know exactly which calling search space is being used for forwarded calls.
Configure the Call Forward Busy and Call Forward No Answer calling search spaces with values
that allow them to reach the DNs for the voicemail pilot and voicemail ports but not external PSTN
numbers.
Configure both the Call Forward All calling search space and the Secondary Calling Search Space
for Forward All, according to your company's policy. Many companies choose to restrict forwarded
calls to internal numbers only, to prevent users from forwarding their IP phone lines to a
long-distance number and dialing their local IP phone number from the PSTN to bypass
long-distance toll charges on personal calls.
14-46
Chapter 14
Dial Plan
Dial Plan Elements
The Call Forward Unregistered (CFUR) feature is a way to reroute calls placed to a temporarily
unregistered destination phone. The configuration of CFUR consists of two main elements:
Destination selection
When the DN is unregistered, calls can be rerouted to either of the following destinations:
Voicemail
Calls can be sent to voicemail by selecting the voicemail checkbox and configuring the CFUR
calling search space to contain the partition of the voicemail pilot number.
A directory number used to reach the phone through the PSTN
This approach is preferred when a phone is located within a site whose WAN link is down. If
the site is equipped with Survivable Remote Site Telephony (SRST), the phone (and its
co-located PSTN gateway) will re-register with the co-located SRST router. The phone is then
able to receive calls placed to its PSTN DID number.
In this case, the appropriate CFUR destination is the corresponding PSTN DID number of the
original destination DN. Configure this PSTN DID in the destination field, preferably in E.164
format, including the + sign (for example, +1 415 555 1234). This allows the CFUR destination
to be processed by the calling phone's local route group, whether or not it uses the same off-net
access code and PSTN prefixes as the unregistered phone.
The Call Forward Unregistered functionality can result in telephony routing loops if a phone is
unregistered while the gateway associated with the phone's DID number is still under control of
Unified CM, as is the case if a phone is simply disconnected from the network. In such a case, the initial
call to the phone would prompt the system to attempt a first CFUR call to the phone's DID through the
PSTN. The resulting incoming PSTN call would in turn trigger another CFUR attempt to reach the same
phone's DN, triggering yet another CFUR call from the central PSTN gateway through the PSTN. This
cycle could repeat itself until system resources are exhausted.
The service parameter MaximumForwardUnRegisteredHopsToDn controls the maximum number of
CFUR calls that are allowed for a DN at the same time. The default value of 0 means the counter is
disabled. If any DNs are configured to reroute CFUR calls through the PSTN, loop prevention is
required. Configuring this service parameter to a value of 1 would stop CFUR attempts as soon as a
single call is placed through the CFUR mechanism. This setting would also allow only one call to be
forwarded to voicemail, if CFUR is so configured. Configuring this service parameter to a value of 2
would allow for up to two simultaneous callers to reach the voicemail of a DN whose CFUR setting is
configured for voicemail, while also limiting potential loops to two for DNs whose CFUR configuration
sends calls through the PSTN.
14-47
Chapter 14
Dial Plan
Note
Extension Mobility DNs should not be configured to send Call Forward Unregistered calls to the PSTN
DID associated with the DN. The DNs of Extension Mobility profiles in the logged-out state are deemed
to be unregistered, therefore any calls to the PSTN DID number of a logged-out DN would trigger a
routing loop. To ensure that calls made to Extension Mobility DNs in the logged-out state are sent to
voicemail, ensure that their corresponding Call Forward Unregistered parameters are configured to send
calls to voicemail.
14-48
Chapter 14
Dial Plan
Dial Plan Elements
On the receiving side, all directory URIs learned through GDPR are put into a single local repository to
be consulted when routing a non-numeric URI does not find a local URI match. All learned URIs are
treated as equivalent from the class-of-service perspective.
In contrast to this, numeric patterns and numbers learned through GDPR are put into local partitions
based on the type of information. Four separate partitions can be configured for +E.164 alternate
numbers, enterprise alternate numbers, +E.164 patterns, and enterprise patterns. The default partitions
for these different types of learned information are Global Learned E164 Numbers, Global Learned
E164 Patterns, Global Learned Enterprise Numbers, and Global Learned Enterprise Patterns. To
avoid unnecessary inter-digit timeout when dialing remote destinations learned through GDPR, the
pattern urgency for learned destinations can be configured per class.
Cisco strongly recommends configuring +E.164 numbers and fixed length +E.164 patterns to be inserted
into local digit analysis as urgent.
For details of how calls to directory URIs and numeric destinations learned through GDPR are routed,
see the section on Routing of SIP Requests in Unified CM, page 14-49.
Is LHS numeric?
no
yes
yes
no
yes
no
Block call
yes
No match
Route based on
RHS
348610
Route/block based
on rules for
numeric routing
Offer call
14-49
Chapter 14
Dial Plan
alice@cisco.com
routestring: sfo.route
ILS
Exchange
carol@cisco.com
348611
routestring: fra.route
Figure 14-26 shows an example of how a dialed directory URI might be routed by Unified CM. In this
example the bottom Unified CM cluster advertises the local directory URI carol@cisco.com. All local
directory URIs of this Unified CM cluster are advertised under the SIP routestring fra.route. As part of
this information exchange through GDPR, the Unified CM cluster at the top populated its learned
directory URI table with the association of carol@cisco.com to the SIP routestring fra.route. If someone
then places a call from the phone registered in the top cluster to directory URI carol@cisco.com, the
local lookup of directory URI carol@cisco.com will fail because carol@cisco.com is not a local
directory URI. The next step in the routing process is to search for carol@cisco.com in the table of
directory URIs learned through GDPR. This search will find the information learned from the bottom
cluster, and the originating cluster at the top then takes the learned SIP routestring fra.route and tries to
14-50
Chapter 14
Dial Plan
Dial Plan Elements
find a route by matching this SIP routestring fra.route against the configured SIP route patterns
addressed by the calling device's calling search space. A SIP route pattern fra.route is configured and
points to a route list that ultimately leads to the SIP trunk pointing to the target Unified CM cluster. The
originating Unified CM cluster thus routes the call down to the destination Unified CM cluster. The
destination in the sent SIP request will be carol@cisco.com. On the destination cluster, the same routing
logic as shown in Figure 14-25 then tries to match carol@cisco.com against all local directory URIs on
the destination cluster, which leads to a full match and the target device rings.
The above example shows that the SIP route string namespace is completely independent of the directory
URI namespace. There is no requirement to use SIP route strings that are related in any way to the
structure of the namespace used for the host portion of directory URIs. This allows to optimize the SIP
route string namespace based on the desired routing topology. To disambiguate between SIP route
patterns used to directly match on the URI host portion and SIP route patterns used to route directory
URIs based on SIP route strings, Cisco highly recommends using an independent namespace for SIP
route string route patterns (for example, ".route" or ".ils").
In the above example, the SIP route strings chosen basically identify the individual call controls
(fra.route, nyc.route), and the SIP route pattern grid used to route directory URI SIP requests based on
learned SIP route strings uses explicit patterns (fra.route, nyc.route) to create the desired reachability. In
a hierarchical topology, hierarchical SIP route strings (for example, sjc.us.route, nyc.us.route,
fra.de.route, and muc.de.route) might be used together with wildcard SIP route patterns (*.de.route,
*.us.route) routing to the respective aggregating Cisco Unified Communications Manager Session
Management Edition (SME) clusters responsible for the addressed set of Unified CM clusters.
Is RHS the
IP address of a
cluster member?
yes
Analyze LHS
no
Does RHS
match
CFQDN?
yes
no
Does RHS
match
OTLD?
no
Match RHS
against SIP Route
Pattern?
Route or block
yes
Does LHS
find a match?
no
Route or block
292528
yes
Offer call
The first step is to check whether the right-hand side of the SIP URI is an IP address or hostname of any
server that is a member of the Unified CM cluster or matches the Cluster Fully Qualified Domain Name
(CFQDN) configured in Unified CM enterprise parameters. In this case the left-hand side of the URI is
considered to be a local numeric pattern and will be matched against numeric patterns existing in local
digit analysis using the calling device's calling search space.
14-51
Chapter 14
Dial Plan
The next step is to check whether the right-hand side of the SIP URI matches the Organization Top Level
Domain (OTLD) configured in Unified CM enterprise parameters. If this is the case, again Unified CM
will try to route the call numerically using the calling device's calling search space. But if no match can
be found, then routing will fall back to route the call by matching the right-hand side of the SIP URI
against the configured SIP route patterns.
Assuming a Unified CM cluster with cluster members having IP addresses 192.168.10.10,
192.168.10.11, 192.168.20.10, and 192.168.20.11, cluster fully qualified domain name configured as
ucm1.cisco.com, and organization top-level domain configured as cisco.com, then all of the following
SIP URIs would be routed to local directory number 1234:
1234@192.168.10.10
1234@192.168.10.11
1234@192.168.20.10
1234@192.168.20.11
1234@ucm1.cisco.com
1234@cisco.com
Assuming that no local match for 1234 exists, the first five calls would fail immediately while
Unified CM would try to route the sixth call by matching cisco.com against the configured SIP route
patterns.
Numeric matching can result in a match on any type of numeric pattern existing locally. This does not
only include directory numbers and route patterns and other regular numeric patterns, but can also lead
to a match on any numeric pattern learned through GDPR (+E.164 number or pattern, or enterprise
number or pattern). If a GDPR learned destination is matched, this immediately leads to a secondary
lookup matching the SIP route string of the matched GDPR information against configured SIP route
patterns. For the secondary lookup to match the SIP route string, the same calling search space is used
that also has been used for the initial numeric lookup. This behavior can be used to restrict access to
information learned as part of certain GDPR catalogs by defining a CSS that does not provide access to
the SIP route pattern routing the associated SIP route strings.
Note
To be able to reach destinations learned through GDPR, the calling device's calling search space has to
include the partition that the GDPR learned pattern is residing in and also the partition that the SIP route
pattern resides in, which matches the SIP route string associated with the GDPR learned destination.
Cisco VCS Addressing Schemes: SIP URI, H.323 ID, and E.164 Alias, page 14-53
14-52
Chapter 14
Dial Plan
Dial Plan Elements
Cisco VCS Addressing Schemes: SIP URI, H.323 ID, and E.164 Alias
The Cisco TelePresence Video Communication Server (VCS) enables communications using H.323 and
SIP, and it allows any addressing scheme inherently supported by these protocols.
The dialable address formats are:
IPv4/IPv6 address
Endpoints and multipoint devices can be called using IP addressing, either IPv4 or IPv6.
H.323 ID
The H.323 ID is an alphanumeric identifier for H.323 endpoints. It can be any string of alphanumeric
characters. Where SIP and H.323 registration is required for endpoints (dual registration), this alias
usually matches the SIP URI.
E.164 alias
E.164 uses the same numbering scheme as the PSTN. It is an option that can be configured in H.323
(numbering plan used in the PSTN) together with the H.323 ID.
SIP URI
This is an alias that always takes the form username@domain.
ENUM
ENUM dialing allows an endpoint to be contacted by a caller dialing an E.164 number (a telephone
number) even if that endpoint has registered using a different format of alias.
In principle, any SIP URI can be made using E.164 aliases. The username portion of the alias will be the
E.164 number, and the hostname portion will be the domain. When configuring this kind of E.164
mapping using SIP, the alias loses information about the user. In this case, FindMe can be configured
with the proper alias username@domain, thus hiding the complexity of many different addressing
schemes. The FindMe alias can be associated to any dialable device, regardless of its addressing scheme.
14-53
Chapter 14
Dial Plan
A priority order, which defines the sequential order the VCS uses to analyze the rules or transforms
A matching expression (pattern string) against which the dialed pattern is checked
A replacement string, which is the expression used to derive the destination alias
Even though regular expressions allow for complex string manipulation, there are some very common
simple applications. One of the most common string manipulations on VCS occurs by adding or
stripping the domain part of an alias. An example of this is the following:
Alias: 88302
Search rule matching expression (using a regular expression): (\d+)
Search rule replacement string: \1@cisco.com
Following this simple rule, any dialed number arriving at the VCS will be translated into
number@domain. In this case, 88302 will be translated into 88302@cisco.com.
Search rules have the following additional characteristics that can be useful when creating a dialing
scheme:
A target zone (mandatory). The target zone could be the Local Zone for VCS internal calls, or any
other zone as a neighbor, traversal client or server, or DNS zone. It might include a policy server as
well. The destination zone is selected based on the user's dialed pattern.
A source zone (optional). Starting with Cisco VCS release 7.2, it is possible to apply a rule only to
endpoints calling from a specific zone or subzone.
On VCS, there is a difference between an alias matching a pattern and an alias that is found and that
addresses a device able to answer the call.
If an alias is checked against a search rule matching expression, and the expression matches the alias,
VCS will check if the alias exists in the target zone.
If the search rule matches the alias and the alias is found, the call is sent to the target zone.
If the search rule matches the alias but the alias is not found, this means that it does not exist in the target
zone. In this case the behavior of VCS depends on what is configured for the On Successful Match field
of the search rule. If this field is set to stop, the routing engine stops even if the alias is not found, and
the call is sent to the destination zone. If the field is set to continue, the searching process goes on
analyzing the remaining lower priority rules until the alias is found, until a rule matches the alias with
the On Successful Match field set to stop, or until all the rules have been analyzed.
14-54
Chapter 14
Dial Plan
Recommended Design
This behavior is useful when the administrator does not know where a specific alias is, as in the case of
alphanumeric SIP URIs with the same domain registered on multiple call control platforms. As an
example, there might be multiple VCSs inside the same company, sharing the same domain
company.com. A call for user1@company.com cannot be routed properly if the destination VCS is not
known; however, with the routing logic of VCS, it is possible to search for that alias in multiple VCSs
or other call control systems and to send the call only after the alias is found.
Recommended Design
This section provides design guidance and outlines how to implement an end-to-end enterprise dial plan.
When a call enters the system, the destination number and the calling number are accepted in their
local format but are then immediately globalized by the system. For calls originating from endpoints
registered with Unified CM, this is achieved through dialing normalization translation patterns. For
calls inbound on trunks, inbound called party transformations serve the same purpose.
14-55
Chapter 14
Dial Plan
Recommended Design
Once globalized, the called number is used to route the call to its destination through the use of route
patterns expressed in the global form. The global form may be a combination of a global internal,
enterprise-specific form such as 81001234 and/or a globalized PSTN representation of a DID
number, such as the +E.164 form (for example, +12125551234).
Once a destination has been identified, the calling and called numbers are localized to the form
required by the endpoint, the network, or the system to which the call is to be delivered.
Represent calling and called party numbers in a global form such as that described by the
International Telecommunications Union's E.164 recommendation.
Present calls to external networks (for example, the PSTN) in a manner compatible with the local
requirements for calling party number, called party number, and their respective numbering types.
Derive the global form of the calling party number on incoming calls from gateways, based on the
calling number digits and the numbering type.
Control the establishment of calls, as well as the initiation of mid-call features, between endpoints
based on policies acting on each endpoint's geolocation, to comply with regulatory requirements in
certain countries.
14-56
Chapter 14
Dial Plan
Recommended Design
Note
Some service providers might not be able to accept calling party numbers representing foreign telephone
numbers, due to either technical limitations of their equipment, company policies, or governmental
regulations. If calling party numbers cannot be accepted by the provider, the provider will either screen
and overwrite the calling party number or reject the call. In some networks two calling party identities
can exist for a call: user provided and network provided.
14-57
Chapter 14
Dial Plan
Recommended Design
14-58
Chapter 14
Dial Plan
Recommended Design
Unified CM offers per-gateway settings for this feature, which allow different prefixes for each
numbering type to be applied to calls entering different gateways. The settings can be configured on the
gateway itself, on the gateway's device pool, or through the cluster-wide service parameters, in order of
precedence. A blank entry signifies that no digits will be prefixed; to inherit the settings from the
lower-precedence setting, the entry must be set to Default.
Due to the global significance of the settings at the service parameter level, Cisco highly recommends
using the settings at the device pool level and allowing these settings to be shared for all gateways
sharing the same device pool, or at the gateway level if only a single gateway exists that has to use the
specific set of transformations. To avoid confusion, Cisco recommends always using only
device-pool-level settings or device-level settings in any given installation and not mixing them (using
device-level settings for some and device-pool-level settings for others).
For all calls within a given numbering type, the prefix and strip-digits operations are applied, with no
consideration for the calling party number originally received.
Note
Calls coming from SIP trunks or from SIP gateways are all associated with calling party numbering type
Unknown.
In particular, the SIP protocol as implemented on SIP gateways and SIP trunks effectively places the
incoming calling party number of all calls in the numbering type Unknown. This prevents Unified CM
from applying different calling party number modifications for different calling party number categories.
Unified CM allows the use of Incoming Calling Party Settings Calling Search Spaces (CSSs) for each
number type. These CSSs are used to apply modifications to the calling party based on Calling Party
Transformation Patterns. These patterns use regular expressions to match a subset of cases, followed by
separate digit manipulation operations for each subset. This new capability enables Unified CM to apply
different calling party number modifications for different calling party number categories. For example,
a SIP trunk used to connect to the PSTN could present calls from local, national, and international parties
with the numbering type set to Unknown; then each call's calling party number would be used to match
a Calling Party Transformation Pattern in the trunk's CSS associated with number type Unknown, thus
allowing Unified CM to apply different calling party number modifications for different calling party
number categories.
Logical Partitioning
Some countries such as India have Telecom regulations requiring an enterprise's voice infrastructure to
use the local PSTN exclusively when connecting calls outside the enterprise. This requires that the voice
system be partitioned logically into two systems: one for Closed User Group (CUG) communications
within the enterprise, and a second one to access the local PSTN. A call from an enterprise user in
location A to another enterprise user in location B could be made within the CUG system; however, a
call from an enterprise user in location A to a PSTN destination, no matter the location, must be made
through local access to the PSTN in location A.
While existing dial plan tools can be used to prevent a call from completing if it were placed between
endpoints outside the CUG, they are not able to prevent new call legs from being established while the
call is in progress. For example, assume that an enterprise user in London, England, calls a co-worker in
Delhi, India, over the enterprise network. Once the call is established, the user in Delhi conferences in
a customer in India, from the same line on which the call from London was received. This mid-call
addition (on the same line) of a destination outside the closed user group is not preventable solely by
using the existing dial plan tools in Unified CM (such as Calling Search Spaces and Partitions).
Unified CM 7.1 and later releases offer logical partitioning functionality, which allows the establishment
and enforcement of policies that apply not only to the initial onset of calls, but also to mid-call features
such as conference and transfer.
14-59
Chapter 14
Dial Plan
Recommended Design
The combination of globalization features available in Unified CM allows the system to accept calls in
the local format preferred by the originating users and carriers, to route the calls on-net using global
representations of the called and calling numbers, and to deliver the calls to phones or gateways in the
local format required by the destination user or network. These three aspects of the dial plan design
approach can be summarized as:
Note
For definitions of Type-A and Type-B phones, see Dial Plan Elements, page 14-13.
The calling party number for calls originating from phones is set to the number configured as the
directory number of the line from which the call originates. Following the concepts of a globalized dial
plan design approach, the calling party information of all calls should be globalized. If the directory
number format is not identical to the format chosen for the globalized internal calling party information
14-60
Chapter 14
Dial Plan
Recommended Design
(+E.164 recommended), then the correct handling of calling party information has to be achieved by
properly globalizing the directory number by using the Caller ID for Calls from this Phone Calling
Party Transformation CSS. This is the recommended way to globalize the calling party information of
calls from phones to +E.164, because this method also is compatible with URI-dialed call flows for
which calling party transformations in translation patterns are not applicable.
In the gateway configuration, configure Call Routing Information > Inbound Calls, where the
quantity of significant digits to be retained from the original called number and the prefix digits to
be added to the resulting string are used to globalize the called number. The prefix digits should be
used to add the applicable + sign and country, region, and city codes.
Place translation patterns in partitions referenced by the gateway's calling search space. The
translation patterns should be configured to match the called party number form used by the trunks
connected to the gateway, and should translate it into the global form. The prefix digits should be
used to add the applicable + sign and country, region, and city codes.
Use the incoming calls called party transformation settings available on the gateway and on the
gateway's device pool. There you can define strip and prefix digit instructions or alternatively
configure a called party transformation calling search space per numbering type. This is the
recommended method.
The globalization of the calling party number should be implemented by using the Incoming Calling
Party Settings configured either on the gateway directly or in the device pool controlling the gateway.
Note
If the administrator sets the prefix to Default, this indicates call processing will use the prefix at the next
level setting (device pool or service parameter). Otherwise, the value configured is used as the prefix
unless the field is empty, in which case there is no prefix assigned.
For example, assume a call is placed to Cisco's US headquarters (+1 408 526 4000) from a US number,
and the call is delivered to a gateway located in San Jose, California. The called number provided to the
gateway is 526 4000. This information is sufficient for the Cisco Unified Communications system to
derive the full destination number for the call. A call delivered by the service provider on this specific
trunk group should inherit an implied country code and area code based on the characteristics of the
trunk group connected to the gateway, which presumes that all destination DID numbers handled by the
trunk group are from the North American Numbering Plan country code (1) and for area code 408.
Therefore, the derived global form of the number is +1 408 526 4000. The calling number provided to
the gateway is 555 1234, with the numbering type set to Subscriber. The numbering type allows the
system to infer the country code and area code from the configured characteristics of the trunk group.
Thus, the system knows that the calling number is +1 408 555 1234.
On a different call, if the calling number is 33158405858 with numbering type International, this is an
indication that the global form of the calling number should represented as +33158405858.
14-61
Chapter 14
Dial Plan
Recommended Design
Note
On some newer phones, the calling party number stored in the missed and received calls directories is
left in its globalized form to allow one-touch dialing from the directories without requiring manual
editing of the directory's stored number string. On other phones, the missed and received calls directories
store the transformed calling party number. To avoid problems with one-touch dialing from directories,
the formats of both the transformed and untransformed calling party number need to match a supported
dialing habit.
Note
Many phone users are becoming accustomed to the globalized form of PSTN numbers, mainly due to the
common use of mobile phones across international boundaries. The system administrator can forgo the
configuration of Calling Party Transformation Patterns for phones if displaying the global form of
incoming numbers is preferred.
14-62
Chapter 14
Dial Plan
Recommended Design
As the call is delivered to the San Francisco gateway, the calling party number matches both Calling
Party Transformation Patterns. However, the first one is a more precise match and is selected to process
the calling party number. Thus, the resulting transformed number is 5551234, with a calling party type
set to Subscriber.
If the gateway had not been able to process the call (for example, if all ports were busy), the call would
have been sent to the Chicago gateway to egress to the PSTN. The Chicago gateway is configured with
the following two Calling Party Transformation Patterns:
As the call is delivered into the Chicago gateway, the calling party number matches only the second
Calling Party Transformation Pattern. Therefore, the resulting calling party number offered to the
gateway is 4155551234, with a calling party number type set to National.
14-63
Chapter 14
Dial Plan
Recommended Design
For example, assume that a call to a San Francisco user (+1 415 555 2222) is routed through a route list
featuring a San Francisco gateway as a first choice and a Chicago gateway as a second choice. The San
Francisco gateway is configured with two Called Party Transformation Patterns:
As the call is delivered to the San Francisco gateway, the called party number matches both of the Called
Party Transformation Patterns. However, the first one is a more precise match and is selected to process
the called party number. Thus, the resulting transformed number is 5552222, with a called party type set
to Subscriber.
If the gateway had not been able to process the call (for example, if all ports were busy), the call would
have been sent to the Chicago gateway to egress to the PSTN. The Chicago gateway is configured with
the following two Called Party Transformation Patterns:
As the call is delivered into the Chicago gateway, the called party number matches only the second
Called Party Transformation Pattern. Therefore, the resulting called party number offered to the gateway
is 4155552222, with a called party number type set to National.
Note
When a call egresses to a gateway, the calling and called party transformation patterns are applied to the
calling and called numbers respectively.
Note
SIP does not offer an indication of the numbering type. Therefore, SIP gateways are not able to receive
an indication of the called or calling party number type set by Unified CM.
Applicable local telephony services such as emergency calls, directory, and operator services
Figure 14-28 shows how to support dialing in the globalized form using local habitual dialing for an
example site in the US.
14-64
Chapter 14
Dial Plan
Recommended Design
Figure 14-28
CSSs
Partitions
SJCInternational
DN
Route Lists
Route Groups
DN
All IP Phone DNs (urgent)
SJCtoE164
1XXX, Prefix +1408555
9.[2-9]XXXXXX, Pre-Dot, Prefix +1408
UStoE164
9011.!, Urgent, Pre-Dot, Prefix +
9011.!#, Urgent, Pre-Dot, Prefix +
9.1[2-9]XX[2-9]XXXXXX,
Pre-Dot, Prefix +
PSTNInternational
\+[^1]!
\+[^1]!#
PSTN route
patterns
USPSTNNational
\+1XXXXXXXXXX, Urgent
LOC RL
Local
Route
Group
SJCPSTNLocal
348612
\+1408[2-9]XXXXXX, Urgent
In Figure 14-28, a US IP phone user dials 9011496100773, connects to the destination in Germany, and
then releases the call. The called party in Germany calls the US user back, connects, and then releases
the call. The US user then goes into the Received calls directory, selects the entry for the last received
call (+49 6100 773), and presses Dial.
In this example, the US user initiates two separate calls to the same destination (+496100773). For the
first call, the form of the destination number localized for US dialing habits is used, and the
corresponding translation pattern 9011.! is matched by the user's input. Once translated, the same calling
search space is used for the secondary lookup (Use Originator's Calling Search Space set on the
translation pattern) and the route pattern \+[^1]! is used to route the call. For the second call, the
globalized form of the destination number is used and the route pattern \+[^1]! is used directly.
Comparing these call flows clearly shows the two-step routing process implemented in this dial plan
approach: first normalize all dialing habits to +E.164 and then route based on +E.164 patterns. The
effective PSTN access level is defined by the PSTN route patterns addressed by the calling search space.
More granular access levels can be implemented by adding more specific route patterns.
All directory numbers in partition DN are configured as urgent DNs to avoid potential inter-digit timeout
if an on-net destination is called, and the dialed on-net destination overlaps with the variable length
off-net route pattern in partition PSTNInternational.
14-65
Chapter 14
Dial Plan
Recommended Design
The first translation in partition SJCtoE164 implements 4-digit intra-site dialing, assuming that all local
DIDs of the site are in the range +1 408 555 1XXX. Local dialing (9+7) for the site in San Jose is
implemented by the second translation pattern in the same partition by again transforming the local
habitual dialing to +E.164. The same is true for partition UStoE164, which implements the globalization
of US habitual PSTN dialing to international and national destinations.
All dialing normalization translation patterns have Use Originator's Calling Search Space set (CSS
Inheritance) so that the calling search space used for the secondary lookup, after applying the called
party transformations defined in the translation pattern, is identical to the activating calling search space.
The single calling search space creating the requested class of service can be used as a line or device
calling search space. In deployments that support mobility features such as extension mobility or device
mobility, the line calling search space has to be used to enable the user to keep his class of service when
roaming.
A user with extension +1 408 555 1234 can now be reached from other users using the calling search
space in the example by dialing:
1234 Translation pattern in partition SJCtoE164 transforms dialed digits to +14085551234, and
then there is a match on the directory number in partition DN.
95551234 Translation pattern in partition SJCtoE164 globalizes the dialed digits, and then the
directory number in partition DN is matched.
914085551234 Translation pattern in partition UStoE164 globalizes the dialed digits, and then
the directory number in partition DN is matched.
14-66
Chapter 14
Dial Plan
Recommended Design
CSSs
Partitions
SJCNational
DN
Route Groups
DN
All IP Phone DNs (urgent)
SJCtoE164
1XXX, Prefix +1408555
9.[2-9]XXXXXX, Pre-Dot, Prefix +1408
UStoE164
9011.!, Urgent, Pre-Dot, Prefix +
9011.!#, Urgent, Pre-Dot, Prefix +
9.1[2-9]XX[2-9]XXXXXX,
Pre-Dot, Prefix +
PSTN route
patterns
Route Lists
USPSTNNational
\+1XXXXXXXXXX, Urgent
SJCPSTNLocal
\+1408[2-9]XXXXXX, Urgent
LOC RL
Local
Route
Group
348613
Figure 14-29
Having access to dialing normalization patterns for the local habitual international dialing 9011 is
required even for class of service "national" because we need to support international dialing to
international on-net destinations (directory numbers in partition DN outside the US).
More restrictive classes of service such as "local" and "internal" are built following the same schema of
simply removing access to the partitions holding the inappropriate PSTN route patterns.
The naming convention used for partitions and calling search spaces in the preceding illustrations helps
to identify which pieces of the dial plan need to be replicated to support multiple classes of service, sites,
and dialing domains. If the name includes the specification of a site (for example, SJC in partition name
SJCtoE164), then that element needs to be replicated for every site. If the name includes the specification
of a class of service (for example, International in SJCInternational), then that element needs to be
14-67
Chapter 14
Dial Plan
Recommended Design
replicated for every class of service. If the name does not include the specification of a site (for example,
partition USPSTNNational), then it can be reused for all sites sharing the same dialing habit (in this case
all sites in the US).
Avoid PSTN Hairpinning for Unassigned DNs and Support for Non-+E.164 DNs
CSSs
SJCNational
DN
Partitions
Route Lists
Route Groups
DN
All IP Phone DNs
E164OnNet
+E.164 patterns of DN ranges
(urgent)
SJCtoE164
348614
14-68
Chapter 14
Dial Plan
Recommended Design
Another potential purpose of the intercept patterns in E164OnNet is to map from +E.164 to the format
of the directory numbers. For example, if the directory numbers are configured as E.164 (without the
plus) for a site with DID range +1 408 555 1XXX, a translation pattern \+.14085551XXX with called
party transformation "strip pre-dot" (removing the +) would need to be configured in E164OnNet.
Although it is highly recommended to configure directory numbers as +E.164, in some cases the
directory numbers might be configured in a different globalized format such as E.164 (without the +),
an abbreviated enterprise numbering scheme, or 10 digits in the US. Not configuring directory numbers
as +E.164 requires additional number normalization to be configured for globalized caller IDs. Also,
some CTI applications (for example, attendant console applications) might require additional number
normalization if configured directory numbers do not match the format of numbers stored in global
directories.
If +E.164 DNs are used and the rare hairpinning of on-net calls to unassigned DNs is not considered
critical, the effort to maintain the list of on-net DN ranges in partition E.164OnNet should be avoided
and the simplified dial plan approach shown in Figure 14-28 and Figure 14-29 should be deployed.
Emergency Calls
Access to emergency services has to be granted to all users. This can be achieved either by adding the
partition with the emergency number route patterns to each calling search space or by enabling access
to the emergency number route patterns through the device-level calling search space. If access to
emergency numbers is granted through the device calling search space, then in roaming scenarios (for
example, extension mobility) the user has to dial emergency services using the habitual dialing of the
visited site, while access to emergency numbers through the line calling search space would allow the
user to dial emergency services using the habitual dialing of the home site. This differentiation obviously
is important only if the habitual dialing of emergency services differs between home and visiting sites
as, for example, in the case of a European user (emergency number 112) logging into an US phone
(emergency number 911).
Typically the recommend method is to provide emergency calling services via the emergency number
local to the physical location of the calling device. Although this might create overlap between the
emergency number and other dialing habits (for example, between 911 and four-digit intra-site dialing
starting with 9 for a non-US user from a site with 9XXX abbreviated dialing who logs into a phone in
the US), this at least guarantees that any phone in a given location at any time is allowed to place
emergency calls using the local habitual emergency dialing independent of whether a remote user from
a region with a different emergency number is logged in or not.
To implement this behavior, the emergency patterns needs to be addressed by the device calling search
space.
Simplified configuration of call routing, especially when considering local egress to the PSTN
14-69
Chapter 14
Dial Plan
Recommended Design
Adaptive call routing for speed dials originating from roaming extension mobility users or
roaming devices
One-touch dialing from phone directory entries, including dual-mode phones
One-touch dialing from missed and received call lists in IP phone directories
Note
The AAR mask or the external phone number mask must be configured in the globalized form on the
called destinations to enable AAR, even if the called directory number might already be configured in
the globalized form. AAR will be activated only if either the AAR mask or the external phone mask is
configured.
Furthermore, in most cases the sole function of the AAR CSS is to route the call to the calling phone's
co-located gateway; therefore, it can be configured with only a single route pattern (\+!) pointing to a
route list that contains the Standard Local Route Group. Because calls routed by this single route pattern
will always be routed through the Local Route Group associated with the calling endpoint, that unique
AAR CSS can be used by all phones at all sites, no matter in which region or country they are located.
The 911 CTI route point to Call Forward No Answer (CFNA) and Call Forward Busy (CFB) to 912,
through a calling search space that contains the partition of the 912 CTI route point
The 912 CTI route point to CFNA and CFB to 911, through a calling search space that contains a
global partition, itself containing a route pattern 911 pointing to a route list that contains the
Standard Local Route Group
If both CTI route points become unregistered, calls to 911 will be forwarded through the local route
group as determined by the calling phone's device pool. If Device Mobility is configured, roaming
phones will be associated with the visited site's device pool, and thus associated with the visited site's
Local Route Group.
14-70
Chapter 14
Dial Plan
Recommended Design
Site
DID Range
Enterprise Alternate
Number Mask
Enterprise Alternate
Number Partition
SJC
+14085551XXX
1XXX
SJCToE164
RTP
+19195552XXX
2XXX
RTPToE164
NYC
+12125551XXX
1XXX
NYCToE164
Using the settings shown in Table 14-4 for directory numbers +14085551234 and +12215551234, the
exact same enterprise alternate number 1234 would be created, but both are in different partitions so that
the site specificity is preserved.
Although the schema shown in Table 14-4 demonstrates how GDPR enterprise alternate numbers added
to local digit analysis can be used to implement abbreviated intra-site dialing without adding dialing
normalization translation patterns for this dialing habit, enterprise alternate numbers with only local site
significance should never be advertised across GDPR. On the receiving cluster, overlapping (and
possibly even identical) enterprise alternate numbers would need to be learned, which causes routing
ambiguities.
Note
Cisco highly recommends advertising only enterprise alternate numbers with global significance over
GDPR. Typically these enterprise alternate numbers follow an enterprise abbreviated on-net numbering
plan.
14-71
Chapter 14
Dial Plan
Recommended Design
Table 14-5 shows a potential enterprise alternate number schema based on an enterprise abbreviated
on-net numbering plan using 8 as the access code and two-digit site numbers.
Table 14-5
Site
DID Range
Enterprise Alternate
Number Mask
Enterprise Alternate
Number Partition
SJC
+14085551XXX
8011XXX
DN
RTP
+19195552XXX
8022XXX
DN
NYC
+12125551XXX
8031XXX
DN
These enterprise alternate numbers now have global significance and thus can simply be added into the
DN partition implementing the abbreviated inter-site dialing habit for all local directory numbers. The
traditional approach to implement the equivalent abbreviated inter-site on-net dialing habit based on
dialing normalization translation patterns is shown in Figure 14-31.
Figure 14-31
CSSs
Partitions
SJCInternational
Route Lists
Route Groups
DN
All IP Phone DNs (urgent)
SJCtoE164
DN
OnNet
8011XXX, Mask +14085551XXX
8022XXX, Mask +19195552XXX
348615
Both schemes (adding enterprise alternate numbers to local digit analysis or using dialing normalization)
implement equivalent user experiences. The only difference again is that with dialing normalization
patterns, calls to unassigned numbers dialed using this overlay dialing habit are routed to the PSTN and
then are hairpinned back. On the other hand, adding explicit enterprise alternate numbers for each
directory number to local digit analysis enlarges the local dial plan significantly, which might add
complexity to troubleshooting the local dial plan.
Similar to enterprise alternate numbers, +E.164 alternate numbers are also defined by masking the
directory number. To define a +E.164 alternate number for a +E.164 DN, the mask simply can be left
empty. A +E.164 alternate number of a +E.164 DN should obviously not be added to the local dial plan,
but it is still required to be able to advertise the +E.164 alternate number or a +E.164 PSTN failover
number to remote call controls.
14-72
Chapter 14
Dial Plan
Recommended Design
Using dialing normalization translation patterns to implement abbreviated on-net dialing habits, instead
of defining enterprise alternate numbers for each directory number, reduces the complexity of digit
analysis because fewer patterns are actually added into the digit analysis. To the same extent, advertising
+E.164 and enterprise alternate patterns instead of individual alternate numbers per directory, minimizes
the number of advertised dial plan elements and thereby reduces the complexity of dial plans of remote
call controls that import the advertised information from GDPR. Advertising only summaries in the form
of +E.164 and enterprise patterns is highly recommended.
Call controls that learn dial plan information from GDPR can put the learned information into different
partitions based on the type: +E.164 alternate number, enterprise alternate number, +E.164 pattern, and
enterprise pattern. If this type-based differentiation is not required to implement the required classes of
service, then all numeric dial plan information learned from GDPR can be put into a single partition
(such as the OnNet partition in Figure 14-31) that then is added to all calling search spaces implementing
classes of service with access to remote on-net destinations.
Differentiated class of service can also be achieved based on limiting access to the SIP route patterns
creating the routing schema for the location information in the form of SIP route strings advertised over
GDPR. This allows for limiting the reachability of destinations advertised by certain call controls or as
part of certain imported GDPR catalogs based on the reachability of the advertised SIP route strings.
H.323 legacy endpoints can be registered to the VCS, which will perform protocol conversion and
content interoperability between H.323/H.239 and SIP Binary Floor Control Protocol (BFCP). Note that
VCS behaves as a signaling and media gateway in this scenario, and as such it has to handle the media
too, therefore the Interworking feature has to be turned on.
H.323 endpoints connected to the VCS share the same numbering plan used by Unified CM.
Alias manipulation and normalization is done on VCS using the standards-based Portable Operating
System Interface for Unix (POSIX) format for regular expression syntax. POSIX is a collection of
standards that define some of the matching and replacement functionality that an operating system
(UNIX) should support.
Figure 14-32 shows an example topology for interconnecting Cisco VSC and Unified CM to enable
end-to-end communications between voice and video endpoints registered with Unified CM and VCS
and also communications peers outside the enterprise via VCS Expressway.
14-73
Chapter 14
Dial Plan
Recommended Design
Figure 14-32
B2B
VCS Expressway
VCS
+14085551XXX
cisco.com
H.323
Third-party
SIP
+14085551XXX
+14085551XXX
cisco.com
*.*
+14085551XXX
348616
(?!.*@%localdomains%.*$).*
14-74
Chapter 14
Dial Plan
Recommended Design
Each H.323 client registered to the VCS dialing the internal range would match this rule.
If a +E.164 call comes into VCS from Unified CM, the address dialed would be in any of the following
forms:
+14085551XXX@cisco.com:5060 (@ followed by the domain and port number 5060 or 5061 if the
trunk configuration on Unified CM uses a domain name and a DNS SRV record)
The recommended way to configure the trunk from Unified CM to VCS is to use IP addresses on
Unified CM to define VCS as a peer.
The pattern string (\+14085551\d{3})(@.*) in Example 14-5 matches all three of the above formats, and
the defined replacement string strips the right-hand side of the received SIP URI to make sure that the
received +E.164 address can successfully be matched against the H.323 IDs configured on VCS.
It is possible to use a more stringent pattern matching if a better pattern selection is needed. For example,
([^@]*)@(%ip%|[^@]*cisco.com(.*)). This pattern would match all URIs starting with a sequence of
characters that do not include the @, followed by the @ and the IP address of any of the VCS peers in
the VCS cluster, or anything that includes "cisco.com" and the port number.
If some SIP endpoints are also registered to VCS, they will automatically add the domain. The search
rules above strip the domain even in this case.
For numeric +E.164 calls routed from VCS to Unified CM, a domain has to be added to the SIP URI in
the outgoing request because H.323 endpoints do not automatically add a domain. A search rule has to
be created in order to add a domain for calls sent to Unified CM, as illustrated in Example 14-6.
Example 14-6 Search Rule "To UCM"
Search Rule "To UCM"
Description: To UCM +E164
Priority: 100
mode: alias pattern match
pattern type: regex
pattern string: (\+14085551\d{3})(.*)
pattern behavior: replace string: \1@cisco.com
On successful Match: Stop
Target: UCM Zone
The search rule in Example 14-6 makes sure that all numeric dialing from VCS matching
\+14085551XXX but not matched by any local client is sent to Unified CM and that the host portion of
the SIP URI sent to Unified CM is set to "cisco.com". According to the SIP routing mechanisms of
Unified CM as documented in the section on Routing of SIP Requests in Unified CM, page 14-49, and
14-75
Chapter 14
Dial Plan
Recommended Design
especially Figure 14-27, the organization top level domain (OTLD) on Unified CM has to be set to
"cisco.com" so that Unified CM routes these numeric SIP URIs according to the numeric +E.164 dial
plan configured on Unified CM. This rule is also matched by SIP endpoints registered to VCS, if there
are any.
To enable B2B connectivity, VCS has to route all B2B calls identified by having a non-local SIP URI
host portion to VCS Expressway in the B2B building block. The search rule in Example 14-7
accomplishes this by matching everything that has a domain other than cisco.com and sending it to the
VCS Expressway and to the Internet.
Example 14-7 B2B Search Rule
Search Rule "External"
Description: for B2B
Priority: 110
mode: alias pattern match
pattern type: regex
pattern string: [^@]*@[^@]*(?<!cisco.com)
pattern behavior: leave
On successful Match: Stop
Target: VCS-E
On Unified CM the +E.164 prefix hosted on VCS has to be added by adding a specific +E.164 route
pattern to the Unified CM dial plan and making sure that this route pattern addresses the trunk to VCS
by means of an appropriate route list and route group configuration.
If endpoints registered to VCS share the same DN range than the endpoints registered to Unified CM,
then the dial plan configuration on Unified CM has to ensure that all +E.164 numbers from the local
prefix that are unknown on Unified CM are routed to VCS. Figure 14-33 shows how this can be achieved
with a globalized dial plan approach.
Figure 14-33
CSSs
Partitions
Route Lists
Route Groups
DN
All IP Phone DNs
E164OnNet
+E.164 patterns of DN ranges
(urgent)
DN
UnassignedDN
RL_VCS
RL_VCS
348617
\+14085551XXX, Urgent
VSCInbound
The globalized dial plan in Figure 14-33 uses the approach as discussed in the section on Globalized Dial
Plan Approach on Unified CM, page 14-55. Simply put, the DN calling search space referenced by all
dialing normalization translation patterns and the urgent translation patterns matching on the known
on-net +E.164 prefixes, has to be extended to include a route pattern matching on the +E.164 prefix that
14-76
Chapter 14
Dial Plan
Recommended Design
is shared with VCS. All +E.164 patterns from this range not matched by directory numbers on
Unified CM will be matched by this route pattern and sent to VCS. To make sure that no routing loop is
created, the inbound calling search space of the trunk coming from VCS should not have access to this
route pattern pointing back.
Also, the dial plan on Unified CM has to make sure that all calls dialed as a URI (non-numeric) that do
not address a directory URI local to Unified CM, are routed to VCS. The easiest way to achieve this is
to add a "catch-all" SIP route pattern (for example, *.*) on Unified CM that also addresses the trunk to
VCS through an appropriate route list and route group configuration. Again, to make sure that routing
loops are avoided, the inbound calling search space of the trunk coming from VCS should not have
access to this "catch-all" SIP route pattern.
If the H.323 endpoints are also addressed using alphanumeric aliases of the same form of a SIP URI
instead of using +E.164 aliases, the "To VCS" search rule in Example 14-5 can be replaced by the one
in Example 14-9.
Example 14-9 Modified Search Rule "To VCS" Supporting URI Dialing of H.323 Registered Endpoints
Search Rule "To VCS"
Description: To Local H.323 aliases
Priority: 50
mode: alias pattern match
pattern type: suffix
pattern string: cisco.com
pattern behavior: leave
On successful Match: Continue
Target: Local Zone
"Continue" has to be enabled because, if the alias is not found in the Local Zone, this means that the alias
is not local and it will be sent to Unified CM following the next rule of priority 100 ("To CUCM").
However, if the call comes from Unified CM, it will not be sent back to the CUCM zone where the call
came from, thus prohibiting routing loops.
On Unified CM a SIP route pattern has to be created to match on "cisco.com" pointing to the same route
list used for +E.164 routing to VCS.
14-77
Chapter 14
Dial Plan
Special Considerations
If a user on Unified CM dials alice@cisco.com, Unified CM will first match this URI against the locally
configured SIP URIs and then as a fallback will match the host portion (cisco.com) against the
configured SIP route patterns so that the above SIP route pattern is matched and the call is routed to VCS.
If the URI is known on VCS, the call is routed to the endpoint, but the call will not be sent back to
Unified CM if the URI is unknown because the call comes from that zone and has not been manipulated.
Special Considerations
This section describes dial plan considerations related to a number of Cisco Unified CM features,
including:
Note
AAR does not support CTI route points as the origin or the destination of calls. Also, AAR is
incompatible with the Extension Mobility feature when users roam across different sites. Refer to
Extension Mobility, page 14-83, for more details.
You must provide the following main elements for AAR to function properly:
14-78
Chapter 14
Dial Plan
Special Considerations
Note
By default, the directory number configuration retains the AAR leg of the call in the call history, which
ensures that the AAR forward to the voice messaging system will select the proper voice mailbox. If you
choose "Remove this destination from the call forwarding history," the AAR leg of the call is not present
in the call history, which would prevent the automated voice mailbox selection and would offer the caller
the generic voicemail greeting.
The AAR Destination Mask is used to allow the destination phone number to be determined
independently of the External Phone Number mask. For example, if Caller ID policy for a company
required a phone's external phone number mask to be the main directory number of an office (such as
415 555 1000), the AAR destination mask could be set to+1 415 555 1234, to provide AAR with the
phone's specific PSTN number.
For example, assume phone A in San Francisco (DN = 2345) dials an on-net DN (1234) configured on
phone B located in New York. If locations-based call admission control denies the call, AAR retrieves
the AAR Destination Mask of the New York phone (+1212555XXXX) and uses it to derive a number
(+12125551234) that can be used to route the call on the PSTN.
It is best to configure the AAR destination mask to yield a fully qualified E.164 number, including the
+ sign, because this will greatly simplify the overall configuration of AAR. For example, a phone in
Paris is configured with an AAR destination mask of +33 1 58 04 58 58. Because this number is a fully
qualified E.164 number, it contains all the information required for the Cisco Unified Communications
system to derive a routable PSTN number as required by the calling phone's gateway to the PSTN,
regardless of whether it is located in France, in Canada, or anywhere else in the world. The following
sections elaborate on this approach.
This is the simplest case; the AAR destination contains + as a wildcard to be replaced by the appropriate
access codes require at each gateway. The destination number is ready to be routed to an appropriate
route pattern and then transformed at the point of egress to the PSTN by the appropriate called party
transformation patterns.
Example 1: A phone in Ottawa, Canada calls a phone in Paris, which triggers AAR due to a lack of
bandwidth on the WAN. The AAR destination is +33 1 58 04 58 58. The AAR calling search space of
the calling phone contains a route pattern \+!, which routes the call to the Standard Local Route Group.
The call is routed to the local gateway in Ottawa, where called party transformation patterns will replace
the + with the applicable international access code 011. The resulting call is placed to 011 33 1 58 04 58
58.
14-79
Chapter 14
Dial Plan
Special Considerations
Example 2: A phone in Nice, France calls a phone in Paris, which triggers AAR due to a lack of
bandwidth on the WAN. The AAR destination is +33 1 58 04 58 58. The AAR calling search space of
the calling phone contains a route pattern \+!, which routes the call to the Standard Local Route Group.
The call is routed to the local gateway in Nice, where called party transformation patterns will replace
the + 33 with the applicable national access code 0. The resulting call is placed to 01 58 04 58 58.
If the AAR Destination Mask Yields a Number Including the Country Code
The destination number (assumed to include the country code) might require a prefix to be routed
properly by the origination branch's dial plan. Furthermore, if the point of origin is located in a different
area code or even a different country, then other prefixes such as international dialing access codes (for
example, 00 or 011) might be required as part of the dialed string.
When configuring AAR, you place the DNs in AAR groups. For each pair of AAR groups, you can then
configure prefix digits to add to the DNs for calls between the two groups, including prefix digits for
calls originating and terminating within the same AAR group.
As a general rule, place DNs in the same AAR group if they share the same inter-country dialing
structure. For example, all phones in the UK dial numbers outside the UK with 9 as a PSTN access code,
followed by 00 for international access; all phones in France and Belgium use 0 as a PSTN access code,
followed by 00 for international access; all phones in the NANP use 9 as a PSTN access code, followed
by 011 for international access.
This yields the following AAR group configuration:
AAR Group
NANP
Cent_EU
UK
NANP
9011
9011
Cent_EU
000
000
000
UK
900
900
Example 3: A phone in Ottawa, Canada calls a phone in Paris, which triggers AAR due to a lack of
bandwidth on the WAN. The AAR destination is 33 1 58 04 58 58. The AAR group of the calling phone
is NANP and that of the destination phone is Cent-EU, thus yielding a prefix of 9011. The AAR calling
search space of the calling phone contains a site-specific route pattern 9011!, which routes the call to a
route list in Ottawa, stripping the 9. The call is routed to the local gateway in Ottawa. The resulting call
is placed to 011 33 1 58 04 58 58.
Example 4: A phone in Brussels, Belgium calls a phone in Paris, which triggers AAR due to a lack of
bandwidth on the WAN. The AAR destination is 33 1 58 04 58 58. The AAR group of the calling phone
and that of the destination phone is Cent-EU, thus yielding a prefix of 000. The AAR calling search space
of the calling phone contains a site-specific route pattern 000!, which routes the call to a route list in
Brussels, stripping the leading 0. The call is routed to the local gateway in Brussels. The resulting call
is placed to 00 33 1 58 04 58 58.
These examples clearly show the benefit of a +E.164 dial plan where no specific AAR groups need to
be configured.
These examples clearly show the benefit of a dial plan with +E.164 directory numbers. No specific AAR
groups or PSTN prefixes need to be configured. The dialed on-net destinations are already in a format
(+E.164) used by the core routing of the dial plan, so that the dialed directory number can be used
directly as the PSTN address for the alternate call.
14-80
Chapter 14
Dial Plan
Special Considerations
Voicemail Considerations
AAR can direct calls to voicemail. The voicemail pilot number is usually dialed without the need for an
off-net access code (if the voicemail pilot number is a fully qualified on-net number, such as 8 555 1000).
When AAR is configured to send calls to voicemail, the AAR group mechanism will still prefix the
configured access code(s). This configuration requires the creation of an AAR group to be used by all
DNs whose desired AAR destination is voicemail (for example, vmail_aar_grp). Ensure that the
configuration for this voicemail AAR group uses no prefix numbers when receiving calls from other
AAR group DNs.
For example: Assume that DNs located in sites San Francisco and New York are configured with AAR
group NANP, which prefixes 9 to calls made between any two DNs in the group. If a DN in San Francisco
is configured to send AAR calls to voicemail (for example, 8 555 1000), a call would be placed to
985551000, which would result in a failed call. Instead, the San Francisco DN should be configured with
AAR group vmail. The prefix digits for calls from AAR group NANP to AAR group vmail are <none>,
as shown in the following table. The call will be placed successfully to 85551000.
Note
AAR Group
NANP
Cent_EU
UK
vmail
NANP
9011
9011
<none>
Cent_EU
000
000
000
<none>
UK
900
900
<none>
When Device Mobility is not used, the AAR group configuration of a DN remains the same even as the
device is moved to different parts of the network. With Device Mobility, the AAR group can be
determined dynamically based on where in the network the phone is physically located, as determined
by the phone's IP address. See Device Mobility, page 14-82, for more details.
14-81
Chapter 14
Dial Plan
Special Considerations
Note
If you have configured additional route patterns to force on-net internal calls dialed as PSTN calls,
ensure that these patterns are not matched by the AAR feature. In a globalized dial plan with +E.164
directory numbers, the partition holding these +E.164 directory numbers must not be part of the AAR
calling search space.
Note
To avoid denial of re-routed calls due to call admission control, AAR functionality requires the use of a
LAN as the IP path between each endpoint and its associated gateway to the PSTN. Therefore, AAR dial
plans cannot rely on centralized gateways for PSTN access.
Note
When Device Mobility is configured, the AAR calling search space can be determined dynamically
based on where in the network the phone is physically located, as determined by the phone's IP address.
See Device Mobility, page 14-82, for more details.
Device Mobility
Device Mobility offers functionality designed to enhance the mobility of devices within an IP network.
(For example, a phone initially configured for use in San Francisco is physically moved to New York.)
Although the device still registers with the same Unified CM cluster, it now will adapt some of its
behavior based on the new site where it is located. Those changes are triggered by the IP subnet in which
the phone is located.
When roaming, a phone will inherit the parameters associated with the device pool associated with the
device's current subnet. From a dial-plan perspective, the functionality of the following five main
configuration parameters can be modified due to the physical location of the phone. For these parameters
to be modified, the device must be deemed as roaming outside its home physical location but within its
home device mobility group.
14-82
Chapter 14
Dial Plan
Special Considerations
When a device is roaming in the same device mobility group, Unified CM uses the Device Mobility CSS
to reach the local gateway. If a user sets Call Forward All at the phone, if the CFA CSS is set to None,
and if the CFA CSS Activation Policy is set to With Activating Device/Line CSS, then:
The Device CSS and Line CSS get used as the CFA CSS when the device is in its home location.
If the device is roaming within the same device mobility group, the Device Mobility CSS from the
Roaming Device Pool and the Line CSS get used as the CFA CSS.
If the device is roaming within a different device mobility group, the Device CSS and Line CSS get
used as the CFA CSS.
The section on Device Mobility, page 21-13, explains the details of this feature.
Extension Mobility
The Extension Mobility feature enables a user to log in to an IP phone and automatically apply his or
her profile to that phone, including extension number, speed dials, message waiting indicator (MWI)
status, and calling privileges. This mechanism relies on the creation of a device profile associated with
each Extension Mobility user. The device profile is effectively a virtual IP phone on which you can
configure one or more lines and define calling privileges, speed dials, and so on.
When an IP phone is in the logged-out state, (that is, no Extension Mobility user has logged into it), the
phone characteristics are determined by the device configuration page and the line configuration page(s).
When a user logs in to an IP phone, the device configuration does not change, but the existing line
configuration is saved in the Unified CM database and is replaced by the line configuration of the user's
device profile.
One of the key benefits of Extension Mobility is that users can be reached at their own extensions
regardless of where they are located, provided that they can log in to an IP phone controlled by the same
Unified CM cluster. When Extension Mobility is applied to multisite deployments with centralized call
processing, this capability is extended to multiple sites geographically separated from each other.
14-83
Chapter 14
Dial Plan
Special Considerations
However, if you combine the Extension Mobility feature with the AAR feature described in the section
on Automated Alternate Routing, page 14-78, some limitations exist. Consider the example shown in
Figure 14-34, where Extension Mobility and AAR are deployed in a centralized call processing
Unified CM cluster with one site in San Jose and one in New York.
Figure 14-34
IP WAN
San Jose
PSTN
New York
PSTN
New York
San Jose
DN: 1000
Ext. mask: 4085551000
EM user
moves
IP
DN: 2000
Ext. mask: 2125552000
IP
IP
114718
DN: 1001
Ext. mask: 4085551001
IP
In this example, assume that an Extension Mobility user who is normally based in San Jose has a DN of
1000 and a DID number of (408) 555-1000. That users external phone number mask (or AAR mask, if
used) is therefore configured as 4085551000. The user now moves to the New York site and logs in. Also,
assume that the IP WAN bandwidth between San Jose and New York has been entirely utilized.
When the user in San Jose with extension 1001 tries to call 1000, AAR is triggered and, based on the
AAR calling search space of the calling party and the AAR groups of both parties, a new call to
914085551000 is attempted by the San Jose phone. This call uses the San Jose gateway to access the
PSTN, but because the DID (408) 555-1000 is owned by that same gateway, the PSTN sends the call
back to it. The San Jose gateway tries to complete the call to the phone with extension 1000, which is
now in New York. Because no bandwidth is available to New York, the AAR feature is invoked again,
and one of the following two scenarios will occur:
Tip
If the gateway's AAR calling search space contains external PSTN route patterns, this is the
beginning of a loop that eventually uses all the PSTN trunks at the San Jose site.
If, on the other hand, the gateway's AAR calling search space contains only internal numbers, the
call fails and the caller hears a fast-busy tone. In this case, one PSTN call is placed and one is
received, so two PSTN trunks are utilized on the San Jose gateway for the duration of the call setup.
To prevent routing loops such as the one described here, always configure all calling search spaces on
the gateway configuration pages to include only internal destinations and no route patterns pointing to
route lists or route groups containing that same gateway.
This example highlights the fact that Extension Mobility leverages the dynamic aspect of Cisco IP
Communications and, therefore, requires that the call routing between sites use the IP network. Because
the E.164 numbers defined in the PSTN are static and the PSTN network is unaware of the movements
of the Extension Mobility users, the AAR feature, which relies on the PSTN for call routing, cannot be
used to reach Extension Mobility users who move to a site other than their home site.
14-84
Chapter 14
Dial Plan
Special Considerations
Note
However, if the Extension Mobility user moves to a remote site that belongs to the same AAR group as
his or her home site, he or she can use the AAR feature to place calls to other sites when the available
IP WAN bandwidth is not sufficient. This is because the path of such a call is determined by the AAR
calling search space of the phone from which the call originates. This AAR calling search space does
not change when users log in or out of Extension Mobility, and it should be configured to use the visited
remote site's gateway.
Tip
Configure unregistered Extension Mobility profile DNs to send calls to voicemail. See Call-Forward
Calling Search Spaces, page 14-45, for details.
Note
14-85
Chapter 14
Dial Plan
Special Considerations
External Phone Number Mask: 408 555 0000 (This is the enterprise's main business number.)
Line Calling Search Space: R_L_CSS
Device Calling Search Space: R_D_CSS
The following sections explain the effects of the above mobility parameters on call routing.
Note
Note
Even if the dialing privileges assigned to Ringo's phone do not allow for external calls, the call to the
remote destination is handled by the rerouting calling search space associated with Paul's remote
destination profile.
14-86
Chapter 14
Dial Plan
Special Considerations
The following examples assume that the service parameter Inbound Calling Search Space for Remote
Destination is set to Trunk or Gateway Inbound Calling Search Space.
For example:
Paul uses his mobile phone to call Ringo at his desk. The call comes into the gateway from the PSTN,
with a calling party number of 514 000 9876 and a called party number of 408 555 0001. The call is
routed to Ringo's phone. The number displayed as the calling party number on Ringo's phone is Paul's
desk phone number, 8 555 1234. This allows Paul's mobile phone number to remain confidential and
allows Ringo's calls placed from the missed and received calls lists to ring into Paul's IP phone, thus
making the full set of enterprise mobility features available.
When the call comes into the gateway, the PSTN offers a calling party number of 514 000 9876 and a
called party of 408 555 0001. The gateways configuration will retain the last seven significant digits of
the called number and prefix 8, yielding 8 555 0001 as the destination number.
The system detects that the calling party number matches Paul's remote destination number. Upon
detecting this match, the system will:
1.
2.
Route the call to the called number using the incoming gateway's calling search space. Specifically,
the routing is done through the GW_CSS calling search space.
The destination (called) number presented by the gateway should be the DN of the phone, and the calling
party substitution illustrated in step 1 above renders possible the use of one-touch dialing from the
missed/received calls lists.
Note
There is no way to partition remote destination numbers. This is worth noting in case multiple user
groups (such as different companies, sub-contractors, and so forth) are using the same cluster. When the
service parameter Inbound Calling Search Space for Remote Destination is set to Trunk or Gateway
Inbound Calling Search Space, the call routing is based on the incoming trunks or gateway's CSS,
regardless of whether or not the calling number matches a remote destination. However, the calling party
number substitution still occurs if the calling party matches any remote destination. This means that calls
from one tenant's remote destination numbers to another tenant's DID numbers will be presented with a
transformed calling party number that matches the caller's on-net extension DN.
Note
Any incoming external call where Calling Party Number is not available will be routed according to the
incoming gateway's CSS. This also applies to incoming calls from IP trunks, such as SIP or H.323 trunks.
Remote Destination Profile's Calling Party Transformation CSS and Transformation Patterns
Calls originating from an enterprise IP phone to a mobility-enabled DN are forked to both the enterprise
destination IP phone's DN and one (or multiple) external destinations. One challenge this creates is to
deliver calling party numbers adapted to each destination phone's dial plan. This is to allow for redialing
of calls from missed calls and received calls lists. For an enterprise phone, the calling party numbers
should be redialable enterprise phone numbers. For a remote destination on the PSTN (such as a home
phone or a mobile phone), the calling party number should be transformed from the enterprise number
associated with the calling IP phone to a number redialable from the PSTN (generally, the DID number
of the calling phone).
When a call is placed to a mobility-enabled enterprise DN, the associated remote destination profile's
calling party transformation calling search space is used to find a match to the caller's calling party
number. It contains partitions which themselves contain transformation patterns.
14-87
Chapter 14
Dial Plan
Special Considerations
Transformation patterns control the adaptation of calling party numbers from enterprise format to PSTN
format. They differ from all other patterns in Unified CM in that they match on the calling number, not
the called number. The matching process is done through a regular expression (for example, 8 555
XXXX), and the transformation process allows for the optional use of the calling DN's external phone
number mask as well as transformation patterns and digit prefixing.
Once matched, they perform all configured transformations, and the resulting calling party number is
used to reach all remote destinations associated with the Remote Destination Profile for which the match
occurred.
For example:
When Ringo calls Paul, we want Paul's IP phone to display the calling party number as 8 555 0001 and
Paul's mobile phone to display 408 555 0001.
For this case, we create a transformation pattern with the following parameters:
Pattern: 8 555 XXXX
Partition: SJ_Calling_Transform
Use calling party's external phone number mask: un-checked
Calling Party Transformation mask: 555 XXXX
Prefix Digits (outgoing calls): 408
We also have to ensure that partition SJ_Calling_Transform is placed in calling search space
P_CPT_CSS.
When the call from Ringo is anchored on Paul's phone, two separate call legs are attempted. The first
rings Paul's IP phone and offers the caller's DN as Calling Party Number (that is, 8 555 0001). The
second call leg is attempted through Paul's Remote Destination Profile. The RDP's calling party
transformation CSS, P_CPT_CSS, is used to find a match for 8 555 0001 in all the referenced partition's
transformation patterns. Pattern 8 555 XXXX is matched in partition SJ_Calling_Transform. The
transformation mask is applied to the calling party number and yields 555 0001. The prefix digits are
added, and the resulting calling party number 408 555 0001 is used when placing the call to the remote
destinations.
Note that, in this example, we chose not to use the external phone number mask because it is set to a
number different than that of Ringo's DID. This offers flexibility in situations where the calling party
number offered to off-net destinations is required to be different based on the relationship of the caller
to the called party. The call from Ringo to Paul is between co-workers, thus the disclosure of Ringo's
DID number is deemed acceptable. Ringo's next call could be to a customer, in which case the main
enterprise number 408 555 0000 is the desired Calling Party Number to be offered to the destination.
Note
Calling Party Transformation calling search spaces do not implicitly include the <none> partition;
therefore, transformation patterns left in the <none> partition do not apply to any Calling Party
Transformation calling search space. This is different from all other patterns in Unified CM, where all
patterns left in the <none> partition are implicitly part of every calling search space.
14-88
Chapter 14
Dial Plan
Special Considerations
For example:
Assume the number 514 000 9876 is configured as Paul's remote destination number. This corresponds
to the form used by the PSTN to identify calls coming into the enterprise. But it differs from the form
used by the enterprise dial plan for outgoing calls, which requires that 91 be prefixed. In this case, we
need to create an application dial rule to adapt the remote destination form to the enterprise dial plan's
form:
Application Dial Rule:
Name: 514000_ten
Description: Used to prefix 91 to ten-digit numbers beginning with 514000
Number begins with: 514000
Number of Digits: 10
Total digits to be removed: 0
Prefix with Pattern: 91
In this example, calls made from Paul's mobile phone into the enterprise are identified as coming from
514 000 9876. This matches the form in which his number is configured as a remote destination, thus
allowing the match to be made and triggering the anchoring of the call on Paul's desk phone as well as
adapting the Calling Party Number offered to the on-net destination. (For example, when a call is placed
to Ringo's DID number, he sees the call as coming from 8 555 1234.)
When a call is placed to Paul's enterprise DN number, the call leg forked to his remote destination
number will be processed by the application dial rule above. The string 514 000 matches the beginning
of Paul's remote destination number, and it is ten digits long, so no digits are removed and 91 is prefixed.
This yields 91 514 000 9876 as a number to be routed through Paul's Remote Destination Profile calling
search space (P_RDP_CSS in this case).
Note
This approach offers the ability to reuse calling search spaces already defined to route calls made from
IP phones. Creating new calling search spaces not requiring prefixes for outbound calls (that is, ones
able to route calls to 514 000 9876 directly) is less preferable because it can create situations where
external patterns overlap with on-net patterns.
Time-of-Day Routing
To use this feature, configure the following elements:
Time period
Time schedule
The time period allows you to configure start and end times for business hours. The start and end times
indicate the times during which the calls can be routed. In addition to these times, you can set the event
to repeat itself on a weekly or yearly basis. Moreover, you can also configure non-business hours by
selecting "No business hours" from the Start Time and End Time options. All incoming calls will be
blocked when this option is selected.
A time schedule is a group of specific time periods assigned to the partition. It determines whether the
partition is active or inactive during the specified time periods. A matching/dialing pattern can be
reached only if the partition in which the dialing pattern resides is active.
As illustrated in Figure 14-35, two hunt pilots with the same calling pattern (8000) are configured in two
partitions (namely, RTP_Partition and SJC_Partition). Each of these partitions is assigned a time
schedule, which contains a list of defined time periods. For example, RTP phones can be reached using
14-89
Chapter 14
Dial Plan
Special Considerations
Hunt Pilot 1 from 8:00 AM to 12:00 PM EST (GMT - 5.00) Monday through Friday as well as 8:00 AM
to 5:00 PM on Sundays. In the same way, SJC phones can be reached using Hunt Pilot 2 from 8:00 AM
to 5:00 PM PST (GMT - 8.00) Monday through Friday and 8:00 AM to 5:00 PM on Saturdays. Both of
the hunt pilots in this example are inactive on July 4th.
Figure 14-35
Time-of-Day Routing
Unified CM
Cluster
RTP Site
M
M
SJC
Partition
Hunt Pilot 2
DN 8000
Hunt Pilot 1
DN 8000
RTP
Partition
Time Period
9.00 to 17.00
Mon-Fri
Hunt List
Hunt List
Time Period
8.00 to 12.00
Mon-Fri
Time Period
8.00 to 17.00
Sat
Line Group
Line Group
Time Period
No Business
Jul 4th
IP
IP
RTP IP Phones
San Jose
Time Schedule
Time Period
8.00 to 17.00
Sun
IP
Time Period
No Business
Jul 4th
RTP
Time Schedule
IP WAN
IP
San Jose
IP Phones
IP
126916
IP
For the example in Figure 14-35, an incoming call to the hunt pilot (8000) on Wednesday at 3:00 PM
will be forwarded to the SJC phones, while a person calling the hunt pilot on July 4th will get a fast busy
tone unless there is another pattern that matches 8000.
Logical Partitioning
The elements of logical partitioning include:
Device types, where phones are classified as interior, and gateways and trunks are defined as border.
Table 14-6 lists the endpoint types for different devices.
Geolocations, where endpoints are assigned a civic address to be used in policy decisions.
Geolocation filters, where policy decisions can be made on a subset of the geolocation objects.
14-90
Chapter 14
Dial Plan
Special Considerations
Policies, where communications between endpoints are either allowed or denied based on their
comparative (filtered) geolocations and device types.
Note
Policies are not applied if all participants in a call (or call attempt) are classified as interior. This means
that calls between phones on the same cluster are never subjected to logical partitioning policies.
Note
Geolocations are not to be confused with locations configured in Unified CM, which are used for call
admission control, or with physical locations used for Device Mobility.
Table 14-6
Device Types
Interior
H.225 trunk
SIP trunk
Geolocation Creation
The (RFC) 4119 standard provides the basis for geolocations. Geolocations use the civic location format
specified by the following objects:
Name
Description
Neighborhood (A5)
14-91
Chapter 14
Dial Plan
Special Considerations
Note
Street (A6)
Landmark (LMK)
Floor (FLR)
Geolocation Assignment
Devices are assigned a geolocation from either the device page, the device pool, or the default
Geolocation as configured under Enterprise Parameters, in that order of precedence.
14-92
Chapter 14
Dial Plan
Special Considerations
Note
Each set of geolocation objects configured in a policy is considered in association with a single device
type. For example, a set of geolocation objects such as Country=India, State=Karnataka, City=Bangalore
needs to be associated with device type Interior for actions pertaining to Bangalore phones, and
separately associated with device type Border for actions pertaining to Bangalore gateways.
Note
When the geolocation identifiers of two devices are being evaluated by logical partitioning, no policy is
applied if both devices are of device type Interior. This means that no call, conference, transfer, or so
forth, between IP phones within the same cluster will ever be denied due to logical partitioning policies.
For example, consider phones A and B located in Bangalore, India, and gateway C located in Ottawa,
Canada. Phone A calls phone B. Because both devices are of type Interior, no policy is invoked. The call
is established, and then the user at phone A invokes a conference, which would bring in gateway C.
Before the action is allowed, Unified CM will check the geolocation identifiers of A and C, as well as
those of B and C, for a match with the preconfigured policies. If any of the matching policies results in
a deny action, the new call leg cannot be established.
Note
The default policy in Unified CM is deny; in other words, if no policy is configured explicitly to permit
a call leg, the call leg will be denied.
In the example above, unless an explicit policy is configured to allow Bangalore Interior devices to
connect to Ottawa Border devices, the call leg will be denied.
14-93
Chapter 14
Dial Plan
Special Considerations
14-94
CH A P T E R
15
Emergency Services
Revised: June 15, 2015
Emergency services are of great importance in the proper deployment of a communications system. This
chapter presents a summary of the following major design considerations essential to planning for
emergency calls:
This chapter presents some information specific to the 911 emergency networks as deployed in Canada
and the United States. Many of the concepts discussed here are adaptable to other locales. Please consult
with your local telephony network provider for appropriate implementation of emergency call
functionality.
In the United States, some states have already enacted legislation covering the 911 functionality required
for users in a multi-line telephone system (MLTS). The National Emergency Number Association
(NENA) has also produced the NENA Technical Requirements Document on Model Legislation E9-1-1
for Multi-Line Telephone Systems, available online at
http://www.nena.org/
This chapter assumes that you are familiar with the generic 911 functionality available to residential
PSTN users in North America.
Note
The topics discussed in this chapter apply to Cisco Emergency Responder only when it is used in
conjunction with Cisco Unified Communications Manager (Unified CM). Cisco TelePresence Video
Communication Server (VCS) currently does not support emergency services.
15-1
Chapter 15
Emergency Services
Described in
Revision Date
A distressed caller should be able to dial the emergency services from a fixed line, a mobile phone,
a public phone, or any device capable of making the voice call.
An emergency services call handler must be available to respond to the emergency request and
dispatch the needed services such as police, fire, and medical.
In order to provide help, the call handler should be able to identify the location of the distressed
caller as precisely as possible.
An emergency services network is needed to route the call to the nearest emergency services call
handler with jurisdiction for the location of the caller.
The following sections explain some of the important architectural components of 911 emergency
services architecture.
For a given street address, all 911 calls are routed to the same PSAP.
Exceptional Situation
The physical size of the campus puts some of the buildings in different PSAP jurisdictions.
Some of the 911 calls need to be routed to an on-net location (campus security, building security).
15-2
Chapter 15
Emergency Services
911 Emergency Services Architecture
Selective Router
The selective router is a node in the emergency services network that determines the appropriate PSAP
for call delivery, based on callers geographic area and the automatic number identification (ANI). The
Local Exchange Carrier (LEC) usually operates the selective router. Hence, it is imperative to ensure that
the enterprise IP communications network is designed in such a way that the caller is routed to the
appropriate selective router based on its location.
For a given street address, the 911 network service provider is the incumbent Local Exchange
Carrier (LEC). For a location served by Phone Company X, the corresponding PSAP is also served
by Phone Company X.
All 911 calls are routed directly to an off-net location, or all 911 calls are routed directly to an on-net
location.
15-3
Chapter 15
Emergency Services
Exceptional Situation
The local exchange carrier (LEC) through which the MLTS interfaces to the PSTN is not the same
LEC that serves as 911 network service provider to the PSAP. (For example, the communications
system is served by Phone Company X, but the PSAP is connected to Phone Company Y.) This
situation might require either a special arrangement between the LECs or special, dedicated trunks
between the phone system and the PSAP's 911 network service provider.
Some LECs may not accept 911 calls on their networks. If this is the case, the only two options are
to change LECs or to establish trunks (dedicated to 911 call routing) connected to a LEC that can
route 911 calls to the appropriate PSAPs.
Some (or all) of the 911 calls have to be routed to an on-net location such as campus security or
building security. This situation can easily be accommodated during the design and implementation
phases, but only if the destination of 911 calls for each phone has been properly planned and
documented.
San Francisco Police Department and San Jose Police Department are the appropriate PSAPs
San Francisco Police Department and San Jose Police Department are served by the same 911
network service provider
However, San Francisco Police Department and San Jose Police Department are served by different
E911 selective routers operated by that same 911 network service provider!
This type of situation would require two separate interface points, one per E911 selective router. The
information pertaining to the E911 selective router territories is generally kept by the incumbent LEC,
and the local account representative for that LEC should be able to provide an enterprise customer with
the pertinent information. Many LECs also provide the services of 911 subject matter experts who can
consult with their own account representatives on the proper mapping of 911 access services.
Typical Situation
For single-site deployments or campus deployments, there is usually only one PSAP for 911 calls.
If access to only one PSAP is required, then only one interface point is required. Even if access to
more than one PSAP is required, they might be reachable from the same E911 selective router,
through the same centralized interface. If the enterprise's branch sites are linked via a WAN
(centralized call processing), it is desirable to give each location its own local (that is, located inside
each branch office) access to 911 to prevent 911 isolation during WAN failure conditions where
Survivable Remote Site Telephony (SRST) operation is activated.
Exceptional Situation
The physical size of the campus puts some of the buildings in different PSAP jurisdictions, and
Some of the 911 calls have to be routed to different E911 selective routers, through different
interface points.
15-4
Chapter 15
Emergency Services
911 Emergency Services Architecture
Note
Some of the information required to establish the geographical territories of PSAPs and E911 selective
routers is available online or from various competitive local exchange carrier (CLEC) information web
sites. (For example, https://clec.att.com/clec/hb/shell.cfm?section=782 provides some valuable data
about the territory covered by AT&T in California and Nevada.) However, Cisco strongly recommends
that you obtain proper confirmation of the appropriate interface points from the LEC prior to the design
and implementation phases of 911 call routing.
Interface Type
In addition to providing voice communications, the interfaces used to present 911 calls to the network
must also provide identification data about the calling party.
Automatic Number Identification (ANI) refers to the North American Numbering Plan number of the
calling party, which is used by networks to route a 911 call to the proper destination. This number is also
used by the PSAP to look up the Automatic Location Identification (ALI) associated with a call.
911 calls are source-routed, which means that they are routed according to the calling number. Even
though different locations are all dialing the same number (911), they will reach different PSAPs based
on their location of origin, which is represented by the ANI (calling number).
You can implement 911 call functionality with either of the following interface types:
While dynamic ANI assignment scales better (because it supports multiple ANIs) and lends itself to all
but the smallest of applications, static ANI assignment can be used in a wider variety of environments,
from the smallest to the largest systems.
Integrated Services Digital Network Primary Rate Interface (ISDN-PRI, or simply PRI)
PRI
This type of interface usually connects a communications system to a PSTN Class 5 switch. The calling
party number (CPN) is used at call setup time to identify the E.164 number of the calling party.
Most LECs treat the CPN differently when a call is made to 911. Depending upon the functionality
available in the Class 5 switch and/or upon LEC or government policy, the CPN may not be used as the
ANI for 911 call routing. Instead, the network may be programmed to use the listed directory number
(LDN) or the bill-to number (BTN) for ANI purposes.
If the CPN is not used for ANI, then 911 calls coming from a PRI interface all look the same to the 911
network because they all have the same ANI, and they are all routed to the same destination (which might
not be the appropriate one).
15-5
Chapter 15
Emergency Services
Some LECs offer a feature to provide CPN transparency through a PRI interface for 911 calls. With this
feature, the CPN presented to the Class 5 switch at call setup is used as ANI to route the call. The feature
name for this functionality varies, depending on the LEC. (For example, SBC calls it Inform 911 in
California.)
Note
The CPN must be a routable North American Numbering Plan number, which means that the CPN must
be entered in the routing database of the associated E911 selective router.
Note
For Direct Inward Dial (DID) phones, the DID number could be used as the ANI for 911 purposes, but
only if it is properly associated with an Emergency Service Number in the 911 service provider's
network. For non-DID phones, use another number. (See Emergency Location Identification Number
Mapping, page 15-11, for more information.)
Many Class 5 switches are connected to E911 selective routers through trunks that do not support more
than one area code. In such cases, if PRI is used to carry 911 calls, then the only 911 calls that will be
routed properly are those whose CPN (or ANI) have the same Numbering Plan Area (NPA) as the Class 5
switch.
Example
An MLTS is connected to a Class 5 switch in area code 514 (NPA = 514). If the MLTS were to send a
911 call on the PRI trunk, with a CPN of 450.555.1212, the Class 5 switch would send the call to the
E911 selective router with an ANI of 514.555.1212 (instead of the correct 450.555.1212), yielding
inappropriate routing and ALI lookup.
To use PRI properly as a 911 interface, the system planner must ensure that the CPN will be used for
ANI and must properly identify the range of numbers (in the format NPA NXX TNTN) acceptable on
the link. For example, if a PRI link is defined to accept ANI numbers within the range 514 XXX XXXX,
then only calls that have a Calling Party Number with NPA = 514 will be routed appropriately.
CAMA
Centralized Automatic Message Accounting (CAMA) trunks also allow the MLTS to send calls to the
911 network, with the following differences from the PRI approach:
CAMA trunks are connected directly into the E911 selective router. Extra mileage charges may
apply to cover the distance between the E911 selective router and the MLTS gateway point.
CAMA trunks support 911 calls only. The capital and operational expenses associated with the
installation and operation of CAMA trunks support 911 traffic only.
CAMA trunks for the MLTS market may be limited to a fixed area code, and the area code is
typically implied (that is, not explicitly sent) in the link protocol. The connection assumes that all
calls share the same deterministic area code, therefore only 7 or 8 digits are sent as ANI.
15-6
Chapter 15
Emergency Services
Cisco Emergency Responder
The POTS line can even serve as a backup line in case of power failure.
The POTS line number can be used as the callback number entered into the ALI database.
POTS lines represent the lowest cost 911 support for locations where user density does not justify
local PRI or CAMA access into the PSTN.
All outgoing 911 calls through this type of interface are treated the same by the E911 network, and any
tools that enable ANI manipulation presented to the E911 network (such as translations or
transformations) are irrelevant because the ANI can be only the POTS lines number.
Dynamic association of a phone to an Emergency Response Location (ERL), based on the detected
physical location of the phone.
Dynamic association of the Emergency Location Identification Number (ELIN) to the calling phone,
for callback purposes. In contrast to the general emergency services scenarios outlined in preceding
sections, Cisco Emergency Responder enables the callback to ring the exact phone that initiated the
911 call.
On-site notification to designated parties (by pager, web page, email, or phone call) to inform them
that there is an emergency call in progress. Email, pager, and web page notifications include the
calling party name and number, the ERL, and the date and time details associated with the call.
Phone notification provides the information about the calling number from which the emergency
call was placed.
For more information on ERLs and ELINs, see Emergency Response Location Mapping, page 15-10,
and Emergency Location Identification Number Mapping, page 15-11. For more information on Cisco
Emergency Responder, see Cisco Emergency Responder Design Considerations, page 15-15, and refer
to the Cisco Emergency Responder product documentation available online at
http://www.cisco.com/en/US/products/sw/voicesw/ps842/tsd_products_support_series_home.html
15-7
Chapter 15
Emergency Services
The key functionality of Cisco Emergency Responder relies on the detection of the endpoints location
by discovery of the network port (Layer 2 port, such as a Fast Ethernet switch port) from which the
endpoint made the 911 call. The discovery mechanism relies on two main assumptions:
The wired infrastructure of the enterprise is well established and does not change sporadically.
The infrastructure is available for Cisco Emergency Responder to browse; that is, Cisco Emergency
Responder can establish Simple Network Management Protocol (SNMP) sessions to the underlying
network infrastructure and can scan the network ports for the discovery of connected phones.
Once Cisco Emergency Responder discovers the originating port for the call, it associates the call with
the pre-established ERL for the location of that port. This process also yields an association with a
pre-established ELIN for the location and the selection of the appropriate egress point to the E911
infrastructure, based on the originating ERL.
Cisco Emergency Responder also provides the capability to configure ERLs for IP subnets and to assign
IP endpoint location by IP address. This capability may be used to locate wireless IP phones, IP
softphones, collaboration endpoints that do not support Cisco Discovery Protocol (CDP), and third-party
SIP endpoints registered to Cisco Unified CM, which Cisco Emergency Responder cannot locate by
connected switch port. It may also be used instead of, or in addition to, connected switch port locations
for wired Cisco Collaboration endpoints. If both connected switch port and IP subnet locations are
available for a Cisco Collaboration endpoint, Cisco Emergency Responder will prefer the connected
switch port location because it is usually more specific than the IP subnet location. Using both connected
switch port and IP subnet locations is a best practice because it provides assurance that an appropriate
ERL will be assigned, even in case of any delay or error in detecting the connected switch port.
Cisco Emergency Responder allows for the use of two or more ELINs per ERL. The purpose of this
enhancement is to cover the specific case of more than one 911 call originating from a given ERL within
the same general time period, as illustrated by the following examples.
Example 1
Phone A and phone B are both located within ERL X, and ERL X is associated with ELIN X.
Phone A makes a 911 call at 13:00 hours. ELIN X is used to route the call to PSAP X, and PSAP X
answers and releases the call. Then, at 13:15 hours, phone B makes a 911 call. ELIN X is again used
to route the call to PSAP X.
PSAP X, after releasing the call from phone B, decides to call back phone A for further details
pertaining to phone As original call. The PSAP dials ELIN X, and gets phone B (instead of the
desired phone A).
To work around this situation, Cisco Emergency Responder allows you to define a pool of ELINs for
each ERL. This pool provides for the use, in a round-robin fashion, of a distinct ELIN for each successive
call. With the definition of two ELINs for ERL X in our example, we now have the situation described
in Example 2.
Example 2
Phone A and phone B are both located within ERL X. ERL X is associated with both ELIN X1 and
ELIN X2.
Phone A makes a 911 call at 13:00 hours. ELIN X1 is used to route the call to PSAP X, and PSAP X
answers and releases the call. Then, at 13:15 hours, phone B makes a 911 call, and ELIN X2 is used
to route this call to PSAP X.
PSAP X, after releasing the call from phone B, decides to call back phone A for further details
pertaining to phone As original call. The PSAP dials ELIN X1 and gets phone A.
Of course, if a third 911 call were made but there were only two ELINs for the ERL, the situation would
allow for callback functionality to properly reach only the last two callers in the sequence.
15-8
Chapter 15
Emergency Services
High Availability for Emergency Services
15-9
Chapter 15
Emergency Services
Single building deployments, where all users are located in the same building
Single campus deployments, where the users are located in a group of buildings situated in close
proximity
Multisite deployments, where users are distributed over a wide geographical area and linked to the
call processing site through WAN connectivity
The locations, or type of deployment, affect the criteria used to design and implement 911 services. The
following sections describe the key criteria, along with typical and exceptional situations for each. When
analyzing and applying these criteria, consider how they are affected by the phone locations in your
network.
Note
This document does not attempt to present the actual requirements of any legislation. Rather, the
information and examples presented here are for the purposes of discussion only. The system planner is
responsible for verifying the applicable local requirements.
For example, assume a building has a surface area of 70,000 sq ft and 100 endpoints. In planning for 911
functionality, the building can be divided into 10 zones (ERLs) of 7000 sq ft each, and each endpoint
can be associated with the ERL where it is located. When a 911 call is made, the ERL (which could be
the same for multiple endpoints) is identified by sending the associated ELIN to the PSAP. If the
endpoints were evenly distributed in this example, each group of 10 endpoints would have the same ERL
and, therefore, the same ELIN.
15-10
Chapter 15
Emergency Services
Design Considerations for 911 Emergency Services
The various legislations define a minimum number of endpoints (for example, 49) and a minimum
surface area (for example, 40,000 sq ft) below which the requirements for MLTS 911 are not applicable.
But even if the legislation does not require 911 functionality for a given enterprise, it is always best
practice to provision for it.
They must be routable across the E911 infrastructure. (See the examples in the section on Interface
Type, page 15-5.) If an ELIN is not routable, 911 calls from the associated ERL will, at best, be
handled according to the default routing programmed in the E911 selective router.
Once the ERL-to-ELIN mapping of an enterprise is defined, the corresponding ALI records must be
established with the LEC so that the ANI and ALI database records serving the PSAP can be updated
accurately.
The ELIN mapping process can be one of the following, depending on the type of interface to the E911
infrastructure for a given ERL:
PSAP Callback
The PSAP might have to reach the caller after completion of the initial conversation. The PSAP's ability
to call back relies on the information that it receives with the original incoming call.
The delivery of this information to the PSAP is a two-part process:
1.
The Automatic Number Identification (ANI) is first sent to the PSAP. The ANI is the E.164 number
used to route the call. In our context, the ANI received at the PSAP is the ELIN that the MLTS sent.
2.
The PSAP then uses the ANI to query a database and retrieve the Automatic Location Identification
(ALI). The ALI provides the PSAP attendant with information such as:
Caller's name
15-11
Chapter 15
Emergency Services
Address
Applicable public safety agency
Other optional information, which could include callback information. For example, the phone
number of the enterprise's security service could be listed, to aid in the coordination of rescue
efforts.
Typical Situation
The ANI information is used for PSAP callback, which assumes that the ELINs are dialable
numbers.
The ELINs are PSTN numbers associated with the MLTS. If someone calls the ELIN from the
PSTN, the call will terminate on an interface controlled by the MLTS.
It is the responsibility of the MLTS system administrator to program the call routing so that calls
made to any ELIN in the system will ring a phone (or multiple phones) in the immediate vicinity of
the associated ERL.
Once the ERL-to-ELIN mapping is established, it needs be modified only when there are changes
to the physical situation of the enterprise. If phones are simply added, moved, or deleted from the
system, the ERL-to-ELIN mapping and its associated ANI/ALI database records need not be
changed.
Exceptional Situation
Callback to the immediate vicinity of the originating ERL may be combined with (or even
superseded by) routing the callback to an on-site emergency desk, which will assist the PSAP in
reaching the original caller and/or provide additional assistance with the emergency situation at
hand.
The situation of the enterprise could change, for example, due to area code splits, city or county
service changes requiring a new distribution of the public safety responsibilities, new buildings
being added, or any other change that would affect the desired routing of a call for 911 purposes.
Any of these events could require changes in the ERL-to-ELIN mapping and the ANI/ALI database
records for the enterprise.
15-12
Chapter 15
Emergency Services
Design Considerations for 911 Emergency Services
Emergency Responder's on-site notification capabilities can help in identifying the phone at the origin
of such spurious 911 calls by providing detailed accounts of all calls made to 911, including calls made
by mistake.
In a multisite deployment, the dial plan configuration should ensure that the emergency calls are always
routed through the PSTN gateway local to the site, thereby making sure that the emergency call is routed
to the nearest PSAP within the jurisdiction. One of the mechanism to achieve this could be to use the
Local Route Group feature of Cisco Unified CM. In the case of multisite deployments with centralized
PSTN access, local call routing to the PSAP is not possible. For deployments with centralized PSTN
access, the PSTN provider for centralized connections will route emergency calls to the proper PSAP
based on ANI or ELIN.
Also, in a multisite deployment it is very important to make sure that the emergency number is always
reachable and routed through the local PSTN gateway for the mobility users (extension mobility and
device mobility) independent of the implemented Class of Service (CoS). If the site/device approach is
being used, the device calling search space (CSS) could be used to route the emergency calls.
Cisco recommends enabling Calling Party Modification on Cisco Emergency Responder. When this
feature is enabled, the calling party number is replaced with the ELIN by Cisco Emergency Responder
for the emergency call. If Calling Party Modification is not enabled, either the DID will be sent to the
PSAP or Cisco Unified CM must be configured to replace the calling party with the ELIN.
Gateway Considerations
Consider the following factors when selecting the gateways to handle emergency calls for your system:
Gateway Placement
Within the local exchange carrier (LEC) networks, 911 calls are routed over a locally significant
infrastructure based on the origin of the call. The serving Class 5 switches are connected either directly
to the relevant PSAP for their location or to an E911 selective router, which itself is connected to a group
of PSAPs significant for its region.
With Ciscos IP-based enterprise communications architecture, it is possible to route calls on-net to
gateways that are remotely situated. As an example, an endpoint located in San Francisco could have its
calls carried over an IP network to a gateway situated in San Jose, and then sent to the LEC's network.
For 911 calls, it is critical to choose the egress point to the LEC network so that emergency calls are
routed to the appropriate local PSAP. In the example above, a 911 call from the San Francisco endpoint,
if routed to a San Jose gateway, could not reach the San Francisco PSAP because the San Jose LEC
switch receiving the call does not have a link to the E911 selective router serving the San Francisco
PSAP. Furthermore, the San Jose area 911 infrastructure would not be able to route the call based on a
San Francisco calling party number.
As a general rule, route 911 calls to a gateway physically located with the originating endpoint. Contact
the LEC to explore the possibility of using a common gateway to aggregate the 911 calls from multiple
locations. Be aware that, even if the 911 network in a given region lends itself to using a centralized
gateway for 911 calls, it might be preferable to rely on gateways located with the calling phones to
prevent 911 call routing from being impacted during WAN failures.
15-13
Chapter 15
Emergency Services
Gateway Blocking
It is highly desirable to protect 911 calls from "all trunks busy" situations. If a 911 call needs to be
connected, it should be allowed to proceed even if other types of calls are blocked due to lack of trunking
resources. To provide for such situations, you can dedicate an explicit trunk group just for 911 calls.
It is acceptable to route emergency calls exclusively to an emergency trunk group. Another approach is
to send emergency calls to the same trunk group as the regular PSTN calls (if the interface permits it),
with an alternative path to a dedicated emergency trunk group. The latter approach allows for the most
flexibility.
As an example, we can point emergency calls to a PRI trunk group, with an alternate path (reserved
exclusively for emergency calls) to POTS lines for overflow conditions. If we put 2 POTS lines in the
alternate trunk group, we are guaranteeing that a minimum of two simultaneous 911 calls can be routed,
in addition to any calls that were allowed in the main trunk group.
If the preferred gateway becomes unavailable, it may be acceptable to overflow emergency calls to an
alternate number so that an alternate gateway is used. For example, in North America calls dialed as 911
could overflow to an E.164 (non-911) local emergency number. This approach does not take advantage
of the North American 911 network infrastructure (that is, there is no selective routing, ANI, or ALI
services), and it should be used only if it is acceptable to the applicable public safety authorities and only
as a last resort to avoid rejecting the emergency call due to a lack of network resources.
Answer Supervision
Under normal conditions, calls made to an emergency number should return answer supervision upon
connection to the PSAP. The answer supervision may, as with any other call, trigger the full-duplex audio
connection between the on-net caller and the egress interface to the LEC's network.
With some North American LECs, answer supervision might not be returned when a "free" call is placed.
This may be the case for some toll-free numbers (for example, 800 numbers). In exceptional situations,
because emergency calls are considered "free" calls, answer supervision might not be returned upon
connection to the PSAP. You can detect this situation simply by making a 911 test call. Upon connection
to the PSAP, if audio is present, the call timer should record the duration of the ongoing call; if the call
timer is absent, it is very likely that answer supervision was not returned. If answer supervision is not
returned, Cisco highly recommends that you contact the LEC and report this situation because it is most
likely not the desired functionality.
If this situation cannot be rectified by the Local Exchange Carrier, it would be advisable to configure the
egress gateway not to require answer supervision when calls are placed to the LEC's network, and to cut
through the audio in both directions so that progress indicator tones, intercept messages, and
communications with the PSAP are possible even if answer supervision is not returned.
By default, Cisco IOS-based H.323 gateways must receive answer supervision in order to connect audio
in both directions. To forego the need for answer supervision on these gateways, use the following
commands:
15-14
Chapter 15
Emergency Services
Design Considerations for 911 Emergency Services
Be advised that, in situations where answer supervision is not provided, the call detail records (CDRs)
will not accurately reflect the connect time or duration of 911 calls. This inaccuracy can impede the
ability of a call reporting system to document the relevant statistics properly for 911 calls.
In all cases, Cisco highly recommends that you test 911 call functionality from all call paths and verify
that answer supervision is returned upon connection to the PSAP.
15-15
Chapter 15
Emergency Services
Regardless of which way the video endpoints are discovered, it is important to note that video is not
supported as media for emergency calling to the PSAP.
Note
The topics discussed in this chapter apply to Cisco Emergency Responder only when it is used in
conjunction with Cisco Unified Communications Manager (Unified CM). Cisco TelePresence Video
Communication Server (VCS) currently does not support emergency services.
15-16
Chapter 15
Emergency Services
Design Considerations for 911 Emergency Services
Test Calls
For any enterprise telephony system, it is a good idea to test 911 call functionality, not only after the
initial installation, but regularly, as a preventive measure.
The following suggestions can help you carry out the testing:
Contact the PSAP to ask for permission before doing any tests, and provide them with the contact
information of the individuals making the tests.
During each call, indicate that it is not an actual emergency, just a test.
Confirm the ANI and ALI that the call taker has on their screen.
Confirm that answer supervision was received by looking at the call duration timer on the endpoint.
An active call timer is an indication that answer supervision is working properly.
15-17
Chapter 15
Emergency Services
Note
Cisco Emergency Responder does not support Cisco Unified Communications Manager Express
(Unified CME) or Survivable Remote Site Telephony (SRST). In case of SRST deployment, configure
the appropriate dial-peer to route the 911 calls to the PSTN with the published site number. Unified CME
natively supports E911.
15-18
Chapter 15
Emergency Services
Cisco Emergency Responder Deployment Models
Figure 15-1
Unified CM
Cluster A
IP
IP
LEC
Network
City A
Cisco ER
Group
Primary
LEC
Network
City B
M
M
Unified CM
Cluster B
IP
IP
132073
Secondary
The single Cisco Emergency Responder group in Figure 15-1 interfaces with the following components:
Each Unified CM cluster via SNMP, to collect information about their respective configured
endpoints
Enterprise access switches via SNMP where IP telephony endpoints are connected. This connection
is not required if the endpoint locations are being identified based on IP subnets. For details on
configuring IP subnet-based ERLs, refer to the Cisco Emergency Responder Configuration chapter
in the Cisco Emergency Responder Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps842/prod_maintenance_guides_list.html
Each Unified CM cluster via JTAPI, to allow for the call processing required by any endpoint that
dials 911 for example, identification of the calling endpoint's ERL, assignment of the ELIN,
redirection of the call to the proper gateway (based on the calling endpoint's location), and the
handling of the PSAP callback functionality
The version of the JTAPI interface used by Cisco Emergency Responder is determined by the version of
the Unified CM software to which it is connected. At system initialization, Cisco Emergency Responder
interrogates the Unified CM cluster and loads the appropriate JTAPI Telephony Service Provider (TSP).
Because there can be only one version of JTAPI TSP on the Cisco Emergency Responder server, all
Unified CM clusters to which a single Cisco Emergency Responder group is interfaced must run the
same version of Unified CM software.
For some deployments, this software version requirement might present some difficulties. For instance,
during a Unified CM upgrade, different clusters will be running different versions of software, and some
of the clusters will be running a version of JTAPI that is not compatible with the version running on the
15-19
Chapter 15
Emergency Services
Cisco Emergency Responder servers. When this situation occurs, emergency calls from the cluster
running a version of JTAPI different than that of the Cisco Emergency Responder group might receive
the call treatment provided by the call forward settings of the emergency number's CTI Route Point.
When considering if a single Cisco Emergency Responder group is appropriate for multiple Unified CM
clusters, apply the following guidelines:
Make Unified CM upgrades during an acceptable maintenance window when emergency call
volumes are as low as possible (for example, after hours, when system use is at a minimum).
Use a single Cisco Emergency Responder group only if the quantity and size of the clusters allow
for minimizing the amount of time when dissimilar versions of JTAPI are in use during software
upgrades.
For example, a deployment with one large eight-server cluster in parallel with a small two-server cluster
could be considered for use with a single Cisco Emergency Responder group. In this case, it would be
best to upgrade the large cluster first, thus minimizing the number of users (those served by the small
cluster) that might be without Cisco Emergency Responder service during the maintenance window of
the upgrade. Furthermore, the small cluster's users can more appropriately be served by the temporary
static routing of emergency calls in effect while Cisco Emergency Responder is not reachable because
they can be identified by the single ERL/ELIN assigned to all non-ER calls made during that time.
gateway or, in the case of roaming endpoints, the proper Unified CM cluster
The access switches (via SNMP) to which most of the endpoints associated with the Unified CM of
the Cisco Emergency Responder group are most likely to be connected
This approach allows Unified CM clusters to run different versions of software because each is
interfaced to a separate Cisco Emergency Responder group.
To allow endpoints to roam between various parts of the network and still be tracked by Cisco
Emergency Responder, you might have to configure the Cisco Emergency Responder groups into a Cisco
Emergency Responder cluster. For details on Cisco Emergency Responder clusters and groups, refer to
the chapter on Planning for Cisco Emergency Responder in the Cisco Emergency Responder
Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps842/prod_maintenance_guides_list.html
Figure 15-2 presents a sample topology illustrating some of the basic concepts behind Cisco Emergency
Responder clustering.
15-20
Chapter 15
Emergency Services
Cisco Emergency Responder Deployment Models
Figure 15-2
Unified CM
Cluster A
Cisco ER
Group A
Switch A1
Phone A1
IP
Gateway A
I am looking
for A3!
LEC
Network
City A
IP
Switch A2
Phone A2
Phone A3
LDAP
IP
Roaming
phone
Switch B1
IP
IP
LEC
Network
City B
Phone B1
Gateway B
IP
Cisco ER
Group B
Switch B2
Phone B2
Unified CM
Cluster B
132074
Endpoint Movements Within the Tracking Domain of a Cisco Emergency Responder Group
The emergency call processing for endpoints moving between access switches controlled by the same
home Cisco Emergency Responder group is the same as the processing done for a deployment with a
single Unified CM cluster. For example, an endpoint moving between access switches A1 and A2
remains registered with Unified CM cluster A, and its location is determined by Cisco Emergency
Responder group A both before and after the move. The endpoint is still under full control of Cisco
Emergency Responder group A, for both the discovery of the endpoint by Unified CM cluster A and the
determination of the endpoint's location on switch A2 by Cisco Emergency Responder. The endpoint is
therefore not considered to be an unlocated phone.
15-21
Chapter 15
Emergency Services
Endpoint Movements Between the Various Tracking Domains of a Cisco Emergency Responder Cluster
A Cisco Emergency Responder cluster is essentially a collection of Cisco Emergency Responder groups
that share location information. Each group shares the location of any endpoint it finds on an access
switch or in an IP subnet.
Cisco Emergency Responder groups also share information about endpoints that cannot be located
within a Cisco Emergency Responder group's tracking domain (in switches or IP subnets) but which are
known to be registered in the groups associated Unified CM cluster. Such endpoints are deemed
unlocated.
If an endpoint is roaming between access switches monitored by different Cisco Emergency Responder
groups, those groups must be configured in a Cisco Emergency Responder cluster so they can exchange
information about the endpoint's location. For example, endpoint A3 is registered with Unified CM
cluster A, but it is connected to an access switch controlled by Cisco Emergency Responder group B.
Cisco Emergency Responder group A is aware that endpoint A3 is registered with Unified CM cluster
A, but group A cannot locate endpoint A3 in any of the site A switches. Therefore, endpoint A3 is
deemed unlocated by Cisco Emergency Responder group A.
Cisco Emergency Responder group B, on the other hand, has detected the presence of endpoint A3 in
one of the switches that it monitors. Because the endpoint is not registered with Unified CM cluster B,
endpoint A3 is advertised through the Cisco Emergency Responder LDAP database as an unknown
endpoint.
Because the two Cisco Emergency Responder groups are communicating through an LDAP database,
they can determine that Cisco Emergency Responder group B's unknown endpoint A3 is the same as
Cisco Emergency Responder group A's unlocated endpoint A3.
The Unlocated Phone page in Cisco Emergency Responder group A will display the endpoint's MAC
address along with the remote Cisco Emergency Responder group (in this, case Cisco Emergency
Responder group B).
Endpoint A3 sends the emergency call string to Unified CM cluster A for processing.
2.
Unified CM cluster A sends the call to Cisco Emergency Responder group A for redirection.
3.
Cisco Emergency Responder group A determines that endpoint A3 is located in Cisco Emergency
Responder group B's tracking domain, so it redirects the call to a route pattern that points to
Unified CM cluster B.
4.
5.
Unified CM cluster B sends the call to Cisco Emergency Responder group B for redirection.
6.
Cisco Emergency Responder group B identifies the ERL and ELIN associated with endpoint A3's
location and redirects the call to Unified CM cluster B. The calling number is transformed into the
ELIN associated with the ERL of endpoint A3, and the called number is modified to route the call
to the proper gateway.
15-22
Chapter 15
Emergency Services
Emergency Call Routing Using Unified CM Native Emergency Call Routing
7.
Unified CM cluster B routes the call according to the new called number information obtained from
Cisco Emergency Responder group B.
8.
Unified CM cluster B sends the call out the gateway toward the Emergency PSTN network.
Dynamic association of the ELIN to the calling phone for callback purposes
For mobile devices, Device Mobility Groups used to track mobile devices with Native Emergency
Call Routing
Automatic replacement of the calling party number with the appropriate ELIN
Routing emergency calls to the appropriate gateway for emergency call completion
When designing an emergency call routing plan using Cisco Unified CM Native Emergency Call
Routing services, give special consideration to the boundaries of an emergency location inside a
building. An emergency location should be an identifiable location with physical or logical boundaries
to reduce the amount of time for emergency services to locate an individual in an emergency situation.
Examples of physical or logical boundaries can include: a single floor of a building, a lab, an office, or
a directional floor indicator (for example, West side of first floor).
The design for Native Emergency Call Routing requires an ELIN to be defined and assigned to devices
or devices pools, but the Native Emergency Call Routing feature does not allow the administrator to
define the ERL information to be associated with the ELIN. The ERL definition for a given ELIN must
be done outside of Cisco Unified CM and uploaded to the local PSAP per the instructions provided by
the local exchange carrier when establishing E911 services.
15-23
Chapter 15
Emergency Services
ALI Formats
Similar to a Cisco Emergency Responder deployment, Native Emergency Call Routing can support
multiple unique and concurrent calls to emergency services from the same location. Native Emergency
Call Routing allows the creation of a pool of ELINs that are associated with an emergency location. The
number of locations that can be defined is based on the number of ELINs assigned to an individual
Emergency Location (ELIN) Group. Native Emergency Call Routing supports a maximum of 100
ELINs. If the deployment requires only one concurrent call per a location, then the system can support
100 unique Emergency Location Groups. If the deployment requires the ability to track 2 concurrent
callers from the same location, then the administrator must define 2 ELINs for a single Emergency
Location (ELIN) Group. If 2 ELINs are required for a single location, Unified CM will be able to support
50 locations (2 ELINs 50 ERLs = 100 ELINs). Using more ELINs to support concurrent and uniquely
identified callers from a location will reduce the total number of locations that can be defined. The
following formula can be used to determine the maximum number of locations that can be defined based
on the number of concurrent and unique callers from an ERL:
100/(Number of unique and concurrent callers per ERL) = Max ERLs
ELINs are not required to be the same for each Emergency Location (ELIN) Group. If one ERL covers
a high-density user population, the Emergency Location (ELIN) Group may contain 4 ELINs to support
4 concurrent and unique emergency callers. But if the same building has a large lab floor or warehouse
that has a small number of regular employees, then that location might have only one ELIN assigned to
the Emergency Location (ELIN) Group.
If the PSAP needs to call back and get additional information from the caller, the call will return to
Unified CM using the ELIN that originated the call. To route the return call correctly, the dial plan must
be configured so that the inbound called number matches the ELIN defined in Unified CM. If the
inbound trunk delivers only the last 5 digits of the called party, then the administrator must include a
translation pattern to expand the collected digits to match the ELIN. For proper return call operation, the
called number must match exactly the ELIN number as defined in Unified CM. Although ELINs can be
any number in a customer's DID range, Cisco recommends keeping the ELIN numbers contiguous to use
as few call translation patterns as possible.
ALI Formats
In multi-cluster configurations, there might be instances where the physical locations of ERLs and
ELINs defined in a single Cisco Emergency Responder group span the territory of more than one phone
company. This condition can lead to situations where records destined for different phone companies
have to be extracted from a common file that contains records for multiple LECs.
Cisco Emergency Responder exports this information in ALI records that conform to National
Emergency Number Association (NENA) 2.0, 2.1, and 3.0 formats. However, many service providers do
not use NENA standards. In such cases, you can use the ALI Formatting Tool (AFT) to modify the ALI
records generated by Cisco Emergency Responder so that they conform to the formats specified by the
service provider. The service provider can then use the reformatted file to update their ALI database.
The ALI Formatting Tool (AFT) enables you to perform the following functions:
Select a record and update the values of the ALI fields. AFT allows you to edit the ALI fields to
customize them to meet the requirements of various service providers. The service provider can then
read the reformatted ALI files and use them to update their ELIN records.
Perform bulk updates on multiple ALI records. Using the bulk update feature, you can apply
common changes to all the records that you have selected.
Selectively export ALI records based on area code, city code, or a four-digit directory number. By
selecting to export all the ALI records in an area code, for example, you can quickly access all the
ELIN records for each service provider, thereby easily supporting multiple service providers.
15-24
Chapter 15
Emergency Services
ALI Formats
Given the flexibility of the AFT, a single Cisco Emergency Responder group can export ALI records in
multiple ALI database formats. For a Cisco Emergency Responder group serving a Unified CM cluster
with sites in the territories of two LECs, the basic approach is as follows:
1.
Obtain an ALI record file output from Cisco Emergency Responder in standard NENA format. This
file contains the records destined for multiple LECs.
2.
Make a copy of the original file for each required ALI format (one copy per LEC).
3.
Using the AFT of the first LEC (for example, LEC-A), load a copy of the NENA-formatted file and
delete the records of all the ELINs associated with the other LECs. The information to delete can
usually be identified by NPA (or area code).
4.
Save the resulting file in the required ALI format for LEC-A, and name the file accordingly.
5.
For more information about the ALI formatting tools, refer to the online documentation available at
http://www.cisco.com/en/US/products/sw/voicesw/ps842/prod_maintenance_guides_list.html
For LECs not listed at this URL, the output from Emergency Responder can be formatted using standard
text file editing tools, such as spreadsheet programs and standard text editors.
15-25
Chapter 15
Emergency Services
ALI Formats
15-26
CH A P T E R
16
16-1
Chapter 16
The considerations presented in this chapter apply to Cisco Unified CM as well as the following
applications bundled with it: Cisco Extension Mobility, Cisco Unified Communications Manager
Assistant, WebDialer, Bulk Administration Tool, and Real-Time Monitoring Tool.
For Cisco Unity, refer to the Cisco Unity Design Guide and to the following white papers: Cisco Unity
Data and the Directory, Active Directory Capacity Planning, and Cisco Unity Data Architecture and
How Cisco Unity Works, also available at
http://www.cisco.com
Described in
Revision Date
16-2
Chapter 16
IT Group
M
IP Telephony
Applications
User
Prov
ision
ing
ooku
L
User
tion
ntica
e
Auth
Auth
e
Corporate
LDAP Directory
IP Telephony Application
Administrators
ntica
tion
IP Telephony
Endpoints
IP Telephony End-users
153279
IP
One common requirement is to enable user lookups (sometimes called the "white pages" service) from
IP phones or other voice and/or video endpoints, so that users can dial contacts quickly after looking up
their numbers in the directory.
Another requirement is to provision users automatically from the corporate directory into the user
database for applications. This method avoids having to add, remove, or modify core user information
manually each time a change occurs in the corporate directory.
Authentication of end users and administrators of the voice and/or video applications using their
corporate directory credentials is also a common requirement. Enabling directory authentication allows
the IT department to deliver single log-on functionality while reducing the number of passwords each
user needs to maintain across different corporate applications.
As shown in Table 16-2, within the context of a Cisco Unified Communications system, the term
directory access refers to mechanisms and solutions that satisfy the requirement of user lookups for
Cisco Unified Communications endpoints, while the term directory integration refers to mechanisms
and solutions that satisfy the requirements of user provisioning and authentication (for both end users
and administrators).
16-3
Chapter 16
Table 16-2
Requirement
Cisco Solution
Directory access
User provisioning
Directory integration
LDAP Synchronization
Directory integration
LDAP Authentication
Directory integration
LDAP Authentication
The remainder of this chapter describes how to address these requirements in a Cisco Unified
Communications system based on Cisco Unified CM.
Note
Another interpretation of the term directory integration revolves around the ability to add application
servers to a Microsoft Active Directory domain in order to centralize management and security policies.
Cisco Unified CM is an appliance that runs on a customized embedded operating system, and it cannot
be added to a Microsoft Active Directory domain. Server management for Unified CM is provided
through the Cisco Real Time Monitoring Tool (RTMT). Strong security policies tailored to the
application are already implemented within the embedded operating system.
16-4
Chapter 16
Figure 16-2
Directory Access for Cisco Unified IP Phones Using the Cisco Unified IP Phone
Services SDK
LDAP
COM
Corporate
Directory
IIS
HTTP
Cisco
Unified CM
LDAP COM
Object
Microsoft
Windows
Server
HTTPS
User
Lookup
Authentication
IP Phone
153280
Directories
button
In the example shown in Figure 16-2, the web server proxy function is provided by the Cisco LDAP
Search Component Object Model (COM) server, which is included in the Cisco Unified IP Phone
Services Software Development Kit (SDK). You can download the latest Cisco Unified IP Phone
Services SDK from the Cisco Developer Community at
http://developer.cisco.com/web/ipps/home
The IP Phone Services SDK can be installed on a Microsoft Windows web server running IIS 4.0 or later,
but it cannot be installed on a Unified CM server. The SDK includes some sample scripts to provide
simple directory lookup functionality.
To set up a corporate directory lookup service using the IP Phone Services SDK, perform the following
steps:
Step 1
Modify one of the sample scripts to point to your corporate LDAP directory, or write your own script
using the LDAP Search COM Programming Guide provided with the SDK.
Step 2
In Unified CM, configure the URL Directories parameter (under System > Enterprise Parameters) to
point to the URL of the script on the external web server.
Step 3
16-5
Chapter 16
Note
If you want to offer the service only to a subset of users, configure the URL Directories parameter
directly within the Phone Configuration page instead of the Enterprise Parameters page.
In conclusion, the following design considerations apply to directory access with the
Cisco Unified IP Phone Services SDK:
When querying Microsoft Active Directory, you can perform lookups against the Global Catalog by
pointing the script to a Global Catalog server and specifying port 3268 in the script configuration.
This method typically results in faster lookups. Note that a Global Catalog does not contain a
complete set of attributes for users. Refer to Microsoft Active Directory documentation for details.
There is no impact on Unified CM when this functionality is enabled, and only minimal impact on
the LDAP directory server.
The sample scripts provided with the SDK allow only a minimal amount of customization (for
example, you can prefix a digit string to all returned numbers). For a higher degree of manipulation,
you will have to develop custom scripts, and a programming guide is included with the SDK to aid
in writing the scripts.
This functionality does not entail provisioning or authentication of Unified CM users with the
corporate directory.
For a list of supported LDAP directories, refer to the latest version of the Cisco Unified Communications
Manager System Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
16-6
Chapter 16
Embedded
database
Device configuration
IP
DB
User profiles
Replication
DB
Subscriber 1
DB
Subscriber 2
153282
By default, all users are provisioned manually in the publisher database through the Unified CM
Administration web interface. Cisco Unified CM has two types of users:
End users All users associated with a physical person and an interactive login. This category
includes all Unified Communications users as well as Unified CM administrators when using the
User Groups and Roles configuration (equivalent to the Cisco Multilevel Administration feature in
prior Unified CM versions).
Application users All users associated with other Cisco Unified Communications features or
applications, such as Cisco Attendant Console, Cisco Unified Contact Center Express, or Cisco
Unified Communications Manager Assistant. These applications need to authenticate with
Unified CM, but these internal "users" do not have an interactive login and serve purely for internal
communications between applications.
Table 16-3 lists the application users created by default in the Unified CM database, together with the
feature or application that uses them. Additional application users can be created manually when
integrating other Cisco Unified Communications applications (for example, the ac application user for
Cisco Attendant Console, the jtapi application user for Cisco Unified Contact Center Express, and so
forth).
16-7
Chapter 16
Table 16-3
Application User
Used by:
CCMAdministrator
CCMQRTSecureSysUser
CCMQRTSysUser
CCMSysUser
IPMASecureSysUser
IPMASysUser
WDSecureSysUser
Cisco WebDialer
WDSysUser
Based on these considerations, Figure 16-4 illustrates the default behavior in Unified CM for
user-related operations such as lookups, provisioning, and authentication.
Figure 16-4
DirSync
Authentication
Identity Management
System (IMS) library IMS
Web
Service
User
Lookup
WWW
HTTPS
HTTPS
HTTP
Password
Authentication
Password
Authentication
Embedded
database
DB
HTTP
PIN
Authentication
User
Lookup
IPMA
IP Telephony users,
administrators
(End-Users)
Attendant
Console
Extension Mobility
login
Directories
button
348717
Unified
CCX
End users access the Unified CM User Options page via HTTPS and authenticate with a user name and
password. If they have been configured as administrators by means of User Groups and Roles, they can
also access the Unified CM Administration pages with the same credentials.
16-8
Chapter 16
Similarly, other Cisco features and applications authenticate to Unified CM via HTTPS with the user
name and password associated with their respective application users.
The authentication challenge carried by the HTTPS messages are relayed by the web service on
Unified CM to an internal library called Identity Management System (IMS). In its default
configuration, the IMS library authenticates both end users and application users against the embedded
database. In this way, both "physical" users of the Unified Communications system and internal
application accounts are authenticated using the credentials configured in Unified CM.
End users may also authenticate with their user name and a numeric password (or PIN) when logging
into the Extension Mobility service from an IP phone. In this case, the authentication challenge is carried
via HTTP to Unified CM but is still relayed by the web service to the IMS library, which authenticates
the credentials against the embedded database.
In addition, user lookups performed by Unified Communications endpoints via the Directories button
communicate with the web service on Unified CM via HTTP and access data on the embedded database.
The importance of the distinction between End Users and Application Users becomes apparent when
integration with a corporate directory is required. As mentioned in the previous section, this integration
is accomplished by means of the following two separate processes:
LDAP synchronization
This process uses an internal tool called Cisco Directory Synchronization (DirSync) on Unified CM
to synchronize a number of user attributes (either manually or periodically) from a corporate LDAP
directory. When this feature is enabled, users are automatically provisioned from the corporate
directory in addition to local user provisioning through the Unified CM administration GUI. This
feature applies only to End Users, while Application Users are kept separate and are still provisioned
via the Unified CM Administration interface. In summary, End Users are defined in the corporate
directory and synchronized into the Unified CM database, while Application Users are stored only
in the Unified CM database and do not need to be defined in the corporate directory.
LDAP authentication
This process enables the IMS library to authenticate user credentials of LDAP synchronized End
Users against a corporate LDAP directory using the LDAP standard Simple_Bind operation. When
this feature is enabled, End User passwords of LDAP synchronized End Users are authenticated
against the corporate directory, while Application User passwords and passwords of local End Users
are still authenticated locally against the Unified CM database. Cisco Extension Mobility PINs are
also still authenticated locally.
Maintaining and authenticating the Application Users internally to the Unified CM database provides
resilience for all the applications and features that use these accounts to communicate with Unified CM,
independently of the availability of the corporate LDAP directory.
Cisco Extension Mobility PINs are also kept within the Unified CM database because they are an
integral part of a real-time application, which should not have dependencies on the responsiveness of the
corporate directory.
The next two sections describe in more detail LDAP synchronization and LDAP authentication, and they
provide design best-practices for both functions.
Note
As illustrated in the section on Directory Access for Unified Communications Endpoints, page 16-4,
user lookups from endpoints can also be performed against a corporate directory by configuring the
Cisco Unified IP Phone Services SDK on an external web server.
16-9
Chapter 16
LDAP Synchronization
Synchronization of Unified CM with a corporate LDAP directory allows the administrator to provision
users easily by mapping Unified CM data fields to directory attributes. Critical user data maintained in
the LDAP store is copied into the appropriate corresponding fields in the Unified CM database on a
scheduled or on-demand basis. The corporate LDAP directory retains its status as the central repository.
Unified CM has an integrated database for storing user data and a web interface within Unified CM
Administration for creating and managing user accounts and data. When LDAP synchronization is
enabled, the local database is still used, and additional local end-user accounts can be created.
Management of end-user accounts is then accomplished through the interface of the LDAP directory and
the Unified CM administration GUI. (See Figure 16-5.). Accounts for application users can be created
and managed only through the Unified CM Administration web interface.
The user account information is imported from the LDAP directory into the database located on the
Unified CM publisher server. Information that is imported from the LDAP directory may not be changed
by Unified CM. Additional user information specific to Cisco Unified Communications is managed by
Unified CM and stored only within its local database. For example, device-to-user associations, speed
dials, call forward settings, and user PINs are all examples of data that is managed by Unified CM and
does not exist in the corporate LDAP directory. The user data is then propagated from the Unified CM
publisher server to the subscriber servers through the built-in database synchronization mechanism.
User information synchronized from the LDAP directory can be converted to local user information so
that the user information then can be edited locally on Unified CM. Local end users can be added
manually using the Unified CM administration GUI. During an LDAP sync, a local end user is converted
to an active LDAP user, and if a user with the same user ID is found in LDAP, the locally configured data
is replaced with data from the directory.
16-10
Chapter 16
Figure 16-5
DirSync
DB
LDAP(S)
Authentication
Identity Management
System (IMS) library IMS
Web
Service
Embedded
database
User
Lookup
WWW
HTTPS
Authentication
HTTP
User
Lookup
IP Phone
153284
Directories
button
When LDAP synchronization is activated, only one type of LDAP directory may be chosen globally for
the cluster at any one time. Also, one attribute of the LDAP directory user is chosen to map into the
Unified CM User ID field. Unified CM uses standard LDAPv3 for accessing the data.
Cisco Unified CM imports data from standard attributes. Extending the directory schema is not required.
Table 16-4 lists the attributes that are available for mapping to Unified CM fields. The data of the
directory attribute that is mapped to the Unified CM User ID must be unique within all entries for that
cluster. The attribute mapped to the Cisco UserID field must be populated in the directory and the sn
attribute must be populated with data, otherwise those records are skipped during this import action. If
the primary attribute used during import of end-user accounts matches any application user in the
Unified CM database, that user is not imported from the LDAP directory.
Table 16-4 lists the attributes that are imported from the LDAP directory into corresponding Unified CM
user fields, and it describes the mapping between those fields. Some Unified CM user fields might be
mapped from one of several LDAP attributes.
16-11
Chapter 16
Table 16-4
Unified CM
User Field
User ID
One of:
One of:
One of:
sAMAccountName
mail
employeeNumber
telephoneNumber
userPrincipalName
uid
mail
employeeNumber
telephoneNumber
userPrincipalName
uid
mail
employeeNumber
telephonePhone
First Name
givenName
givenName
givenname
Middle Name
One of:
One of:
initials
middleName
initials
middleName
initials
Last Name
sn
sn
sn
Title
title
title
title
Manager ID
manager
manager
manager
Department
department
department
departmentnumber
Phone Number
One of:
One of:
telephonenumber
telephoneNumber
ipPhone
telephoneNumber
ipPhone
Mobile Phone
Number
mobile
mobile
mobile
Home Phone
Number
homePhone
homePhone
homePhone
Pager Number
pager
pager
pager
Mail ID
Table 16-5 contains a list of additional attributes that are imported by the DirSync process and copied
into the Unified CM only if Microsoft Active Directory is used as the LDAP directory. The attribute
msRTCSIP-PrimaryUserAddress is populated in AD when Microsoft Lync is used. This table is included
for completeness.
Table 16-5
objectGUID
objectGUID
OCSPrimaryUserAddress
msRTCSIP-PrimaryUserAddress
In addition to the direct mapping of directory attributes to local user attributes, other characteristics of
the synchronized users are determined by settings on the LDAP directory synchronization agreement.
Access control group membership of users created through LDAP synchronization is directly configured
in the LDAP directory configuration setting. Further user capabilities are determined by the feature
16-12
Chapter 16
group template selected. The selection of a feature group template on an LDAP directory
synchronization agreement is optional. The feature group templates allow administrators to define user
characteristics, including home cluster selection, IM and Presence capabilities, mobility features,
services profiles, and user profiles. The user profiles allow administrators to define a universal line
template which is considered for automatic creation of directory numbers for LDAP synchronized users
by Unified CM.
The synchronization is performed by a process called Cisco DirSync, which is enabled through the
Serviceability web page. When enabled, it allows one to 20 synchronization agreements to be configured
in the system. This number is reduced to 10 if more than 80,000 users are synchronized. An agreement
specifies a search base that is a position in the LDAP tree where Unified CM will begin its search for
user accounts to import. Unified CM can import only users that exist in the domain specified by the
search base for a particular synchronization agreement.
In Figure 16-6, two synchronization agreements are represented. One synchronization agreement
specifies User Search Base 1 and imports users jsmith, jdoe and jbloggs. The other synchronization
agreement specifies User Search Base 2 and imports users jjones, bfoo, and tbrown. The CCMDirMgr
account is not imported because it does not reside below the point specified by a user search base. When
users are organized in a structure in the LDAP directory, you can use that structure to control which user
groups are imported. In this example, a single synchronization agreement could have been used to
specify the root of the domain, but that search base would also have imported the Service Accts. The
search base does not have to specify the domain root; it may specify any point in the tree.
Figure 16-6
User
Search
Base 2
User
Search
Base 1
dc=vse, dc=lab
ou=Eng
jsmith
ou=Mktg
jdoe
jjones
bfoo
tbrown
153285
ou=Service Accts
jbloggs
To import the data into the Unified CM database, the system performs a bind to the LDAP directory
using the account specified in the configuration as the LDAP Manager Distinguished Name, and reading
of the database is done with this account. The account must be available in the LDAP directory for
Unified CM to log in, and Cisco recommends that you create a specific account with permissions to
allow it to read all user objects within the sub-tree that was specified by the user search base. The sync
agreement specifies the full Distinguished Name of that account so that the account may reside anywhere
within that domain. In the example in Figure 16-6, CCMDirMgr is the account used for the
synchronization.
16-13
Chapter 16
It is possible to control the import of accounts through use of permissions of the LDAP Manager
Distinguished Name account. In this example, if that account is restricted to have read access to ou=Eng
but not to ou=Mktg, then only the accounts located under Eng will be imported.
Synchronization agreements have the ability to specify multiple directory servers to provide redundancy.
You can specify an ordered list of up to three directory servers in the configuration that will be used when
attempting to synchronize. The servers are tried in order until the list is exhausted. If none of the
directory servers responds, then the synchronization fails, but it will be attempted again according to the
configured synchronization schedule.
Synchronization Mechanism
The synchronization agreement specifies a time for synchronizing to begin and a period for
re-synchronizing that can be specified in hours, days, weeks, or months (with a minimum value of 6
hours). A synchronization agreement can also be set up to run only once at a specific time.
When synchronization is enabled for the first time on a Unified CM publisher server, user accounts that
exist in the corporate directory are imported into the Unified CM database. Then either existing
Unified CM end-user accounts are activated and data is updated, or a new end-user account is created
according to the following process:
Note
1.
If end-user accounts already exist in the Unified CM database and a synchronization agreement is
configured, all pre-existing accounts that have been synchronized from LDAP previously are
marked inactive in Unified CM. The configuration of the synchronization agreement specifies a
mapping of an LDAP database attribute to the Unified CM UserID. During the synchronization,
accounts from the LDAP database that match an existing Unified CM account cause that
Unified CM account to be marked active again.
2.
After the synchronization is completed, any LDAP synchronized accounts that were not set to active
are permanently deleted from Unified CM when the garbage collection process runs. Garbage
collection is a process that runs automatically at the fixed time of 3:15 AM, and it is not
configurable.
3.
Subsequently when changes are made in the corporate directory, the synchronization from Microsoft
Active Directory occurs as a full re-synchronization at the next scheduled synchronization period.
On the other hand, the Sun ONE directory products perform an incremental synchronization
triggered by a change in the directory. The following sections present examples of each of these two
scenarios.
Once users are synchronized from LDAP into the Unified CM database, deletion of a synchronization
configuration will cause users that were imported by that configuration to be marked inactive in the
database. Garbage collection will subsequently remove those users.
16-14
Chapter 16
8:00am
11:00pm
Jan. 1st
8:00am:
Account disabled
3:15am
Jan. 2nd
11:00pm:
Periodic Resync
time
Jan. 3rd
3:15am:
Garbage Collection
153292
Figure 16-7
After the initial synchronization, the creation, deletion, or disablement of an account will propagate to
Unified CM according to the timeline shown in Figure 16-7 and as described in the following steps:
1.
At 8:00 AM on January 1, an account is disabled or deleted in AD. From this time and during the
whole period A, password authentication (for example, Unified CM User Options page) will fail for
this user because Unified CM redirects authentication to AD. However, PIN authentication (for
example, Extension Mobility login) will still succeed because the PIN is stored in the Unified CM
database.
2.
The periodic re-synchronization is scheduled for 11:00 PM on January 1. During that process,
Unified CM will verify all accounts. Any accounts that have been disabled or deleted from AD will
at that time be tagged in the Unified CM database as inactive. After 11:00 PM on January 1, when
the account is marked inactive, both the PIN and password authentication by Unified CM will fail.
3.
Garbage collection of accounts occurs daily at the fixed time of 3:15 AM. This process permanently
deletes user information from the Unified CM database for any record that has been marked inactive
for over 24 hours. In this example, the garbage collection that runs at 3:15 AM on January 2 does
not delete the account because it has not been inactive for 24 hours yet, so the account is deleted at
3:15 AM on January 3. At that point, the user data is permanently deleted from Unified CM.
If an account has been created in AD at the beginning of period A, it will be imported to Unified CM at
the periodic re-synchronization that occurs at the beginning of period B and will immediately be active
on Unified CM.
16-15
Chapter 16
8:00am
3:15am
Jan. 1st
8:00am:
Account
deleted
Jan. 3rd
Jan. 2nd
~8:02am:
Incremental sync
time
3:15am:
Garbage Collection
153293
Figure 16-8
An account is deleted from the corporate directory at 8:00 AM on January 1, which causes an
incremental update to be sent from the LDAP server to Unified CM. Unified CM sets its
corresponding copy of the data to inactive. Because LDAP authentication is configured, the user will
be unable to log in via password as soon as the LDAP server has deleted the record. Also, the PIN
may not be used for login at the moment the Unified CM record is marked inactive.
2.
During period B, the users record is still present in Unified CM, albeit inactive.
3.
When the garbage collection runs at 3:15 AM on January 2, the record has not yet been inactive for
24 hours. The data remains in the Unified CM database until the beginning of period C on January 3,
when the garbage collection process runs again at 3:15 AM and determines that the record has been
inactive for 24 hours or more. The record is then permanently deleted from the database.
Accounts that are newly created in the directory are synchronized to Unified CM via incremental updates
as well, and they may be used as soon as the incremental update is received.
If the mask is left empty, then Unified CM takes all digits and also a leading "+" (if present) from
the directory.
16-16
Chapter 16
Examples for Directory Number Creation from LDAP Phone Numbers Based on Masks
Number in LDAP
Mask
Result
14085551234
14085551234
14085551234
+XXXXXXXXXX
+14085551234
14085551234
+XXXXXXXXXXXXXXXXX
+14085551234
14085551234
XXXX
1234
+14085551234
+14085551234
+14085551234
+XXXXXXXXXXXXXXXXX
+14085551234
+496100123
+XXXXXXXXXXXXXXXXX
+496100123
As an alternative to creating directory numbers based on information from LDAP, directory numbers for
new users can also be taken from predefined number pools. Each pool is defined by a start and end
number. Directory number pools support +E.164 numbers. Up to five pools can be defined. Numbers are
assigned from the first pool until all numbers of that pools have been assigned. Number assignment then
starts to take numbers from the next pool.
Automatic Line Creation is enabled only if both of the following conditions are met:
A Universal Line Template is selected in the User Profile selected in the Feature Group Template.
Figure 16-9 shows the hierarchy of configuration elements required to define line-level settings for
automatic line creation.
16-17
Chapter 16
Figure 16-9
Relation of LDAP Directory Configuration, Feature Group Template, User Profile, and
Universal Line Template
LDAP Directory
Feature Group Template
User Profile
Universal Device Template
Universal Line Template
Description
Partition
CSS
AAR Settings
Alternate Numbers
...
348824
Ultimately the Universal Line Template defines the characteristics for all directory numbers that are
automatically created for users added through the corresponding LDAP synchronization definition.
Design Considerations
The calling search space defined in the Universal Line Template determines the class of service of
devices using any of the auto-generated directory numbers. This implies that all directory numbers
created through the same LDAP synchronization agreement share the same class of service, and thus if
directory numbers for multiple sites and multiple classes of service need to be auto-generated, then
multiple LDAP synchronization agreements (one per site and class of service) need to be configured. For
each of these synchronization agreements, disjunct LDAP filters need to be defined, each exactly
matching on only the users belonging to one of the site-specific and class-of-service-specific user
groups. This mapping from LDAP attributes to site and class of service groups can be challenging unless
the group membership based on site and class of service is explicitly encoded in few LDAP attributes
(potentially even in a custom attribute). Also, the maximum number of supported LDAP agreements is
limited, which limits the number of distinct user groups for which directory numbers can be created
automatically.
Automatic creation of directory numbers applies only to users created during LDAP directory
synchronization. Adding, changing, or updating the Universal Line Template for a given LDAP
synchronization agreement will not create directory numbers for already existing users and will not
change the settings of already existing directory numbers.
The Universal Line Template allows administrators to define call forward unregistered destinations and
either to select voicemail as the forward destination or to define an explicit destination. To reach
endpoints in remote sites from registered endpoints in case of WAN failure, the call forward unregistered
destination for the remote site's phones must be set to the PSTN alias (+E.164 number) of the remote
16-18
Chapter 16
phone. This cannot be achieved with Universal Line Template settings because this would require
defining the call forward unregistered destination to be set based on the assigned directory numbers
(potentially with a mask applied).
Security Considerations
During the import of accounts, no passwords or PINs are copied from the LDAP directory to the
Unified CM database. If LDAP authentication is not enabled in Unified CM, the password for the end
user is managed by using Unified CM Administration. The password and PIN are stored in an encrypted
format in the Unified CM database. The PIN is always managed on Unified CM. If you want to use the
LDAP directory password to authenticate an end user, see the section on LDAP Authentication,
page 16-22.
The connection between the Unified CM publisher server and the directory server can be secured by
enabling Secure LDAP (SLDAP) on Unified CM and the LDAP server. Secure LDAP enables LDAP to
be sent over a Secure Socket Layer (SSL) connection and can be enabled by adding the LDAP server
into the Tomcat trust store within the Unified CM Platform Administration. For detailed procedure steps,
refer to the Unified CM product documentation available at http://www.cisco.com. Refer to the
documentation of the LDAP directory vendor to determine how to enable SLDAP.
Use a specific account within the corporate directory to allow the Unified CM synchronization
agreement to connect and authenticate to it. Cisco recommends that you use an account dedicated
to Unified CM, with minimum permissions set to "read" all user objects within the desired search
base and with a password set never to expire. The password for this account in the directory must
be kept in synchronization with the password configuration of the account in Unified CM. If the
service account password changes in the directory, be sure to update the account configuration in
Unified CM.
All synchronization agreements on a given cluster must integrate with the same family of LDAP
servers.
Stagger the scheduling of synchronization agreements so that multiple agreements are not querying
the same LDAP servers simultaneously. Choose synchronization times that occur during quiet
periods (off-peak hours).
If security of user data is required, enable Secure LDAP (SLDAP) by checking the Use SSL field
on the LDAP Directory configuration page in Unified CM Administration.
Ensure that the LDAP directory attribute chosen to map into the Unified CM UserID field is unique
within all synchronization agreements for that cluster.
The attribute chosen as UserID must not be the same as that for any of the Application Users defined
in Unified CM.
The LDAP attribute sn(lastname) is a mandatory attribute for LDAP Synchronization of users.
Administer end-user accounts through the LDAP directory's management tools, and manage the
Cisco-specific data for those accounts through the Unified CM Administration web page.
16-19
Chapter 16
LDAP Synchronization is supported only with Microsoft NT LAN Manager (NTLM). Kerberos and
NTLMv2 are not supported.
For AD deployments, the ObjectGUID is used internally in Unified CM as the key attribute of a user.
The attribute in AD that corresponds to the Unified CM User ID may be changed in AD. For
example, if sAMAccountname is being used, a user may change their sAMAccountname in AD, and
the corresponding user record in Unified CM would be updated.
With all other LDAP platforms, the attribute that is mapped to User ID is the key for that account in
Unified CM. Changing that attribute in LDAP will result in a new user being created in Unified CM,
and the original user will be marked inactive.
User
Search
Base 1
root domain
dc=vse, dc=lab
ou=Corporate
User
Search
Base 2
jjones
bfoo
ou=Svc Accts
User
Search
Base 3
Dir Mgr 1
vse.lab
child
domains
dc=amer, dc=vse, dc=lab
ou=Mktg
ou=Svc Accounts
ou=Eng
ou=Mktg
ou=Svc Accounts
Dir Mgr 2
jdoe
jsmith
jbloggs
amer.vse.lab
Dir Mgr 3
jbrown
tross
bwhite
emea.vse.lab
153286
ou=Eng
16-20
Chapter 16
In Figure 16-10, each of the domains and sub-domains contains at least one domain controller (DC)
associated to them, and the three synchronization agreements each specify the appropriate domain
controller. The DCs have information only on users within the domain where they reside, therefore three
synchronization agreements are required to import all of the users.
When synchronization is enabled with an AD forest containing multiple trees, as shown in Figure 16-11,
multiple synchronization agreements are still needed for the same reasons listed above. Additionally, the
UserPrincipalName (UPN) attribute is guaranteed by Active Directory to be unique across the forest and
must be chosen as the attribute that is mapped to the Unified CM UserID. For additional considerations
on the use of the UPN attribute in a multi-tree AD scenario, see the section on Additional Considerations
for Microsoft Active Directory, page 16-25.
Synchronization with Multiple AD Trees (Discontiguous Namespaces)
User
Search
Base 1
User
Search
Base 2
dc=avvid, dc=info
dc=vse, dc=lab
ou=IT
ou=Mktg
ou=Svc
ou=Eng
ou=Mktg
ou=Svc
ou=Users
ou=Users
Dir Mgr 1
ou=Users
ou=Users
Dir Mgr 2
probert
jbloggs
jsmith
jdoe
avvid.info
jbrown
vse.lab
153287
Figure 16-11
Unified CM sends a default LDAP search filter string to AD when performing the synchronization of
accounts. One of the clauses is to not return accounts that have been marked as disabled in AD. An
account marked disabled by AD, such as when failed login attempts are exceeded, will be marked
inactive if synchronization runs while the account is disabled.
16-21
Chapter 16
LDAP Authentication
The LDAP authentication feature enables Unified CM to authenticate LDAP synchronized users against
the corporate LDAP directory. Application users and locally configured users are always authenticated
against the local database. Also PINs of all end users are always checked against the local database only.
This authentication is accomplished with an LDAPv3 connection established between the Identity
Management System (IMS) module within Unified CM and a corporate directory server, as shown in
Figure 16-12.
Figure 16-12
DirSync
DB
LDAP(S)
Corporate
Directory
(Microsoft AD,
Netscape/iPlanet)
Embedded
database
Authentication
LDAP(S)
Identity Management
System (IMS) library IMS
Web
Service
User
Lookup
WWW
HTTPS
Authentication
HTTP
User
Lookup
IP Phone
153288
Directories
button
16-22
Chapter 16
To enable authentication, a single authentication agreement may be defined for the entire cluster. The
authentication agreement supports configuration of up to three LDAP servers for redundancy and also
supports secure connections LDAP over SSL (SLDAP) if desired. Authentication can be enabled only
when LDAP synchronization is properly configured and used. LDAP authentication configuration is
overridden by enabling SSO. With SSO enabled, end users are always authenticated using SSO, and
LDAP authentication configuration is ignored.
The following statements describe Unified CM's behavior when authentication is enabled:
End user passwords of users imported from LDAP are authenticated against the corporate directory
by a simple bind operation.
End user passwords for local users are authenticated against the Unified CM database.
This behavior is in line with the guiding principle of providing single logon functionality for end users
while making the operation of the real-time Unified Communications system independent of the
availability of the corporate directory, and is shown graphically in Figure 16-13.
Figure 16-13
Authenticating End User Passwords, Application User Passwords, and End User PINs
DirSync
DB
Corporate
Directory
(Microsoft AD,
Netscape/iPlanet)
Embedded
database
IMS
WWW
Password
Authentication
PIN
Authentication
Password
Authentication
IPMA
Attendant
Console
Application users
(ac, jtapi, CCMAdministrator, )
348718
Unified
CCX
16-23
Chapter 16
Figure 16-14 illustrates the following process, adopted by Unified CM to authenticate an end user
synchronized from LDAP against a corporate LDAP directory:
1.
A user connects to the Unified CM User Options page via HTTPS and attempts to authenticate with
a user name and password. In this example, the user name is jsmith.
2.
If the user is a local user, the password is checked against the local database.
Figure 16-14
3.
If the user is an LDAP synchronized user, Unified CM issues an LDAP query for the user name
jsmith, using the value specified in the LDAP Search Base on the LDAP Authentication
configuration page as the scope for this query. If SLDAP is enabled, this query travels over an SSL
connection.
4.
The corporate directory server replies via LDAP with the full Distinguished Name (DN) of user
jsmith (for example, "cn=jsmith, ou=Users, dc=vse, dc=lab").
5.
Unified CM then attempts to validate the user's credentials by using an LDAP bind operation to pass
the full DN and password provided by the user.
6.
If the LDAP bind is successful, Unified CM allows the user to proceed to the configuration page
requested.
Authentication Process
Cisco
Unified CM
End-user
Corporate
Directory
153290
Create a specific account within the corporate directory to allow Unified CM to connect and
authenticate to it. Cisco recommends that you use an account dedicated to Unified CM, with
minimum permissions set to "read" all user objects within the desired search base and with a
password set to never expire. The password for this account in the directory must be kept in
synchronization with the password configuration of the account in Unified CM. If the account
password changes in the directory, be sure to update the account configuration in Unified CM. If
LDAP synchronization is also enabled, you can use the same account for both functions.
This method provides single logon functionality to all end users synchronized from LDAP. They can
then use their corporate directory credentials to log in to the Unified CM User Options page.
16-24
Chapter 16
Manage end-user passwords for LDAP synchronized users from within the corporate directory
interface. Note that the password field is no longer displayed for LDAP synchronized users in the
Unified CM Administration pages when authentication is enabled.
Manage end-user PINs from the Unified CM Administration web pages or from the Unified CM
User Options page.
Manage Application User passwords from the Unified CM Administration web pages. Remember
that these application users facilitate communication and remote call control with other Cisco
Unified Communications applications and are not associated with real people.
Enable single logon for Unified CM administrators by adding their corresponding end user to the
Unified CM Super Users user group from the Unified CM Administration web pages. Multiple
levels of administrator rights can be defined by creating customized user groups and roles.
16-25
Chapter 16
Figure 16-15
Cisco
Unified CM
Search: jdoe
Base: dc=avvid,dc=info
4
Response: full DN
Active Directory
Global Catalog
Server
jdoe@avvid.info
********
dc=avvid, dc=info
jdoe@vse.lab
********
ou=Users
bfoo
John Doe
(avvid.info)
John Doe
(vse.lab)
jdoe
avvid.info
dc=vse, dc=lab
ou=Users
ou=other
jsmith
jdoe
ou=other
jbrown
153291
vse.lab
As shown in Figure 16-15, a user named John Doe exists in both the avvid.info tree and the vse.lab tree.
The following steps illustrate the authentication process for the first user, whose UPN is
jdoe@avvid.info:
Note
1.
The user authenticates to Unified CM via HTTPS with its user name (which corresponds to the
UPN) and password.
2.
Unified CM performs an LDAP query against a Microsoft Active Directory Global Catalog server,
using the user name specified in the UPN (anything before the @ sign) and deriving the LDAP
search base from the UPN suffix (anything after the @ sign). In this case, the user name is jdoe and
the LDAP search base is "dc=avvid, dc=info".
3.
Microsoft Active Directory identifies the correct Distinguished Name corresponding to the user
name in the tree specified by the LDAP query. In this case, "cn=jdoe, ou=Users, dc=avvid, dc=info".
4.
Microsoft Active Directory responds via LDAP to Unified CM with the full Distinguished Name for
this user.
5.
Unified CM attempts an LDAP bind with the Distinguished Name provided and the password
initially entered by the user, and the authentication process then continues as in the standard case
shown in Figure 16-14.
Support for LDAP authentication with Microsoft AD forests containing multiple trees relies exclusively
on the approach described above. Therefore, support is limited to deployments where the UPN suffix of
a user corresponds to the root domain of the tree where the user resides. AD allows the use of aliases,
which allows a different UPN suffix. If the UPN suffix is disjointed from the actual namespace of the
tree, it is not possible to authenticate Unified CM users against the entire Microsoft Active Directory
forest. (It is, however, still possible to use a different attribute as user ID and limit the integration to a
single tree within the forest.)
16-26
Chapter 16
Importing users who will not be assigned Unified Communications resources can increase directory
synchronization time.
Importing users who will not be assigned Unified Communications resources can slow Unified CM
searches and overall database performance.
In many cases, the number of user accounts in the LDAP directory store far exceeds the total user
capacity of the Unified CM database.
Unified CM has no enforced limit on the number of accounts that may be added to the system. Cisco
recommends limiting the number of users to twice the supported number of endpoints. There might be
cases where accounts are needed for applications, and some designs might require additional accounts.
Cisco recommends using the control mechanisms described here to minimize the number of user
accounts imported, regardless of the LDAP database size. This will improve the speed of the first and
subsequent periodic synchronizations and will also improve manageability of the user accounts.
16-27
Chapter 16
LDAP Query
Additional control over filtering might be required for any of the following reasons:
The LDAP directory has a flat structure that does not enable adequate control by configuration of
the synchronization agreements. When the aggregate number of users that are imported by all the
synchronization agreements is greater than the maximum number of users supported by the
Unified CM cluster, then it is necessary to control the number of users imported through filters.
You want to import a subset of user accounts into the Unified CM cluster, for administrative
segmentation of users, to control a subset of users that have access and authentication to the cluster.
Any account that is imported into a cluster has some level of access to the web pages and
authentication mechanisms, which might not be desirable in some cases.
The LDAP directory structure does not have an accurate representation of how users are going to be
mapped into the Unified CM clusters. For instance, if OUs are set up according to an organizational
hierarchy but users are mapped to Unified CM by geography, there might be little overlap between
the two.
In these cases, the LDAP Query filter may be used to provide additional control over the synchronization
agreements.
16-28
Chapter 16
Operator
Meaning of Function
Logical NOT
&
Logical AND
Logical OR
Wildcard
Equal to
>=
<=
An attribute specified in the filter can be any attribute that exists in the LDAP directory store, and it does
not have to be one of the attributes that is understood and imported by Unified CM. The attribute is used
only on the LDAP server to select data, and the corresponding entries will have a subset of their data
imported into Unified CM.
Example 16-1 A Single Condition
(givenName=Jack)
The filter in Example 16-1 selects any user with a given name of Jack.
Example 16-2 Multiple Conditions May Be Joined with Logical Characters
(&(objectclass=user)(department=Engineering))
The filter in Example 16-2 selects all users in the engineering department.
16-29
Chapter 16
Default Active Directory Application Mode (ADAM) or Active Directory Lightweight Directory
Services (AD LDS) filter string
(&(objectclass=user)((objectclass=Computer))(!(msDS-UserAccountDisabled=TRUE)))
High Availability
Unified CM LDAP Synchronization allows for the configuration of up to three redundant LDAP servers
for each directory synchronization agreement. Unified CM LDAP Authentication allows for the
configuration of up to three redundant LDAP servers for a single authentication agreement. You should
configure a minimum of two LDAP servers for redundancy. The LDAP servers can be configured with
IP addresses instead of host names to eliminate dependencies on Domain Name System (DNS)
availability.
16-30
Chapter 16
User account information is cluster-specific. Each Unified CM publisher server maintains a unique list
of those users receiving Unified Communications services from that cluster. Synchronization
agreements are cluster-specific, and each publisher has its own unique copy of user account information.
The maximum number of users that a Unified CM cluster can handle is limited by the maximum size of
the internal configuration database that gets replicated between the cluster members. The maximum
number of users that can be configured or synchronized is 160,000. With more than 80,000 users the
maximum number of LDAP synchronization agreements is limited to 10, while with less than 80,000
users the total number of LDAP synchronization agreements is limited to 20. To optimize directory
synchronization performance, Cisco recommends considering the following points:
Note
Directory lookup from phones and web pages may use the Unified CM database or the IP Phone
Service SDK. When directory lookup functionality uses the Unified CM database, only users who
were configured or synchronized from the LDAP store are shown in the directory. If a subset of users
are synchronized, then only that subset of users are seen on directory lookup.
When the IP Phone Services SDK is used for directory lookup, but authentication of Unified CM
users to LDAP is needed, the synchronization can be limited to the subset of users who would log
in to the Unified CM cluster.
If only one cluster exists, and the LDAP store contains fewer than the maximum number of users
supported by the Unified CM cluster, and directory lookup is implemented to the Unified CM
database, then it is possible to import the entire LDAP directory.
When multiple clusters exist and the number of users in LDAP is less than the maximum number of
users supported by the Unified CM cluster, it is possible to import all users into every cluster to
ensure directory lookup has all entries.
If the number of user accounts in LDAP exceeds the maximum number of users supported by the
Unified CM cluster and the entire user set should be visible to all users, it will be necessary to use
the Unified IP Phone Services SDK to off-load the directory lookup from Unified CM.
If both synchronization and authentication are enabled, user accounts that have either been
configured or synchronized into the Unified CM database will be able to log in to that cluster. The
decision about which users to synchronize will impact the decision on directory lookup support.
Cisco supports the synchronization of user accounts up to the limit mentioned above, but it does not
enforce this limit. Synchronizing more user accounts can lead to starvation of disk space, slower
database performance, and longer upgrade times.
16-31
Chapter 16
For more information, refer to the latest versions of the Cisco TelePresence Management Suite
Administrator Guide and the Cisco TelePresence Management Suite Provisioning Extension Deployment
Guide, both available at
http://www.cisco.com/en/US/products/ps11338/tsd_products_support_series_home.html
16-32
Chapter 16
Figure 16-16
Collaboration Applications
Collaboration Clients
and Admin Interfaces
Corporate
LDAP
Identity
Provider
LDAP
SAML 2.0
OAuth
GUI access and
UC Interfaces
16-33
Chapter 16
In general, a single SAML metadata instance describes either a single SAML entity or multiple entities.
A SAML metadata instance describing multiply SAML instances contains a list of descriptions of single
entities. SAML metadata instances created by Cisco Collaboration solutions always describe only a
single SAML instance.
For any SAML instance described by a SAML metadata instance, the metadata contains:
A unique identifier
Organization
Caching period
Contact persons
Description of SAML role of this SAML instance (identity provider, service provider, and so forth)
SAML Authentication
The actors in generic SAML authentication flow are:
IdP Entity performing the user authentication based on user credentials. The actual credentials
and the actual authentication mechanism are hidden by the IdP. The IdP issues SAML assertions
based on the authentication process result.
SAML defines a number of profiles to describe the use of SAML to solve typical use cases. The relevant
profile used for SSO with Cisco Collaboration services is the web browser SSO profile of SAML V2.0.
16-34
Chapter 16
The use case solved by this profile is the multi-domain web single sign-on, illustrated in Figure 16-17.
In this use case, a user already has a login session with some web service (for example,
travel.example.org) and is using this service. As part of the login process, a security context has been
established for travel.example.org. If the same user now moves to another web service (for example,
sales.example.de) and a business agreement exists between travel.example.org and sales.example.de that
establishes a federated identity for the user between these services, then the user is able to access the
web service sales.example.de without having to provide authentication credentials again. In this case the
identity provider site (travel.example.org) asserts to the service provider site (sales.example.de) that the
user is known, has been properly authenticated, and has certain identity attributes. The service provider
site (sales.example.de) trusts this assertion based on the existing business agreement between the sites
and grants access to the service.
Figure 16-17
1
2
348720
Access protected
service
Business agreement
Identity
Information
Authenticate
This description implies that the user first is authenticated by a web service and that this first web service
then provides an identity assertion to enable the user to access the second web service. The web service
accessed first (travel.example.org) acts as the IdP for SP sales.example.de. This is known as IdP initiated
web SSO.
The more typical web SSO flow used with Cisco Collaboration Services is SP initiated web SSO,
illustrated in Figure 16-18. In this case the user directly (without visiting an IdP first) tries to access a
protected resource on an SP. The SP sends the user to the IdP to get authenticated, and then the user
presents the authentication assertion received from the IdP to the SP to get access.
16-35
Chapter 16
Figure 16-18
Authenticate
Identity Provider
(IdP)
3
Trust relationship
Identity
Information
Service Provider
(SP)
348721
The SAML web browser SSO profile provides a variety of options depending on whether the
authentication is initiated by the IdP or SP and on how the messages are exchanged between IdP and SP.
As mentioned above, Cisco Collaboration services use SP initiated SSO only where the SP sends a user
to an IdP first to authenticate when the user is trying to access a protected resource and does not have an
active session with the service provider. The IdP then builds an authentication assertion and sends the
user back to the SP with that assertion.
The binding used for the messages exchange between IdP and SP for Cisco Collaboration services is the
Redirect/POST binding, illustrated in Figure 16-19. Here an HTTP 302 redirect is used to send the
SAML authentication request message from the SP to the IdP, and the authentication response from IdP
to SP is sent using an HTTP POST message.
16-36
Chapter 16
Figure 16-19
Service Provider
Browser
Access resource
access
check
Supply resource
2
7
POST signed
response
6
Signed response
in HTML form
Assertion
Consumer
Service
Trust
relationship
Identity Provider
4
5
SSO
Challenge for
credentials
Credentials
348722
The user tries to access a service or resource by pointing the browser to the URL hosted on the
application server. The browser at this moment does not have an active session with the service.
2.
The SP realizes that the request originates from a client without an active session. Because HTTP is
stateless, an active session can be detected by the SP only if the client sends a session cookie that
has been issued by the SP earlier. Based on the SSO configuration, the SP now generates an SAML
authentication request to be sent to the appropriate IdP defined as part of the SSO configuration. The
SAML request contains information about the SP generating the request. This is required so that the
IdP can identify the SPs sending SAML requests.
The SP does not communicate directly with the IdP to authenticate the user. Instead the SP redirects
the browser to the IdP. The URL used for this redirect is taken from the IdP metadata exchanged
earlier. The SAML request to be sent to the IdP is included in the redirect as a URL query parameter
using Base64 encoding.
This redirecting HTTP 302 might look like this:
HTTP/1.1 302 Found
Location:
https://pingsso.home.org:9031/idp/SSO.saml2?SAMLRequest=nZLNbtswEITveQqCd1m0pKoWY
RlwYxQ1kDZK5OaQG02tYwISqXLJtH37kkra%2FBjwodflcPab3V2iGPqRr7076lv44QEdIb%2BGXiOfXm
rqreZGoEKuxQDIneTt%2BusVz2aMj9Y4I01PL7abmmJWVCxnku07sYCqFAu2KGWVdaycV1AWRbnPPjJZl
Dkld2BRGV3TYEPJFtHDVqMT2oUSm%2BcJq5Ks2LGK5x84K%2B8p2QQ0pYWbfh2dG5Gn6aj0A6KZHc0AM2
MfeACYp6ob07a9nsUEGSWfjZUwJazpQfQIsWEjENUj%2FKs0z1E%2BKd0F0%2FO5908i5F92uyZprtsdJ
WtEsJHu0mj0A9gW7KOS8P326oVXejkk4F94F0WRpyEBjmmkjdip6JXAEyldXSyjhE%2FDsq%2BWdJ5V%2
FOWiq%2FeWy%2FSV4bP9yL8Fi%2B2mMb2Sv%2F%2FnFuK8B%2BHOq2NFdclhknJnhUYF2lHSNrH%2FjQ9
DOCiwNT2ZA1n3vfl5aUG4sD5nPdDVU5K37CFQenrdqz8%3D&RelayState=s249030c0bda8e96a8086c
92d0619e6446b270c463
16-37
Chapter 16
The encoded SAML authentication request shown above can be decoded to:
<samlp:AuthnRequest xmlns:samlp="urn:oasis:names:tc:SAML:2.0:protocol"
ID="s249030c0bda8e96a8086c92d0619e6446b270c463"
Version="2.0"
IssueInstant="2013-09-19T09:35:06Z"
Destination="https://pingsso.home.org:9031/idp/SSO.saml2"
ForceAuthn="false"
IsPassive="false"
ProtocolBinding="urn:oasis:names:tc:SAML:2.0:bindings:HTTP-POST"
AssertionConsumerServiceURL="https://cucm-eu.home.org:
8443/ssosp/saml/SSO/alias/cucm-eu.home.org"
>
<saml:Issuer xmlns:saml="urn:oasis:names:tc:SAML:2.0:assertion">
cucm-eu.home.org</saml:Issuer>
<samlp:NameIDPolicy xmlns:samlp="urn:oasis:names:tc:SAML:2.0:protocol"
Format="urn:oasis:names:tc:SAML:2.0:nameid-format:transient"
SPNameQualifier="cucm-eu.home.org"
AllowCreate="true"
/>
</samlp:AuthnRequest>
Among other details specifying authentication parameters and identifying the requesting SP, the
above SAML authentication request also specifies the Assertion Consumer Service (ACS) URL. The
ACS URL is the URL to which the SAML Authentication Response needs to be POSTed at the end
of the authentication process.
3.
The browser receives the redirect, follows the URL, and issues the corresponding GET to the IdP.
The SAML request is maintained. The browser at this stage does not have an active session with the
IdP.
4.
After receiving the new request from a browser with no active session (browser is not sending a
cookie issued by the IdP earlier), the IdP authenticates the user based on the pre-configured
authentication mechanisms. Possible authentication mechanisms include user/password, PKI/CAC,
or Kerberos. For user/password authentication, the IdP might push a form to the user to enter the
credentials (for example, 200 OK message with an IdP login form). For the actual authentication,
the IdP might depend on back-end systems such as an LDAP server for user/password
authentication.
One key point here is that the exchange of credentials for the purpose of authentication takes place
between the IdP and the browser. The SP is not involved and does not see the credentials.
5.
The browser provides further information required for the authentication process. For the
user/password case, this would be a POST with the information. For other authentication
mechanisms, other details would need to be sent to the IdP by the browser.
6.
The IdP now checks and validates the provided credentials. The check could involve interactions
with respective back-end systems (LDAP bind for user/password authentication against LDAP,
communication with Kerberos server to validate ticket, and so forth).
Finally the IdP generates an SAML response for the SP. This response contains the SAML assertion
documenting the result of the authentication process. The SAML assertion, in addition to the basic
Yes/No information, also contains validity information and information about attributes describing
the authenticated entity. At least the user ID of the authenticated entity has to be included in the well
known attribute uid so that the SP can extract this information from the assertion to relate the
authenticated entity to users existing in the local database.
The SAML assertion is signed by the IdP according to the SSO key information published in the IdP
metadata. This ensures that the SP can verify the authenticity of the SAML assertion.
16-38
Chapter 16
The IdP returns the SAML assertion to the browser in a hidden form in a 200 OK message. The
hidden form instructs the browser to POST the SAML assertion to the Assertion Consumer Service
(ACS) URL of the SP.
The IdP has to establish a security context so that future authentication requests from the same
browser can be answered without going through the exchange of credentials. The IdP will then
realize that it already has a valid session with the browser and will assert the authentication of the
previously authenticated user without prompting for credentials again. This basically enables SSO
against multiple SPs.
7.
The browser follows the hidden POST received in the 200 OK message and POSTs the SAML
assertion to the Assertion Consumer Service on the SP.
8.
The SP extracts the SAML assertion from the POST and validates the signature of the assertion. This
guarantees the authenticity of the SAML assertion and the IdP. The user identifier received in the
SAML assertion in attribute uid is then used to decide whether the user is authorized to access the
requested service. This is based on local access control configuration on the SP.
The SP grants access to the requested resource and sends back the content in a 200 OK message to
the browser. The SP also sets a session cookie in the browser so that, for subsequent access requests
from the same browser to the same SP, the SP does not have to initiate any more exchanges with the
IdP. The IdP will be involved with additional requests from the same browser only after the SP
session has expired.
OS user
This user is specified during installation and has access to the CLI, the Disaster Recovery System
(DRS) GUI, and the OS Admin GUI. Access to the CLI, DRS, and OS Admin GUI always is
authenticated locally. The password is stored in the local database.
Application user
These are functional users created and managed locally. Passwords are stored in the local database.
Application users are not enabled for SSO. With SSO enabled, application users can get access to
only the Admin GUI through the vanity URL on the landing page.
16-39
Chapter 16
PIN-based authentication is always (even with SSO enabled) based on local configuration. Multiple
collaboration services maintain individual PINs.
The following web services are enabled for SSO based on SAML IdP redirects:
OAuth 2.0
OAuth is an open framework for authorization that provides delegated access to services on behalf of
resource owners. The protocol allows authorization of clients to access resources without sharing the
credentials used by the clients for authentication. OAuth essentially allows access tokens to be issued to
clients by a central authorization instance. These tokens then are presented by the clients to the servers
offering a service. The service providers then ask the authorization instance for validation of the
presented tokens to verify the authorization. In this process the content of the authorization token, as
well as the mechanism used to authenticate the client and issue the token, is completely transparent for
the service provider.
To understand the operation of OAuth, a few definitions need to be introduced.
An OAuth Client is the entity that requests an OAuth access token from the OAuth Authorization service.
Depending on the particular use case, an OAuth Client may be a Collaboration service (for example, the
Collaboration Edge) or an end user client (for example, Jabber). If Collaboration Edge requests an
OAuth token on behalf of a user, then Collaboration Edge acts as an OAuth client. In the case of a Jabber
client login flow inside the enterprise, the Jabber client acts as the OAuth client.
Every OAuth client has an unique identifier, the OAuth client_id. This OAuth client_id uniquely
identifies a client type. For example, Jabber for Windows and Jabber for Android use different client_ids,
but all releases of Jabber for Windows use the same client_id unless concrete reasons mandate a changed
client_id to enable the authorization service to differentiate between different client releases (for
example, support variation in the OAuth exchange with different client releases). A set of client_ids is
predefined for Cisco products and also for a third-party client.
For each given client_id an OAuth redirect_uri exists, to which the authorization service returns issued
tokens. For all predefined client_ids, the redirect_uri is fixed to /ssosp/public/oauthcb.
16-40
Chapter 16
When requesting authorization to a protected resource, an OAuth client might request a token with a
particular scope. The scope indicates the range of services that an OAuth token can be used to access.
An OAuth Access Token is granted by the authorization service and is used by bearers (clients) for access
to a protected resource. Typically access tokens are issued to a specific user and have a specific
expiration time. Whenever an access token expires, the client must get a new access token.
OAuth supports a number of call flows, but only two are relevant for the Collaboration architecture:
Implicit Grant Flow An end user's client (for example, Jabber) acts as the OAuth Client and
requests an OAuth token directly from the authorization service. This is the main grant flow used by
collaboration clients accessing Unified Communications APIs.
SAML Bearer Assertion Grant Flow An entity (typically a service) uses an assertion issued on
behalf of an end user to get an OAuth token that is associated with the end user. This grant flow is
used by Collaboration Edge to get tokens on behalf of clients connecting from outside the edge.
OAuth tokens can be used in HTTP based flows by sending them in an Authorization header. OAuth
tokens are transported in an Authorization header marked with the "Bearer" keyword.
OS
Windows
Mac
iOS
Android
Internet
Explorer
Safari
WebKit
WebKit
Underlying Browser
Yes
Yes
No
No
16-41
Chapter 16
Jabber and other endpoints supporting SSO at startup need to get an OAuth authorization token to get
access to the required collaboration APIs. The mechanism to acquire this OAuth authorization token is
an OAuth Implicit Grant Flow, as shown in figure Figure 16-20.
OAuth Implicit Grant Flow
Browser
1
Service
Authz
IdP
348793
Figure 16-20
GET /authorize?...
/authorize?response_type=token
&client_id=c1
Authentication
302 cucm1.ent-pa.com/ssosp/public/oauthcb#token...
GET /ssosp/public/oauthcb#token...
200 OK [javascript]
The components in this flow are the web browser (embedded browser instance), the collaboration service
(Unified CM with IM and Presence, and Unity Connection), the authorization service, and the IdP.
In step 1 the client, via the embedded browser instance, accesses the Implicit Grant API by sending an
HTTP GET request to the /authorize endpoint on the authorization server. This request at least has to
indicate the desired response type (response_type) and also has to have client_id to identify the type of
client. Clients can also indicate a scope to request a token (for example,
scope=UnifiedCommunications).
Assuming that the client does not have an active session with the authorization service (no valid
authorization service session cookie provided in the request), the authorization service initiates an
SP-initiated SAML 2.0 redirect/POST authentication flow by returning a 302 response redirecting the
browser to the IdP in step 2.
Browser and IdP will then exchange the messages required to authenticate the user. The message
exchange in step 3 depends on the authentication method configured on the IdP.
If the SAML authentication succeeds, then as the last step of the SAML exchange the browser POSTs
the SAML assertion to the assertion consumer service on the authorization service in step 4.
After authentication succeeds, the authorization service can then check whether the authenticated user
is eligible for the requested scope. If that is the case, then the authorization service will grant an
authorization token, which is returned in step 5 as fragment of a location sent back in an HTTP 302
message to the browser. With the 302 sent back, an authorization service session cookie also is returned
to the browser.
In step 6 the browser follows the redirect and sends a GET to the OAuth callback service.
16-42
Chapter 16
This service in step 7 returns a 200 OK with a web page indicating successful authorization. This
returned web page might also contain JavaScript code to redirect the browser to some URL with some
locally registered service type. This mechanism can be used to extract the granted OAuth token and make
it available for the application that initiated the Implicit Grant Flow. This mechanism is not required for
clients using embedded browser instances, in which case the client simply waits for the web transaction
to finish and then extracts the authorization token as a fragment of the OAuth callback URL to which the
client was last redirected.
The OAuth token obtained from the Implicit Grant flow can then be used when accessing resources.
Figure 16-21 shows how an OAuth authentication token is used to provide authentication for a web
service.
Using OAuth Token to Authenticate to a Web Service
Browser
1
Service
Authz
GET /resource
Authorization: Bearer token123...
200 OK
200 OK (content)
348794
Figure 16-21
The client browser presents the OAuth token in an HTTP Authorization header to the web service as part
of the HTTP GET. In step 2 the web service asks the authorization service to validate the access token.
The authorization service returns a 200 OK if the access token is valid. This 200 OK also includes the
user ID for which the token initially was granted, so that the web service can verify that the token is not
abused by a different user. If all checks succeed, then the service returns the requested resource to the
client.
The login procedures of collaboration interfaces not based on HTTP (for example, IM&P SOAP and
IM&P XMPP) are extended to support OAuth tokens. Clients using these interfaces present the OAuth
token during the login process, and after validating the token with the authorization service, then access
to the interface is granted.
OAuth tokens used in the Cisco Collaboration solution have a default lifetime of one hour
(3,600 seconds). If an expired token is used to authorize a request, then the authorization service returns
a 401 error indicating that the token is expired. The client then has to refresh the token by calling the
/authorize endpoint again. If the embedded browser of the client now also passes a valid authorization
service session cookie, then the authorization service will not initiate a new SAML authentication flow,
but instead issues a new OAuth token right away. To avoid this forced OAuth token refresh, Cisco
Collaboration clients initiate a token refresh after 75% of the token expiration time has elapsed. Note
that OAuth token expiration does not affect session-based protocols such as XMPP and SIP.
16-43
Chapter 16
Currently Cisco Unified CM with IM and Presence and Cisco Unity Connection run separate
authorization service instances, which means that OAuth tokens achieved from the authorization service
running on Unified CM cannot be used to access APIs on Unity Connection (for example, VMRest).
Without sharing of OAuth tokens between Cisco Unity Connection and Unified CM with IM and
Presence, clients requiring access to interfaces on both Unified CM with IM and Presence and Unity
Connection must access the separate authorization services on Unified CM with IM and Presence and
on Unity Connection to achieve separate authorization tokens.
Collaboration Applications
DMZ
348795
Figure 16-22
Firewall
Expressway-C
Expressway-E
Collaboration Clients
and Admin Access
Corporate
LDAP
Identity
Provider
IdP proxy
LDAP
SAML 2.0
oAuth
GUI access
UC Interfaces
Endpoints and collaboration clients registering through the collaboration edge can use SSO to access
collaboration applications inside the enterprise. During endpoint registration, the SAML redirect/POST
SAML authentication flow is initiated by Expressway-C (proxied by Expressway-E), and the browser on
the client is redirected to a publicly accessible identity provider. This typically is an IdP proxy in the
customer's DMZ acting as proxy for the IdP deployed inside the enterprise. The IdP proxy in the DMZ
essentially is only a generic HTTPS reverse proxy for the enterprise IdP. Some IdP vendors offer an
option to install an IdP instance in the DMZ as an IdP proxy role. If the customer uses Identity as a
Service (IDaaS), then the IdP used in the SAML exchange is located in the public Internet. In this case
no IdP proxy in the customer's DMZ is required. Expressway-E and Expressway-C proxy only
collaboration client requests for services on the collaboration applications. While the SAML
16-44
Chapter 16
authentication flow is redirected to an IdP proxy in the DMZ by making sure that the public DNS
resolved the DNS name of the enterprise IdP to the public IP address of the IdP proxy in the DMZ, the
OAuth exchange to achieve an OAuth token passes through Expressway-C and Expressway-E, and
Expressway-E requests the OAuth token as a proxy for the actual client. This is a variant of an OAuth
SAML Bearer Grant flow as shown in Figure 16-23.
Figure 16-23
Browser
IdP
Edge
Authz
GET /#(domain)/authorize?...
Authentication
Swap SAML assertion for
Oauth token using the cached
authorization parameters
POST /authorize_proxy
Authentication: Basic #(user)
Assertion=[Assertion]
200 OK
[Oauth token]
302 edge.ent-pa.com/oauthcb?access_token=...
GET /oauthcb?access_token...
200 OK [javascript]
348796
To acquire an OAuth token, the browser sends an HTTP GET request to the /authorize endpoint on Edge.
The /authorize endpoint on Edge is accessed using prefix encoding to refer to the customer domain. Edge
in this description refers to a Expressway-C and Expressway-E pair implementing the collaboration
edge.
Edge in step 2 now initiates an SP-initiated SAML 2.0 redirect/POST authentication flow by returning
a 302 response redirecting the browser to the IdP. Edge also caches the authorization parameters from
the client request because they are needed later in the actual OAuth proxy request.
Browser and IdP then exchange the messages required to authenticate the user. The message exchange
in step 3 depends on the authentication method configured on the IdP.
If the SAML authentication succeeds, then as the last step of the SAML exchange the browser POSTs
the SAML assertion to the assertion consumer service on Edge in step 4. Edge now still needs to
exchange this SAML assertion for an OAuth token.
16-45
Chapter 16
To achieve this, Edge in step 5 uses the /authorize_proxy endpoint on the authorization service. This
request is authenticated using the username and password of an application user. The referenced
application user has to have rights to access the AXL API on the authorization service on Unified CM
and Unity Connection, respectively. The /authorize_proxy request contains all authorization parameters
cached earlier.
The authorization service then can check whether the authenticated end user has the required privileges
for the requested scope. If the authenticated user is authorized to access the requested scope, then the
authorization service issues an OAuth token and in step 6 returns that token in a 200 OK.
Edge can then cache the token and still needs to return the token to the client. This in step 7 is achieved
by returning a 302 to the client, redirecting the browser instance on the client to the OAuth callback on
Edge. The target URL of the redirect contains the required information about the token, including the
token itself and the token lifetime.
The embedded browser on the client in step 8 follows the redirection and accesses the OAuth callback
resource on Edge. The 200 OK returned in step 9 finishes the SSO flow through Edge. The client can
now extract the OAuth token from the final URL.
The acquired OAuth token is used to authorize all further client requests through Edge.
When the OAuth token expires, the client will request a new token from the Authz service.
If the Authz session has expired, the client will request a new Assertion from the IdP.
If the IdP session has expired, the IdP will challenge the client to reenter credentials.
All three timer values impact network load and throughput. Whenever a timer expires, the client must
signal over the network to get a new token or establish a new session.
The IdP session timer impacts user experience. If the user is redirected to the IdP due to OAuth token
and Authz session expiration, and if the IdP session has also expired, the user will be challenged for
credentials.
16-46
Chapter 16
While the IdP timers are configured on the IdP, the UC service session timeout is configured on the
individual UC services. The CLI command set webapp session timeout can be used on Unified CM,
Unified CM IM and Presence, and Unity Connection servers to set the UC service session timeout for
all web services running on a specific server. The UC service session timeout needs to be configured on
each server individually, and it affects all web-based services including the authorization service. The
OAuth token expiration timer is configured in the enterprise parameters on Unified CM and Unity
Connection.
Decreasing the OAuth access token timeout to less than 40 minutes leads to an interesting situation in
that, whenever a client tries to refresh the OAuth token after 75% of the OAuth token expiration, the
default UC service session timeout of 30 minutes for the authorization service has not expired yet, so
that the authorization service immediately issues a new token without initiating a new SAML
authentication flow. While this minimizes the complexity of the reauthorization flow, the lack of
re-initiating a new SAML authentication flow makes it impossible to de-authenticate currently logged
in users.
16-47
Chapter 16
16-48
PART
Mobile Collaboration
CH A P T E R
17
Once the network, call routing, and call control infrastructure has been put in place for your Cisco
Unified Communications and Collaboration System, additional applications and services can be added
or layered on top of that infrastructure. There are numerous applications and services that can be
deployed on an existing Cisco Unified Communications and Collaboration infrastructure, and the
following applications and services are typically deployed:
Presence services Provide user availability tracking across user devices and clients.
Contact center Provides call handling, queuing, and monitoring for large call volumes.
Call recording Provides the ability to record audio and video calls for later retrieval and playback.
The chapters in this part of the SRND cover the applications and services mentioned above. Each chapter
provides an introduction to the application or service, followed by discussions surrounding architecture,
high availability, capacity planning, and design considerations. The chapters focus on design-related
aspects of the applications and services rather than product-specific support and configuration
information, which is covered in the related product documentation.
This part of the SRND includes the following chapters:
17-1
Chapter 17
Architecture
Architecture
As with other network and application technology systems, Unified Communications and Collaboration
applications and services must be layered on top of the underlying network and system infrastructures.
Unified Communications and Collaboration applications and services such as voice messaging, rich
media conferencing, presence, mobility, contact center, and call recording rely on the underlying Unified
Communications and Collaboration call routing and call control infrastructure and network
infrastructure for everything from network connectivity to basic Unified Communications and
Collaboration functions such as call control, supplementary services, dial plan, bandwidth management,
and gateway services. For example, voice messaging and presence applications leverage the network
infrastructure for reaching users in campus sites, in branch sites, and on the Internet. Further, these same
applications depend on the Unified Communications and Collaboration voice and video endpoints, call
routing, PSTN connectivity, and media resources provided by the call routing and control infrastructure.
In addition to relying on these infrastructure layers and basic Unified Communications and
Collaboration services, applications and services are also often dependent upon each other for full
functionality.
17-2
Chapter 17
High Availability
As with network, call routing, and call control infrastructures, critical Unified Communications and
Collaboration applications and services should be made highly available to ensure that required features
and functionality remain available if failures occur in the network or applications. It is important to
understand the various types of failures that can occur and the design considerations around those
failures. In some cases, the failure of a single server or feature can impact multiple services because
many Unified Communications and Collaboration applications are dependent on other applications or
services. For example, while the various application service components of a contact center deployment
might be functioning properly, the loss of all call control servers would effectively render the contact
center unusable because the deployment is dependent upon the call control servers to route calls to the
call center applications.
For applications and services such as voice messaging and mobile collaboration, high availability
considerations include temporary loss of functionality due to network connectivity or application server
failures resulting in the inability of callers to leave messages, of users to retrieve messages, and of users
to schedule or attend conferences. In addition, failover considerations for callers and users of voice
messaging and mobile collaboration applications include scenarios in which portions of the functionality
can be handled by a redundant resource that allows end users to continue to access services in the event
of certain failures.
High availability considerations are also a concern for services such as presence and mobility.
Interrupted network connectivity or server failures will typically result in reduced functionality or, in
some case, complete loss of functionality. For presence services, this can mean that some or all devices
and clients will be unable to send or receive presence or availability updates. For mobility services, high
availability considerations include the potential for loss of specific functionality such as two-stage
dialing or dial-via-office, or reduced functionality for features such as single number reach (resulting in
situations where only the enterprise phone rings or only the mobile phone rings). Further, in some failure
scenarios, enterprise endpoints and mobile clients might have to re-register, re-connect and/or
re-authenticate before full functionality is available again.
For contact center deployments, there are numerous servers and components for which high availability
must be considered. Typically, an isolated single-server or single-component failure can be handled
without loss of features or functionality as long as the server or component has been made redundant. In
other situations, loss of multiple servers or components will typically result in loss of some features or
functionality. However, in scenarios where there is complete loss of a particular component such as all
call control servers, more catastrophic loss of features or functionality is possible.
When considering collaboration clients and applications, high availability is certainly important. Not
only can specific collaboration features or functions become unavailable in failure scenarios, but in some
cases presence-capable clients might be unable to connect to the network for even basic functionality
such as registration and making or receiving calls. In other cases, clients or devices might have to
reconnect and re-authenticate in order to return to service.
17-3
Chapter 17
Capacity Planning
Capacity Planning
Network, call routing, and call control infrastructures must be designed and deployed with an
understanding of the capacity and scalability of the individual components and the overall system.
Similarly, deployments of Unified Communications and Collaboration applications and services must
also be designed with attention to capacity and scalability considerations. When deploying various
Unified Communications and Collaboration applications, not only is it important to consider the
scalability of the applications themselves, but you must also consider the scalability of the underlying
infrastructures. Certainly the network infrastructure must have available bandwidth and be capable of
handling the additional traffic load the applications will create. Likewise, the call routing and control
infrastructure must be capable of handling user and device configuration and registration as well as
application integration loads surrounding protocols and connections. For example, with applications and
services such as mobility, presence, and contact center, there are capacity implications for each of these
individual applications in terms of users, devices, and features, but just as important is the scalability of
the underlying infrastructure to handle connections and protocols such as Computer Telephony
Integration (CTI). While a mobility, presence, or contact center application may be able to support many
CTI connections, the underlying call control and routing infrastructure might not have available capacity
to handle the added CTI load of these application and services.
For applications and services such as voice messaging and rich media conferencing, capacity planning
considerations include things like number of mailboxes or users, mailbox size, audio and video ports,
and MCU sessions. In most cases additional capacity can be added by increasing the number of
application servers and MCUs or by upgrading server or MCU hardware with higher-scale models,
assuming the underlying network and call routing and control infrastructures are capable of handling the
additional load.
Capacity planning considerations are also a concern for services such as presence and mobility.
Scalability must be contemplated not only for things like numbers of configured and supported users and
devices, but also for the number of integrations and connections between those applications and others.
The volume of two-stage dialing and dial-via-office calls is of particular concern for mobility
applications from the perspective of both the call control capacity and the PSTN gateway capacity. With
presence services, on the other hand, critical scalability concerns include frequency of presence status
changes and the propagation of these changes to the network, as well as text or instant message volumes.
Typically, additional application servers or hardware upgrades will result in increased capacity for these
applications and services, but the underlying call routing and control infrastructures must be capable of
handling any increases in load.
Contact center deployments are no different than other applications and services in terms of scalability
concerns. Certainly the number of agents and agent devices handling calls is important in terms of user
and device configuration and registration. However, the major concerns in terms capacity for contact
center deployments are the high number of busy hour call attempts (BHCA) common in contact centers
and the number of CTI integrations to the call control and routing infrastructure.
When considering collaboration clients and application capacity planning, device registration and
configuration are the most important scalability concerns. However, certainly there are other scalability
implications in terms of the back-end applications and services such as presence and messaging. Further,
when deploying or integrating various clients with third-party applications and infrastructures, you must
also consider the supported capacities for those third-party deployments.
For a complete discussion of system sizing, capacity planning, and deployment considerations related to
sizing, refer to the chapter on Collaboration Solution Sizing Guidance, page 25-1.
17-4
CH A P T E R
18
Cisco Unified Communications Manager (Unified CM) applications provide numerous operational and
functional enhancements to basic IP telephony. External eXtensible Markup Language (XML)
productivity applications or IP Phone Services can be run on the web server and/or client on most
Cisco Unified IP Phones. For example, the IP phone on a user's desk can be used to get stock quotes,
weather information, flight information, and other types of web-based information. In addition, custom
IP phone service applications can be written that allow users to track inventory, bill customers for time,
or control conference room environments (lights, video screen, temperature, and so forth). Unified CM
also has a number of integrated applications that provide additional functionality, including:
Cisco WebDialer
WebDialer is a click-to-call application for Unified CM that enables users to place calls easily from
their PCs using any supported phone device.
In some cases these integrated applications also invoke IP Phone Services to provide additional
functionality.
This chapter examines the following Unified CM applications:
18-1
Chapter 18
Described in
Revision Date
IP Phone Services
Cisco Unified IP Phone Services are applications that utilize the web client and/or server and XML
capabilities of the Cisco Unified IP Phone. The Cisco Unified IP Phone firmware contains a
micro-browser that enables limited web browsing capability. These phone service applications provide
the potential for value-added services and productivity enhancement by running directly on the users
desktop phone. For purposes of this chapter, the term phone service refers to an application that transmits
and receives content to and from the Cisco Unified IP Phone.
This section examines the following design aspects of the IP Phone Services feature:
User-initiated (pull)
An IP Phone user presses the Services or Applications button, which sends an HTTP GET message
to Unified CM for displaying a list of user-subscribed phone services. Figure 18-1 illustrates this
functionality.
Phone-initiated (pull)
An idle time value can be set within the IP Phone firmware, as indicated by the URL Idle Time
parameter. When this timeout value is exceeded, the IP Phone firmware itself initiates an HTTP GET
to the idle URL location specified by the URL Idle parameter.
Note
Unlike with the user-initiated and phone-initiated pull functionality, whereby the phone's web client is
used to invoke phone services, the phone service-initiated push functionality invokes action on the phone
by posting content (via an HTTP POST) to the phone's web server (not to its client).
18-2
Chapter 18
Figure 18-1 shows a detailed illustration of the user-initiated IP Phone service operation. With Services
Provisioning set to External URL or Both when a user presses the Services or Applications button, an
HTTP GET message is sent from the IP Phone to the Unified CM getservicesmenu.jsp script by default
(step 1). You can specify a different script by changing the Phone URL enterprise parameter. The
getservicesmenu.jsp script returns the list of phone service URL locations to which the individual user
has subscribed (step 2). The HTTP response returns this list to the IP Phone (step 3). Any further phone
service menu options chosen by the user continue the HTTP messaging between the user and the web
server containing the selected phone service application (step 4).
By default the Services Provisioning parameter is set to Internal. With this setting, the IP phone obtains
the list of phone services from its configuration file instead of sending an HTTP GET message to
Unified CM.
Note
If the Service Provisioning enterprise parameter is set to Internal, steps 1 through 3 are bypassed and the
operation of phone services begins with step 4.
Note
The Cisco Unified IP Phone 7960 does not have the ability to parse the list of phone services from its
configuration file, so it sends an HTTP GET to Unified CM to get that list, even if the Service
Provisioning enterprise parameter is set to Internal.
Figure 18-1
Web server
Cisco
Unified CM
IP Phone
IP
1 HTTP Get
getservicesmenu.jsp
2 Menu generation
3 HTTP Response
List of subscribed
services URLs
Phone Service 1
Phone Service 2
Phone Service 3
Phone Service N
Extension Mobility
IPMA
153263
4 HTTP Get
18-3
Chapter 18
IP Phone Services
Figure 18-2 shows examples of both phone-initiated and phone service-initiated push functionality. In
the phone-initiated example, the phone automatically sends an HTTP GET to the location specified
under the URL Idle parameter when the URL Idle Time is reached. The HTTP GET is forwarded via
Unified CM to the external web server. The web server sends back an HTTP Response, which is relayed
by Unified CM back to the phone, and the phone displays the text and/or image on the screen.
In the phone service-initiated push example, the phone service on the external web server sends an HTTP
POST with a Common Gateway Interface (CGI) or Execute call to the phone's web server. Before
performing the CGI or Execute call, the phone authenticates the request using the proxy authentication
service specified by the URL Authentication parameter. This proxy authentication service provides an
interface between the phone and the Unified CM directory in order to validate requests made directly to
the phone. If the request is authenticated, Unified CM forwards an HTTP Response to the phone. The
phone's web server then performs the requested action, and the phone returns an HTTP response back to
the external web server. If authentication fails, Unified CM forwards a negative HTTP Response, and the
phone does not perform the requested CGI or Execute action but in turn forwards a negative HTTP
Response to the external web server.
Figure 18-2
Web server
Cisco
Unified CM
IP Phone
IP
153264
HTTP Response
In addition to XML Services, a new service can be created with a Service Category of Java MIDlet.
When a Java MIDlet-type service is invoked, the configured Service URL contains the URL from which
the MIDlet JAD file can be retrieved. When the application server receives the JAD file request, the
server should return the appropriate JAR file for that device, which the phone's MIDlet-installer will
download and process.
For more information on Java MIDlet support on Cisco IP Phones, refer to the Cisco IP Phone data sheets
at http://www.cisco.com.
18-4
Chapter 18
Note
After a phone has downloaded its configuration file via TFTP, the phone parses the services
configuration to determine whether or not the list of services has changed, and if so, it updates its local
(persisted) services configuration. If any of the changed services were Java MIDlets (which are explicitly
provisioned and stored on the phone), then the phone sequentially walks through the necessary install,
upgrade, downgrade, and uninstall operations to comply with what was provisioned in the configuration
file. If a MIDlet install fails, it will re-attempt the install the next time the phone checks its configuration
file (during boot, reset, or restart).
The administrator has the added ability to specify the Service Type of configured services to be one of
the following: IP Phone Services, Directories, or Messages. This gives the administrator the flexibility
to control which button users must press on the IP phone to access new services. New services can
optionally be configured as Enterprise Subscriptions, which forces them to appear automatically on all
IP phones without the need to update subscriptions for each individual phone. In addition, services can
be enabled or disabled without the need to delete the service from the Unified CM database.
Note
Default services such as Missed Calls, Placed Calls, and Corporate Directory can also be disabled. This
allows the administrator to create a custom service with a Service URL matching that of the
corresponding default service, thus allowing phones to subscribe to these default services on an
as-needed basis.
Unified CM provides the ability to configure a secure IP Phone Services URL using HTTPS in addition
to a non-secure URL. Phones that support HTTPS will automatically use the secure URL. For more
information about Trust Verification Services and security certificate handling for IP phones, along with
a complete list of phones that support HTTPS, refer to the HTTPS information in the latest version of
the Cisco Unified Communications Manager Security Guide, available at
http://cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
18-5
Chapter 18
IP Phone Services
Redundancy in this case depends upon some type of server load balancing (SLB), as illustrated in
Figure 18-3, where a virtual IP address (or DNS-resolvable hostname) is used to point to one or more
Unified CM servers. This virtual IP address (or DNS-resolvable hostname) is used when configuring the
URL Services parameter. The SLB device is configured with the real IP addresses of the Unified CM
subscriber nodes. Thus, a Unified CM server failure does not prevent the IP Phone Services subscription
list from being returned to the phone when the phone's Services or Applications button is pushed. In
addition, phone services such as Extension Mobility and Unified CM Assistant that run on a Unified CM
server are also potentially made redundant by this method. (See High Availability for Extension
Mobility, page 18-15, and High Availability for Unified CM Assistant, page 18-23.)
Most SLB devices can be configured to monitor the status of multiple servers and automatically redirect
requests during failure events.
Figure 18-3
Unified CM-1
Real IP:
10.1.1.11
SLB
Virtual IP:
10.1.1.1
IP
HTTP
153265
Unified CM-2
Real IP:
10.1.1.12
Failure Scenario 2: External Web Server Hosting a Particular IP Phone Service Fails
In this scenario, the connection to the Unified CM server is preserved, but the link fails to the web server
hosting the user-subscribed phone service. This is an easier scenario to provision for redundancy because
the IP phone is still able to access the Unified CM server when the Services or Applications button is
pressed. In this case, the IP phone is similar to any other HTTP client accessing a web server. As a result,
you can again use some type of SLB functionality (similar to the one indicated in Figure 18-3) to redirect
the HTTP request from the phone to one or more redundant web servers hosting the user-subscribed
phone service.
18-6
Chapter 18
Services Provisioning has to be set to External URL or Both, or if you are using a large number phones
that do not have the ability to retrieve the list of services from their configuration file (such as the Cisco
Unified IP Phone 7960), carefully select the node that will provide the Cisco Unified IP Phone Services
list. For example, consider using the Unified CM TFTP servers instead of the Unified CM publisher if
the load on the publisher is already high, or consider using Unified CM subscribers that are not handling
a lot of traffic.
Note
In the case of Extension Mobility and Unified CM Assistant phone service, Unified CM acts as more
than a redirect server, and additional performance impacts should be considered. See the sections on
Extension Mobility, page 18-7, and Unified CM Assistant, page 18-19, for specific performance and
scalability considerations for these applications.
Because the IP Phone is either an HTTP client or server, estimating the required bandwidth used by an
IP Phone service is similar to estimating the bandwidth of an HTTP browser accessing the same text as
HTTP content residing on a web hosting server.
Extension Mobility
The Cisco Extension Mobility (EM) feature enables users to configure a Cisco Unified IP Phone as their
own, on a temporary basis, by logging in to that phone. After a user logs in, the phone adopts the user's
individual device profile information, including line numbers, speed dials, services links, and other
user-specific properties of a phone. For example, when user X occupies a desk and logs in to the phone,
that user's directory number(s), speed dials, and other properties appear on that phone; but when user Y
uses the same desk at a different time, user Y's information appears. The EM feature dynamically
configures a phone according to the authenticated user's device profile. The benefit of this application is
that it allows users to be reached at their own extension on any phone within the Unified CM cluster,
regardless of physical location, provided the phone supports EM.
This section examines the following design aspects of the Extension Mobility feature:
18-7
Chapter 18
Extension Mobility
Note
1.
When the user presses the Services or Applications button on the phone, this action generates a call
to the URL specified under the URL Services parameter on the Enterprise Parameter configuration
page (see step 1 in Figure 18-4).
2.
An HTTP/XML call is generated to the IP Phone Services, which returns a list of all services to
which the user's phone is subscribed (see step 2 in Figure 18-4).
With the Services Provisioning enterprise parameter set to Internal, steps 1 and 2 are bypassed.
Alternatively, with Services Provisioning set to External URL or Both, a Service URL button can be
configured for EM on a user's phone so that the user can press a line or speed-dial button to generate a
direct call to the Cisco Extension Mobility Application service, also bypassing steps 1 and 2.
3.
Next the user selects the Extension Mobility phone service listing. This selection in turn generates
an HTTP call to the Cisco Extension Mobility Application service, which serves as the interface
between the phone and the Cisco Extension Mobility service (see step 3 in Figure 18-4).
4.
The Cisco Extension Mobility Application service then forwards an XML response back to the
phone requesting user login credentials (userID and PIN) or, if the user is already logged in, a
response asking if the user wants to log off the phone (see step 4 in Figure 18-4).
5.
Assuming the user is attempting to log in, the user must use the phone's keypad to enter a valid
userID and PIN. After the user presses the Submit softkey, a response containing the userID and PIN
just entered is forwarded back to the Cisco Extension Mobility Application service (see step 5 in
Figure 18-4).
6.
The Cisco Extension Mobility Application service next forwards this login information to the Cisco
Extension Mobility service, which interacts with the Unified CM database to verify the users
credentials (see step 6 in Figure 18-4). The Cisco Extension Mobility Application service subscribes
to cluster change notification, and it maintains a list of all nodes in the cluster with the Cisco
Extension Mobility service activated. Therefore, in case the Cisco Extension Mobility service is not
running on the same Unified CM node, the Cisco Extension Mobility Application service forwards
the login information to other Unified CM nodes that are running the Cisco Extension Mobility
service.
7.
Upon successful verification of the users credentials, the Cisco Extension Mobility service also
interacts with the Unified CM database to read and select the appropriate user device profile and to
write needed changes to the phone configuration based on this device profile (see step 7 in
18-8
Chapter 18
Figure 18-4).
8.
Once these changes have been made, the Cisco Extension Mobility service sends back a successful
response to the Cisco Extension Mobility Application service (see step 8 in Figure 18-4).
9.
The Cisco Extension Mobility Application service, in turn, sends a reset message to the phone, and
the phone resets and accepts the new phone configuration (see step 9 in Figure 18-4).
Figure 18-4
Cisco Unified CM
Cisco
IP Phone
Services
service
Cisco
Extension
Mobility
Application
service
1
3
4
9
IP
6
8
Cisco
Extension
Mobility
service
Cisco
Unified CM
Database
254300
18-9
Chapter 18
Extension Mobility
Figure 18-5
John Smiths
Info
Home Cluster
Visiting Cluster
WAN
IP
IP
Visiting
Phone
253890
Home Phone
The EM service in the visiting cluster attempts to locate the home cluster of the user by sending out
queries to each of the EMCC remote clusters that have been configured in Unified CM. When the user's
home cluster responds positively, this initiates communications between the EM services of both clusters
to exchange information that essentially brings the device information into the home cluster database
and allows the home cluster to build a configuration file for this visiting phone. This configuration file
incorporates some device configuration from the visiting cluster, configuration parameters from the
home cluster, and the user's device profile in the home cluster. Once the home cluster TFTP server has
a configuration file for this visiting phone, a reset issued by the visiting cluster forces the visiting phone
to download a small configuration from the visiting cluster, which further instructs it to download
certificates and a full configuration from the home cluster. Ultimately, the visiting phone cross-registers
with the home cluster. This means that all call control signaling occurs between a home cluster
Unified CM subscriber and the visiting phone, and the user's home cluster dialing habits are maintained.
For a step-by-step description of the EMCC login process, refer to the Extension Mobility Cross Cluster
information in the latest version of the Feature Configuration Guide for Cisco Unified Communications
Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
Call Processing
EMCC call processing behavior is also critical to understand because it impacts dial plan design. When
a user has logged into a phone in a visiting cluster, any digits dialed by the user are analyzed by the home
cluster according to the visiting phone's assembled call search space (CSS), which is a concatenation of
the Adjunct CSS in the home cluster's device pool for the visiting phone (referred to as the EMCC
roaming device pool), the Line CSS configured on the directory number associated with the user's device
profile, and the EMCC CSS configured on the user's device profile. Figure 18-6 illustrates the resulting
CSS for an EMCC phone.
18-10
Chapter 18
Adjunct CSS
Adjunct CSS
Partition A1
Partition A2
more
Line CSS
Partition L1
Partition L2
Line CSS
Device
Profile
Config
EMCC CSS
Partition D1
Partition D2
Device CSS
Resulting CSS
Partition A1
Partition A2
Partition L1
Partition L1
Partition D1
Partition D2
Reminder: Partition order
in a CSS is the tie-breaker
only when two patterns are
an equal best match.
253891
Figure 18-6
The Adjunct Calling Search Space is a new call routing configuration parameter that is used by EMCC
to intercept and route emergency numbers for users from a visiting cluster. The Adjunct CSS contains a
partition with directory numbers such as 911, 112, or 999, that route the calls to the visiting cluster and
allow the call to reach emergency services local to the physical phone's location. For more information
on Adjunct Calling Search Spaces and the EMCC roaming device pool and how it is associated with a
visiting phone, refer to the Extension Mobility Cross Cluster information in the latest version of the
Feature Configuration Guide for Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
Note
The EMCC roaming device pool associated with the EMCC feature is not related to the roaming device
pool associated with the Device Mobility feature.
EMCC users must be aware that, when placing calls, they will be leveraging their home Unified CM
routes and numbering plan. For example, if a user from Cluster A logs into a phone from Cluster B and
wants to place a call to the directory number of a Cluster B phone located right next to it, the user would
have to dial the appropriate pattern as if the user was placing the call from Cluster A to the phone in
Cluster B. This implies that the home cluster may initiate an intercluster trunk call from Cluster A to
Cluster B, but the media will flow locally between the visiting phone and the remote phone.
If the EMCC clusters have been deployed using +E.164 numbering, then the users should already be
accustomed to dialing the full number of the target number and will not need to alter their dialing habits.
With PSTN routed calls, there are two different configurations that affect call processing behavior:
Route patterns that do not use the Local Route Group (LRG) feature
When an EMCC logged-in user dials a PSTN call, if the digit analysis matches a route pattern that
ultimately leads to a voice gateway (either via the route list and route group construct or configured
directly to a voice gateway), the call is offered out the gateway. If the Standard Local Route Group
(Standard LRG) feature is not in use, and the call involves a voice gateway associated with the home
cluster; therefore media will flow between the visiting phone (typically across a WAN) back to the voice
gateway. When the route pattern leads to a route list configured to use Standard LRG, the behavior
18-11
Chapter 18
Extension Mobility
changes. (For more information about LRG, see Local Route Group, page 14-31.) When Unified CM
logic must invoke a Standard LRG for an EMCC logged-in device, it recognizes the endpoint as an
EMCC device and sends the PSTN call across a designated EMCC-specific SIP trunk to the visiting
cluster to which this visiting phone is normally registered.
Note
Only one SIP trunk with an EMCC trunk service type is required per cluster. There is no destination
information configured on this trunk; that information is gathered dynamically when adding and
updating an EMCC remote cluster.
When a call invite is received on the EMCC SIP trunk in the visiting cluster, the visiting cluster again
performs digit analysis on the called number according to the CSS of the trunk (or alternatively,
according to the CSS of the visiting phone's original device configuration), and routes the call
accordingly. There is additional information included in a SIP invite across an EMCC SIP trunk, namely
the device name of the visiting phone. This enables the visiting cluster to determine the configured
device CSS of the visiting phone in the database (if required); and if the digit analysis results in matching
a route pattern that ultimately points to the Standard LRG, the visiting cluster is able to determine the
configured Standard LRG for this visiting phone. The Standard LRG in the visiting cluster will typically
contain voice gateways associated with the visiting cluster, therefore the PSTN call is offered out a voice
gateway local to the visiting phone.
The difference between LRG and non-LRG call processing behavior is critical when considering calls
to emergency numbers. While the use of Local Route Groups (LRGs) is not required cluster-wide for an
EMCC deployment, the EMCC logged-in phones must have access to an LRG in order to route
emergency calls correctly. An LRG is required to correctly route an emergency call to a visiting cluster
so that the call can be placed through an appropriate voice gateway local to the visiting phone. The
Adjunct Calling Search Space in the roaming device pool configuration for an EMCC device enables an
administrator to add emergency route patterns that will use an LRG for EMCC logged-in devices, but it
will not affect emergency dialing for other devices in the home cluster. As discussed earlier, an EMCC
logged-in phone will be associated with a device pool (by means of geolocations) that represents all
phone devices from another cluster. The device pools Adjunct Calling Search Space allows for the
visiting cluster's emergency route pattern to be configured so that only emergency calls for an EMCC
logged-in phone will be sent through an LRG. So even if the home and visiting clusters use the same
emergency route pattern, the EMCC logged-in phone's emergency call will route through the LRG to the
visiting cluster. Once the call is received at the visiting cluster through the EMCC SIP trunk, the visiting
cluster dial plan will be responsible for further processing of the call.
Note
If any cluster supporting EMCC is also using Cisco Emergency Responder for emergency call
processing, refer to the Cisco Emergency Responder Administration Guide for information on how to
configure the dial plan to support the deployment, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps842/prod_maintenance_guides_list.html.
Note
If Standard LRGs are already deployed for the emergency route pattern, and if the home and visiting
clusters use the same emergency dial string, use of the Adjunct CSS is not required.
For detailed EMCC call processing examples and configuration, refer to the Extension Mobility Cross
Cluster information in the latest version of the Feature Configuration Guide for Cisco Unified
Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
18-12
Chapter 18
Media Resources
All media resources except for RSVP agents are allocated from the home cluster according to the media
resource group list of the device pool assigned to the visiting phone. Conferencing, transcoding, and
music on hold all function as normal, with the difference being that media is streaming between the
visiting phone and media resources across (typically) a WAN separating the home and visiting clusters.
When an EMCC logged-in user makes a call that requires use of an RSVP agent, the Unified CM EMCC
logic is able to determine it is a visiting phone, and it sends a resource request across the EMCC SIP
trunk to the remote cluster to which the visiting phone belongs. The device name of the visiting phone
is included in this request, which enables the visiting cluster to verify the RSVP agent media resources
that are normally assigned to this visiting phone and to allocate its use for the call.
18-13
Chapter 18
Extension Mobility
Mixed Mode
Non-Secure Mode
Secure mode
Non-secure mode
Non-secure mode
Non-secure mode
Visiting Cluster
Note
As of Cisco Unified CM 9.0, the EMCC SIP trunk cannot be configured with a secure profile. Therefore,
calls to the local PSTN do not use a secure channel for signaling. However, the media is encrypted if the
phone and PSTN gateway are configured in a secure mode.
18-14
Chapter 18
Note
While the Cisco Extension Mobility service can be activated on more than two nodes, a
maximum of two nodes can actively handle login/logout requests at any given time. The
other nodes running the Cisco Extension Mobility service should start handling login/logout
requests only in case of failure.
Cisco recommends deploying a server load balancer device to load-balance the requests across two
Unified CM nodes and to provide redundancy. Without a server load balancer, load balancing would be
uneven and the redundancy would be manual. For example, two EM IP Phone services could be
configured on each phone. If one Unified CM node is not reachable, the end user would have to manually
select the other EM IP Phone service to reach the other node.
Note
While it is possible to provide redundancy for the EM IP Phone service by relying on end users to
manually select an EM IP Phone service from a list of EM IP Phone services, achieving high availability
in this manner can be problematic. Because there is no control over which EM IP Phone service a user
might select from the phone services menu (or assigned feature keys), there is no way to ensure that the
EM login/logout load is balanced between Unified CM nodes handling EM login/logout requests.
Further, end user behavior when encountering delay in response of the EM service, which is typical in a
failure scenario, will usually exacerbate the situation as users cancel EM service calls and select
alternate EM IP Phone service. This can lead to added congestion and load on the network as well as on
the remaining Unified CM node handling EM login/logout requests.
A deployment with two Unified CM nodes running the Cisco Extension Mobility service provides the
highest capacity in terms of number of login/logout requests per minute. (See Capacity Planning for
Extension Mobility, page 18-17, for details.) It also provides redundancy. However, in case of failure,
the login/logout request capacity is reduced because there is only one node left. Therefore, to achieve
the highest login/logout capacity and maintain this capacity in case of failure, the Cisco Extension
Mobility service should be activated on additional Unified CM nodes. To load balance evenly across the
active nodes and to ensure that only two nodes are handling login/logout requests at any given time, a
server load balancer device should be deployed.
18-15
Chapter 18
Extension Mobility
Note
Cisco does not recommend a redundancy design using DNS A or SRV records with multiple IP listings.
With multiple IP addresses returned to a DNS request, the phones must wait for a timeout period before
trying the next IP address in the list, and in most cases this results in unacceptable delays to the end user.
In addition, this can result in more than two subscriber nodes with the Cisco Extension Mobility
Application service enabled to handle login/logout requests, which is not supported.
With EMCC, remote clusters are administratively added via Unified CM web administration by
specifying a single FQDN or IP address of a Unified CM subscriber node running the EM service in the
remote cluster. The EM services between the two clusters provide information about the Unified CM
version, an ordered list of EM Service nodes for EMCC EM Service communications, which EMCC SIP
trunk services are enabled (PSTN Access and/or RSVP Agent) in the remote cluster, and an ordered list
of up to three remote Unified CM nodes that handle EMCC SIP trunk operations for each EMCC service.
EMCC EM service communications over HTTPS include locating users' home clusters, exchanging
information during EMCC logins, and remote cluster updates. Upon an initial update, a remote cluster's
Extension Mobility Application service is queried, which will return the first three EM Service nodes in
its list. This ordered list determines which remote cluster EM Service nodes will be used for EMCC
communications.
The remote cluster obtains the information regarding primary, secondary, and tertiary options for EMCC
PSTN Access and RSVP Agent services from the Unified CM Group that is associated with the device
pool of the assigned EMCC SIP trunk for those services. This ensures that, if the primary Unified CM
subscriber handling the EMCC SIP trunk is offline, then the EMCC SIP trunk call will be handled by the
secondary Unified CM subscriber, and so on.
Once a phone is logged in through EMCC, redundancy is provided for the phone in the form of the
Unified CM Group configured in its assigned EMCC device pool. If the visiting phone is located in a
remote site and there is a WAN outage in which both the visiting and home cluster are unreachable, then
the SRST reference from the visiting cluster is maintained by the EMCC phone. Therefore, an EMCC
logged-in phone will still be able to register with the appropriate SRST router in the site where it is
located. The EMCC logged-in user's DID most likely will not be associated with the local gateway(s) at
the SRST site, so incoming calls will still be routed based on the call forwarding rules on the user's home
cluster. While in SRST mode, the user will also have to adapt to the visiting SRST site's configured dial
habits during SRST failover registration. For additional examples of an EMCC logged-in phone's
behavior during a networking failure, refer to the Cisco Extension Mobility Cross Cluster section in the
latest version of the Feature Configuration Guide for Cisco Unified Communications Manager, available
at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
Cisco also recommends configuring a default and backup Unified CM TFTP server to be used for
visiting phones to download EMCC configuration files that will allow them to register with the home
cluster. This is configured under EMCC Feature Configuration.
18-16
Chapter 18
Note
The Cisco Extension Mobility service can be activated on more than two nodes for redundancy purposes,
but Cisco supports a maximum of two subscriber nodes actively handling logins/logouts at any given
time.
Note
18-17
Chapter 18
Extension Mobility
EM users should not move between locations or sites within a cluster when Automated Alternate
Routing (AAR) and/or the Voice over PSTN (VoPSTN) deployment model are in use.
EM functionality relies on the use of the IP network for routing calls. Call routing via the PSTN is
more problematic because E.164 PSTN numbers are static and the PSTN is unable to account for
movement of EM user directory numbers (DNs) from their home sites. AAR relies on the PSTN for
call routing, as does the VoPSTN deployment model. In both cases, EM user movement between
locations and sites is supported only if all sites the user is traversing are in the same AAR group.
For additional information, see Extension Mobility, page 14-83.
Restarting the Cisco Extension Mobility service or the node on which the service is running will
affect auto-logout settings.
If the Cisco Extension Mobility service is stopped or restarted, the system does not auto-logout users
who are already logged in after the expiration of the maximum login interval. These phones will
either have to be logged out manually or wait until the daily database clean-up process runs
(typically at midnight).
Some Cisco endpoints such as the Cisco TelePresence endpoints might not support Extension Mobility.
WebDialer supports the use of phones logged in using Extension Mobility. For more information, please
see WebDialer, page 18-34.
Prior to Unified CM release 9.1(1), EMCC requires that all users must be unique across all clusters
in the enterprise. If LDAP synchronization is maintaining common users for multiple clusters, some
type of filtering must be applied.
Starting with Unified CM release 9.1(1), the same user ID can exist in multiple cluster; however,
only one cluster should be defined as a home cluster for the user. When a user attempts to log in on
a cluster that has the home cluster option selected for the user, the cluster will perform a local EM
login and will not attempt an EMCC login with the remote cluster(s).
Consider the network delay between clusters in combination with the features you plan to use. As
the visiting phone is registered with the home cluster, features will work. However, depending on
the network delay for a given deployment, all applications and features might not meet user
requirements. Testing might be required to determine the usability of features for a given network.
For example, EMCC supports dynamic CTI control of a visiting phone. But if an offhook is issued
via an application and it takes 1 second before the phone goes offhook, this might be acceptable for
an office worker but might not be acceptable for a call center agent.
Phone load firmware is not enforced during the login process. Instead, the visiting cluster phone load
information is maintained so that cross-registration does not result in new phone firmware
downloads.
18-18
Chapter 18
If the home cluster locale is different than that of the visiting cluster, the phone will download the
new locale from the visiting cluster TFTP server. If it is not available, then the phone will not change
locales and instead will maintain the visiting cluster locale.
The total number of EMCC logins is controlled by the total number of EMCC inserted devices in
the Bulk Administration Tool (BAT).
Except for RSVP agents, all other media resources are allocated from the home cluster according to
the media resource group list associated with the EMCC roaming device pool.
Audio and video codecs are determined by the EMCC region settings. These settings override
normal region configuration for EMCC registered phones. All EMCC region parameters must be
configured with the same values in all clusters. If they are different, RSVP Agent for that cluster
will be disabled by the remote cluster update operation.
For the EMCC roaming device pool to be assigned correctly, EMCC-capable phones must have a
geo-location configured via device configuration or a device pool.
Incoming calls for a user's directory number will always be received on a home cluster voice
gateway, therefore RTP media will flow between the visiting phone and the home gateway for
incoming calls.
Calls sent across the EMCC SIP trunk will have gone through digit manipulation in the home cluster.
The called number may require manipulation to match visiting cluster route patterns.
Verify configured codec capabilities of H.323 and SIP gateways in the home cluster. For example,
if home cluster gateways are configured to accept only G.711 calls and the EMCC region bandwidth
is set to 8 kbps (G.729), a transcoder is required to complete the call. Alternatively, the H.323 or SIP
gateway dial peers may be configured to allow for G.729 in addition to G.711.
Design considerations must be made regarding the calling party for EMCC emergency calls.
Depending on dial plan configurations, the calling party number leaving the visiting cluster gateway
may be the user's DID that is normally associated with the home cluster. This would require
transforming the calling number incoming on the EMCC SIP trunk, on route patterns, or egressing
on the visiting gateways.
When EMCC is deployed with Cisco Emergency Responder, Emergency Responder should be
deployed in all clusters handled by a single Emergency Responder cluster. If the visiting cluster is
deployed with Emergency Responder and the home cluster is not, Emergency Responder will not be
able to identify the visiting phone when the call arrives back to the visiting cluster.
Unified CM Assistant
Cisco Unified Communications Manager Assistant (Unified CM Assistant) is a Unified CM integrated
application that enables assistants to handle incoming calls on behalf of one or more managers. With the
use of the Unified CM Assistant Console desktop application or the Unified CM Assistant Console
phone service on the assistant phone, assistants can quickly determine a manager's status and determine
what to do with a call. Assistants can manipulate calls using their phone's softkeys and service menus or
via the PC interface with either keyboard shortcuts, drop-down menus, or by dragging and dropping calls
to the managers proxy lines.
18-19
Chapter 18
Unified CM Assistant
This section examines the following design aspects of the Unified CM Assistant feature:
Manager
Assistant handles
call; redirects
to Manager phone
IP
DN: 60001
3
6XXXX
Assistant
IP
Proxy
Line: 39001
Unified CM
Assistant
2
Route
Point
Redirect to
Proxy Line
4 Given RP failure,
CFNA ensures
fall-through to
Manager phone
CFNA: 6XXXX
IP
153267
Dials: 60001 1
Phone A
18-20
Chapter 18
Note
The CFNA fall-through mechanism illustrated in Figure 18-7 requires configuration of the same
summarized digit-string as the Unified CM Assistant RP directory number in both the Forward No
Answer Internal and Forward No Answer External fields under the Unified CM Assistant RP directory
number configuration page. In addition, the calling search space (CSS) field for each of these call
forward parameters should be configured with the calling search space containing the partition with
which the Manager phone DNs are configured, so that the Manager phone DNs can be reached if the
Unified CM Assistant RP or Unified CM Assistant application fails.
Assistant
Assistant and
Manager
phones ring
IP
Manager
IP
DN: 60001
Shared
Line: 60001
IP
Phone A
153268
Dials: 60001 1
In Unified CM Assistant shared line mode, the Unified CM Assistant RP is not needed or required for
intercepting calls to the Manager phone. However, the Do Not Disturb (DND) feature on the Manager
phone and the Unified CM Assistant Console desktop application still depend on the Cisco IP Manager
Assistant (IPMA) and Cisco CTIManager services. Furthermore, in Unified CM Assistant shared line
mode, features such as call filtering, call intercept, assistant selection, and Assistant Watch are not
available.
Manager and Assistant phones register with the Cisco CallManager Service, and the phones keypad
and softkeys are used to handle call flows (see step 1 in Figure 18-9).
18-21
Chapter 18
Unified CM Assistant
2.
Both the Unified CM Assistant Console desktop application and the Manager Configuration
web-based application communicate and interface with the Cisco IP Manager Assistant service (see
step 2 in Figure 18-9).
3.
The Cisco IP Manager Assistant service in turn interacts with the CTIManager service for
exchanging line monitoring and phone control information (see step 3 in Figure 18-9).
4.
The CTIManager service passes Unified CM Assistant phone control information to the
Cisco CallManager service and also controls the Unified CM Assistant RP (see step 4 in
Figure 18-9).
5.
In parallel, the Cisco IP Manager Assistant service reads and writes Unified CM Assistant
application information to and from the Unified CM database (see step 5 in Figure 18-9).
6.
The Manager may choose to invoke the Unified CM Assistant phone service by pushing the Services
or Applications button, thus generating a call to the IP Phone Services service that will return a list
of all services (including the Unified CM Assistant phone service) to which the phone is subscribed
(see step 6 in Figure 18-9).
The Unified CM Assistant phone service is controlled by the Cisco IP Manager Assistant service, and
configuration changes made by the Manager using the phone are handled and propagated via the Cisco
IP Manager Assistant service.
Note
With the Services Provisioning enterprise parameter set to Internal, steps 1 and 2 are bypassed.
Alternatively, with Services Provisioning set to External URL or Both, a Service URL button can be
configured for the Unified CM Assistant phone service on a user's phone so that the user can press a line
or speed-dial button to generate a direct call to the Cisco IP Manager Assistant service, also bypassing
steps 1 and 2.
18-22
Chapter 18
Figure 18-9
Cisco Unified CM
Cisco
IP Phone
Services
service
6
Manager and
Assistant
Phones
Cisco
Unified CM
service
IP
IP
4
Cisco
CTIManager
service
Unified CM
Assistant Console
3
Cisco IP
Manager
Assistant
service
2
Manager
Configuration
Cisco
Unified CM
Database
Note
254301
While Figure 18-9 shows the IP Phone Services, Cisco CallManager, CTIManager, and Cisco
IP Manager Assistant services all running on the same node, this configuration is not a requirement.
These services can be distributed between multiple nodes in the cluster but have been shown on the same
node here for ease of explanation.
18-23
Chapter 18
Unified CM Assistant
Note
Note
The redundancy scenario depicted in Figure 18-10 shows a special circumstance. During normal
operation it is not possible to have any pair of Unified CM Assistant servers active at the same time. If
an active and backup pair of Unified CM Assistant servers can communicate over the network, then one
server will be in backup mode and cannot handle requests.
18-24
Chapter 18
Figure 18-10
IP WAN
Manager A
IP
Assistant A
Manager B
IP
Assistant B
IP
153272
IP
Site 1
Site 2
As previously mentioned, the publisher is a single point of failure when it comes to writing Unified CM
Assistant information to the Unified CM database. Given a publisher failure, all aspects of the
Unified CM Assistant application will continue to work; however, no changes to the Unified CM
Assistant application configuration can be made. Configuration changes via the Unified CM Assistant
Console desktop application, the Manager configuration web-based application, the phone softkeys, or
the Unified CM Assistant phone service, will not be possible until the publisher is restored. This
condition includes enabling or disabling features such as Do Not Disturb, DivertAll, Assistant Watch,
and call filtering, as well as changing call filter and assistant selection configuration.
18-25
Chapter 18
Unified CM Assistant
more quickly, be careful not to affect the network adversely by sending keep-alives too frequently. This
consideration is especially important if there are a large number of Assistant Console desktop
applications in use.
The Unified CM Assistant Console phone service, unlike the Unified CM Assistant Console desktop
application, requires manual intervention for redundancy given the failure of the primary Unified CM
Assistant server. If the primary Unified CM Assistant server goes down, assistants using the phone
console will not see an indication of this condition. However, the assistant phone will receive a "Host
not found Exception" message upon trying to use a softkey. In order to continue using the phone console
with the backup Unified CM Assistant server, the user must manually select the secondary Unified CM
Assistant phone service from the IP Services menu and log in again.
There are several other failover mechanisms which ensure that Manager and Assistant reachability are
redundant. First, calls sent to a Manager's Assistant via the Unified CM Assistant application (in proxy
line mode) can be forwarded to the Manager's next available Assistant if the call is not answered after a
configured amount of time. If the next Assistant does not answer the call after the configured amount of
time, the call can again be forwarded to the Manager's next available Assistant, and so on. The
mechanism is configured using the Cisco IPMA RNA Forward Calls and Cisco IPMA RNA Timeout
service parameters. Second, as mentioned previously, if all Cisco IP Manager Assistant and CTI services
on cluster nodes fail, the Unified CM Assistant RP will become unavailable. However, based on the
CFNA configuration of the Unified CM Assistant RP, calls to all Manager DNs will fall-through directly
to the Manager phones so that Manager reachability is sufficiently redundant.
A maximum of 33 Managers can be configured for a single Assistant (if each Manager has one
Unified CM Assistant-controlled line).
A maximum of 3500 Assistants and 3500 Managers (7000 total users) can be configured per cluster
using the 7,500-user or 10,000-user VM configuration.
A maximum of three pairs of primary and backup Unified CM Assistant servers can be deployed per
cluster if the Enable Multiple Active Mode advanced service parameter is set to True and a second
and third pool of Unified CM Assistant servers are configured.
In order to achieve the maximum Unified CM Assistant user capacity of 3500 Managers and 3500
Assistants (7000 users total), multiple Unified CM Assistant server pools must be defined. As illustrated
in Figure 18-11, up to three pools can be configured. Each pool consists of a primary and backup
Unified CM Assistant server and a group of Managers and Assistants. Pool 1's Unified CM Assistant
servers are configured with the Cisco IPMA Server (Primary/Backup) IP Address service parameters,
Pool 2's servers are configured with the Pool2: Cisco IPMA Server (Primary/Backup) IP Address
advanced service parameters, and Pool 3's servers are configured with the Pool3: Cisco IPMA Server
(Primary/Backup) IP Address advanced service parameters.
18-26
Chapter 18
Figure 18-11
Unified CM
Assistant
servers
Primary
Pool 1
(default)
IP
IP
Managers
IP
Backup
M
IP
Assistants
IP
IP
Primary
Managers
IP
Pool 2
Backup
M
IP
Assistants
IP
IP
Primary
Managers
IP
Pool 3
Backup
IP
Assistants
191957
The Cisco Unified CM Assistant application interacts with the CTIManager for line monitoring and
phone control. Each line (including Intercom lines) on a Unified CM Assistant or Manager phone
requires a CTI line from the CTIManager. In addition, each Unified CM Assistant route point requires a
CTI line instance from the CTIManager. When you configure Unified CM Assistant, the number of
required CTI lines or connections must be considered with regard to the overall cluster limit for CTI lines
or connections. (For more information on CTI connection limits per cluster, see Capacity Planning for
CTI, page 9-30.) If additional CTI lines are required for other applications, they can limit the capacity
of Unified CM Assistant.
18-27
Chapter 18
Unified CM Assistant
With Unified CM Assistant in proxy line mode, the proxy line number(s) on the assistant phone
should be unique, even across different partitions.
With Unified CM Assistant in proxy line mode, two Managers cannot have the same
Unified CM Assistant controlled line number (DN), even across different partitions.
When enabling Multiple Active Mode and using more than one Unified CM Assistant server pool, ensure
that the appropriate server pool (1 to 3) is selected in the Assistant Pool field under the end user Manager
Configuration page so that Managers and Assistants are evenly distributed between the Unified CM
Assistant server pools. A Manager's associated Assistant will automatically be assigned to the pool
where their Manager is configured.
Unified CM Assistant supports a non-secure or secure connection (Transport Layer Security) to the CTI
Manager.
Some Cisco endpoints such as the Cisco TelePresence System EX90 might not support Cisco
Unified CM Assistant. For details, refer to the latest version of the Feature Configuration Guide for
Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
Note
Unified CM Assistants cannot use EM to log in to their phones because there is no concept of a Mobile
Assistant.
18-28
Chapter 18
Assistant_Everyone partition, containing all the Assistant and other user phone DNs
In addition, two calling search spaces are required, and for the example in Figure 18-12 they are as
follows:
That is the extent of the dial plan for this example. However, it is also important to properly configure
the various phone and Unified CM Assistant RP DNs or lines with the appropriate calling search spaces
so that call routing works as required. In this case all user, Assistant primary (or personal), and Manager
phone lines would be configured with the ASSISTANT_EVERYONE_CSS calling search space so that
all of these lines can reach all the DNs in the Assistant_Everyone and Assistant_Route_Point partitions.
Intercom lines and any other lines configured on devices within the telephony network would be
configured with this same calling search space. All Manager proxy lines and all Assistant_RP lines are
configured with the MANAGER_EVERYONE_CSS calling search space so that all of these lines can
reach the Manager DNs in the Assistant_Manager partition as well as all the DNs belonging to the
Assistant_Everyone partition. In this way, the dial plan ensures that only the Assistant_RP lines and the
Manager proxy lines on the Assistant phones are capable of reaching the Manager phone DNs directly.
18-29
Chapter 18
Unified CM Assistant
Figure 18-12
Calling Search
Spaces
IP
User Phone
Line[1]: 34001
ASSISTANT_EVERYONE_CSS
Partitions
Assistant_Route_Point
6XXXX
5XXXX
IP
Unified CM
Assistant Route
Point DNs
Manager Phone
Line[1]: 60001
Assistant
Primary Line
Assistant_Everyone
30001
IP
Assistant Phone
34001
Line[1]: 30001
Assistant and
All Phone DNs
38001
Line[2]: 38001
Manager
Proxy Line
50001
Assistant_RP
CTI
Line[1]: 6XXXX
Assistant_Manager
MANAGER_EVERYONE_CSS
Manager
Phone DNs
60001
153270
Line[2]: 5XXXX
50002
The example in Figure 18-12 shows the minimum dial plan requirements for Unified CM Assistant in
proxy line mode. However, most real-world telephony networks will have additional or existing dial plan
requirements that must be integrated with the Unified CM Assistant calling search spaces and partitions.
Figure 18-13 illustrates such an integration dial plan. In this example, the previously discussed dial plan
must now handle two additional partitions and an additional calling search space. The On Cluster
partition has been added in Figure 18-13, and it contains some additional phone DNs. The On Cluster
partition has been added to both of the existing Unified CM Assistant calling search spaces
(ASSISTANT_EVERYONE_CSS and MANAGER_EVERYONE_CSS) so that existing devices can
reach these added DNs. The UNRESTRICTED_CSS calling search space has also been added to the
existing dial plan. This calling search space is configured with the Assistant_Route_Point,
Assistant_Everyone, and the recently added On Cluster partitions. In addition, a second new partition
called PSTN has been added, and it contains a set of route patterns used for routing calls to the PSTN
via the common route list (RL), route group (RG), and voice gateway mechanism. This PSTN partition
is configured as part of the UNRESTRICTED_CSS calling search space.
Phone and device line calling search space configurations may be adjusted to incorporate the newly
added partitions and calling search spaces, provided the Assistant_RP and Assistant phone Manager
proxy lines remain assigned to the MANAGER_EVERYONE_CSS calling search space. In this example,
the Manager phone line has been moved from the originally configured ASSISTANT_EVERYONE_CSS
calling search space to the new UNRESTRICTED_CSS because it is likely that a Manager would be
given unrestricted access to the PSTN.
18-30
Chapter 18
Figure 18-13
IP
User Phone
Line[1]: 34001
ASSISTANT_EVERYONE_CSS
Partitions
Assistant_Route_Point
6XXXX
5XXXX
Assistant
Primary Line
IP Assistant Phone
Assistant_Everyone
30001
Line[1]: 30001
34001
Line[2]: 38001
38001
Manager
Proxy Line
MANAGER_EVERYONE_CSS
Assistant_RP
CTI
Unified CM
Assistant Route
Point DNs
On Cluster
35001
35002
Assistant and
All Phone DNs
Additional
Phone DNs
Assistant_Manager
Line[1]: 6XXXX
50001
Line[2]: 5XXXX
50002
Manager
Phone DNs
60001
IP
PSTN
Manager Phone
Line[1]: 60001
9.011!
UNRESTRICTED_CSS
RL
RG
9.911
V
153271
PSTN
As Figure 18-13 illustrates, integrating additional partitions and calling search spaces into a new or
existing Unified CM Assistant dial plan is feasible, but care must be taken to ensure that the underlying
proxy line mode mechanism remains intact.
For Unified CM Assistant shared line mode, no special dial plan provisioning is required. Manager and
Assistant phones can be configured with calling search spaces and partitions like any other phones in the
network because there are no Unified CM Assistant RPs or proxy lines to be concerned about. The only
requirement with regard to shared line mode is that the Manager and Assistant DNs must be in the same
partition so that shared line functionality is possible.
18-31
Chapter 18
Unified CM Assistant
18-32
Chapter 18
While the additional traffic generated by directory searches within the desktop application is nominal,
this traffic can be problematic in centralized call processing deployments when one or more
Unified CM Assistant console applications are running at remote sites. A directory search resulting in a
single entry generates approximately one (1) kilobit of traffic from the Unified CM Assistant server to
the desktop application. Fortunately, a maximum of 25 entries can be retrieved per search, meaning that
a maximum of approximately 25 kilobits of traffic can be generated for each search made by the desktop
application. However, if directory searches are made by multiple Unified CM Assistant Console desktop
applications across low-speed WAN links from the Unified CM Assistant server, the potential for
congestion, delay, and queuing is increased. In addition, directory retrieval traffic is sourced from
Unified CM on TCP port 2912, like all other Unified CM Assistant traffic to the desktop. This means
that directory retrieval traffic is also marked with DSCP 24 (PHB CS3) and therefore is queued like call
signaling traffic. As a result, directory retrieval could potentially congest, overrun, or delay call control
traffic.
Note
If a directory search generates more than 25 entries, the assistant is warned via a dialog box with the
message: Your search returned more than 25 entries. Please refine your search.
Given the potential for network congestion, Cisco recommends that administrators encourage
Unified CM Assistant Console users to do the following:
To reduce the number of entries returned, enter as much information as possible in the Name field
and avoid wild-card or blank searches when using the feature.
These recommendations are especially important if either of the following conditions is true:
There are many assistants separated from the Unified CM and/or Unified CM Assistant servers by
low-speed WAN links.
18-33
Chapter 18
WebDialer
WebDialer
WebDialer is a click-to-call application for Unified CM that enables users to place calls easily from their
PCs using any supported phone device. There is no requirement for administrators to manage CTI links
or build JTAPI or TAPI applications because Cisco WebDialer provides a simplified web application and
HTTP or Simple Objects Access Protocol (SOAP) interface for those who want to provide their own user
interface and authentication mechanisms.
This section examines the following design aspects of the WebDialer feature:
WebDialer Architecture
The WebDialer application contains two servlets: the WebDialer servlet and the Redirector servlet. Both
servlets are enabled when the Cisco WebDialer Web service is activated on a subscriber server. While
related, they each serve different functions and can be configured to run simultaneously.
WebDialer Servlet
Figure 18-14 illustrates a simple WebDialer example. In this example, user John Smith launches
WebDialer from a web-based or desktop application (step 1). WebDialer responds with a request for
login credentials. The user must respond with a valid userID and password as configured in the
Unified CM end-user directory. In this case, John Smith submits userID = jsmith and password = cisco
(step 2). Next, based on this login, WebDialer responds with the Cisco WebDialer Preferences
configuration page, and the user must indicate either User preferred device or Use Extension
Mobility (assuming the user has an EM device profile). In this case, user John Smith selects User
preferred device and selects the appropriate MAC address (SEP00036BC7B973) and directory number
(10001) for his phone from drop-down menus on the configuration page (step 3). Finally, the user is
presented with a screen requesting the phone number to be called (this value may already be indicated)
and must click Dial. In this case, John Smith enters 10002 and, after clicking Dial, a call is automatically
generated from his phone to Phone B at number 10002 (step 4).
Note
If the user has previously logged in to the WebDialer application and a web browser and server cookie
are still active, the user will not be prompted to log in again during subsequent requests. The user will
be prompted to log in again when the cookie has been cleared at the browser or by a restart of the
WebDialer server. Alternatively, the user web browser cookie can be set to expire automatically after a
certain number of hours as configured by the User Session Expiry WebDialer service parameter.
18-34
Chapter 18
Figure 18-14
UserID: jsmith
Password: cisco
User launches WebDialer 1
via web-based or
desktop application
Phone A
IP
DN: 10001
MAC: 00036BC7B973
Cisco Unified CM
(w./ WebDialer service)
IP
4
User clicks Dial and call is
generated from user device
to destination phone
Phone B
DN: 10002
153275
Redirector Servlet
The Redirector servlet provides WebDialer functionality in a multi-cluster or distributed call processing
environment. This functionality allows the use of a single enterprise-wide web-based WebDialer
application between all Unified CM clusters. Figure 18-15 illustrates the basic operation of the
Redirector servlet as part of the WebDialer application. In this example, the enterprise has three
Unified CM clusters: New York, Chicago, and San Francisco. All three clusters have been configured
with a single WebDialer application. The San Francisco cluster has been designated as the Redirector.
The enterprise-wide web-based application points to the San Francisco Redirector and is launched by
the New York user (see step 1 in Figure 18-15). Next the Redirector requests user login, and the New
York user responds back with their userID and password (see step 2 in Figure 18-15).
Note
If the user has previously logged in to the WebDialer application and a web browser and server cookie
are still active, the user will not be prompted to log in again during subsequent requests. Alternatively,
the user web browser cookie can be set to expire automatically after a certain number of hours as
configured by the User Session Expiry WebDialer service parameter.
The Redirector then broadcasts an isClusterUser HTTPS request to every WebDialer in the enterprise
simultaneously (as configured in the List of WebDialers service parameter). In this example, the requests
go to the Chicago and New York WebDialer servers (see step 3 in Figure 18-15). Because the New York
user is local to the New York cluster, the New York WebDialer responds with a positive response (see
step 4 in Figure 18-15). Finally, the New York user is redirected to their local WebDialer server, which
will handle the application request (see step 5 in Figure 18-15). The user is not notified of the redirect;
however, the URL in the browser address bar will be changed as the user is redirected from the
18-35
Chapter 18
WebDialer
Redirector to the local WebDialer server). In this example, only one Redirector is deployed; but in order
to provide redundancy for the Redirector, configure the Redirector on multiple clusters, as discussed in
the section on Service and Component Redundancy, page 18-40.
Figure 18-15
Chicago
Cluster
WebDialer
M
10.2.1.10
M
San Francisco
Cluster
3
Redirector/
WebDialer
M
10.1.1.10
M
Redirector sends
isClusterUser HTTPS
request to every
WebDialer in enterprise
New York
Cluster
WebDialer
10.3.1.10
4
NY WebDialer
(local to user)
sends positive
response.
2 Redirector requests
User launches web- 1
based application which
points to Redirector
5 User is redirected
IP
New York
user
Note
to local WebDialer
server which
handles request
153276
Web-based
application
Because the Redirector application is an enterprise-wide application that requires user authentication
against the Unified CM Database, Cisco highly recommends that all end-user userIDs be unique across
all Unified CM clusters. If they are not, then it is possible that more than one positive response to the
isClusterUser request could be received by the Redirector application. If this happens, the user will be
asked by the Redirector application to select their local WebDialer server manually. The user will then
have to know which server is their local server. If the wrong server is chosen, the WebDialer request will
fail.
18-36
Chapter 18
WebDialer Architecture
The architecture of the WebDialer application is as important to understand as its functionality.
Figure 18-16 depicts the message flows and architecture of WebDialer. The following sequence of
interactions and events can occur:
1.
WebDialer user phones register and make and receive calls via the Cisco CallManager service (see
step 1 in Figure 18-16).
2.
The WebDialer application on the user's PC communicates with the Cisco WebDialer Web Service
(see step 2 in Figure 18-16) via one of the following interfaces:
HTML over HTTPS
This interface is used by web-based applications based on the HTTPS protocol. This is the only
interface that provides access to the WebDialer and Redirector servlets.
Simple Object Access Protocol (SOAP) over HTTPS
The WebDialer Web service reads user and phone information from the Unified CM Database (see
step 3 in Figure 18-16).
4.
The WebDialer Web service in turn interacts with the CTIManager service for exchanging line and
phone control information (see step 4 in Figure 18-16).
5.
The CTIManager service passes WebDialer phone control information to the Cisco CallManager
service (see step 5 in Figure 18-16).
Figure 18-16
WebDialer Architecture
Cisco Unified CM
Cisco
Unified CM
service
IP
5
Cisco
CTIManager
service
2
WebDialer
Cisco WebDialer
Web Service
service
Cisco
Unified CM
Database
254302
18-37
Chapter 18
WebDialer
Note
Although Figure 18-16 shows the Cisco CallManager, CTIManager, and WebDialer Web Service
services all running on the same node, this configuration is not a requirement. These services can be
distributed among multiple nodes in the cluster, but they are shown on the same node here for ease of
explanation.
WebDialer URLs
The WebDialer application can be accessed from web-based applications via the HTML-over-HTTPS
interface using the following URLs:
WebDialer servlet
https://<Server-IP_Addr>:8443/webdialer/Webdialer?destination=<Number_to_dial>
(where <Server_IP-Address> is the IP address of any node in the cluster running the Cisco
WebDialer Web Service service, and where <Number_to_dial> is the number that the WebDialer
user wishes to dial)
Redirector servlet
https://<Server-IP_Addr>:8443/webdialer/Redirector?destination=<Number_to_dial>
(where <Server_IP-Address> is the IP address of any node in the enterprise running the Cisco
WebDialer Web Service service, and where <Number_to_dial> is the number that the WebDialer
user wishes to dial)
Figure 18-17 gives an example of HTML source code used in a click-to-call web-based application
calling the Cisco WebDialer application. In this example, the URL
https://10.1.1.1:8443/webdialer/Webdialer?destination=30271 in the HTML source view corresponds to
the "Phone: 30721" link for user Steve Smith within the web browser view. A user clicking on this link
would launch the WebDialer application and, after logging in and clicking Dial, would generate a call
from the user's phone to Steve Smith's phone. The same code could be used for a click-to-call application
using the Redirector function by changing the URL to
https://10.1.1.1:8443/webdialer/Redirector?destination=30271.
18-38
Chapter 18
Figure 18-17
153278
For information and examples of SOAP-over-HTTPS source code to be used in click-to-call desktop
applications, refer to the WebDialer API Programming information in the Cisco Unified
Communications Manager WebDialer Developer Guide, available at
https://developer.cisco.com/site/webdialer/webdialer/index.gsp
18-39
Chapter 18
WebDialer
Each WebDialer service can handle up to 4 call requests per second per node.
The following general formula can be used to determine the number of WebDialer calls per second (cps):
(Number of WebDialer users) ((Average BHCA) / (3600 seconds/hour))
When performing this calculation, it is important to estimate properly the number of BHCA per user that
will be initiated specifically from using the WebDialer service. The following example illustrates the use
of these WebDialer design calculations for a sample organization.
18-40
Chapter 18
Company XYZ wishes to enable click-to-call applications using the WebDialer service, and their
preliminary traffic analysis resulted in the following information:
Note
50% of all calls are dialed outbound, and 50% are received inbound.
Projections estimate 30% of all outbound calls will be initiated using the WebDialer service.
These values are just examples used to illustrate a WebDialer deployment sizing exercise. User dialing
characteristics vary widely from organization to organization.
10,000 users each with 6 BHCA equates to a total of 60,000 BHCA. However, WebDialer deployment
sizing calculations must account for placed calls only. Given the initial information for this sizing
example, we know that 50% of the total BHCA are placed or outbound calls. This results in a total of
30,000 placed BHCA for all the users enabled for click-to-call using WebDialer.
Of these placed calls, the percentage that will be initiated using the WebDialer service will vary from
organization to organization. For the organization in this example, several click-to-call applications are
made available to the users, and it is projected that 30% of all placed calls will be initiated using
WebDialer.
(30,000 placed BHCA) 0.30 = 9,000 placed BHCA using WebDialer
To determine the number of WebDialer servers required to support a load of 9,000 BHCA, we convert
this value to the average call attempts per second required to sustain this busy hour:
(9,000 call attempts / hour) (hour/3600 seconds) = 2.5 cps
Each WebDialer service can support up to 4 cps, therefore one node can be configured to run the
WebDialer service in this example. This would allow for future growth of WebDialer usage. In order to
maintain WebDialer capacity during a server failure, additional backup WebDialer servers should be
deployed to provide redundancy.
Keep in mind that the Cisco WebDialer application interacts with the CTIManager for phone control.
When enabled, each WebDialer service opens a single persistent CTI connection to the CTIManager. In
addition, each WebDialer individual MakeCall (or EndCall) request generates a temporary CTI
connection. The number of CTI connections required to handle WebDialer call rates also applies against
the CTI connection limits per cluster. (For more information on CTI connection limits per cluster, see
Capacity Planning for CTI, page 9-30.)
The administrator should ensure that all WebDialer users are associated with a phone or device
profile in the Unified CM end-user directory.
If the user selects "Use permanent device" under the Cisco WebDialer Preferences screen with
no phone association, then the following message is received when the Dial button is pressed:
No supported device configured for user
18-41
Chapter 18
If the user selects Use Extension Mobility under the Cisco WebDialer Preferences screen with
no device profile association (or the user is not logged in using a profile), then the following
message is received when the Dial button is pressed:
Call to <dialed_ number> failed: User not logged in on any device
An application interfaces with the WebDialer and Redirector servlets through HTTPS.
If using Client Matter Codes (CMC) or Forced Authorization Codes (FAC), WebDialer users must
enter the proper code at the tone by using the phone's keypad. Failure to enter the appropriate code
at the tone will result in call failure signaled by a reorder tone.
Cisco WebDialer is available on any Cisco endpoints that support Cisco Computer Telephony
Integration (CTI).
For a list of Cisco endpoints that support Cisco Computer Telephony Integration (CTI), refer to the Cisco
CTI Supported Device Matrix, available at
http://developer.cisco.com/web/jtapi/wikidocs/-/wiki/Main/Cisco+CTI+Supported+Device+Matrix
Cisco Unified Attendant Consoles have a client attendant console application that installs on an
attendant's Windows PC. With Cisco Unified Attendant Console Standard, there is no attendant console
server application installed. However, with Cisco Unified Attendant Console Advanced, an attendant
console server application is installed on a separate physical server than Unified CM. The attendant
console application communicates with the attendant console server application, and the attendant
console server application communicates with Unified CM securely through CTI and AXL over Secure
Socket Layer (SSL). Multiple attendant consoles can connect to a single attendant console server.
This section examines the following design aspects of the attendant consoles:
18-42
Chapter 18
A call comes into Unified CM, and the called number matches the directory number configured on
a CTI route point.
2.
The CTI route point is CTI-controlled by the attendant console server application and is associated
with a Queue Direct Dial In (DDI) configured on the server.
3.
The attendant console server application immediately redirects the call internally to one of its
Computer Telephony (CT) Gateway Devices. As part of this process, the attendant console server
application sends a CTI redirect message to the CTI Manager service to redirect the call to a CTI
port.
Note
A CTI redirect message does not result in a connected call; the call is not answered and there is
no media connection.
4.
The attendant console server application now associates the call with the CT Gateway Device and
controls the call on the CTI port.
5.
At this point, the call is presented to the attendant console client applications in the system that are
associated with the Queue DDI.
6.
Once an attendant chooses to answer the call through the attendant console client application,
another CTI redirect message is sent to the CTI Manager service, which moves the call from the CTI
port to the answering attendant's physical phone. The call is automatically connected on the
attendant's phone, either to the handset or the headset, depending on the phone configuration. The
region and location settings of the attendant's phone and the initiating gateway or phone dictate the
codec used for media.
7.
When a transfer to another extension is required, the attendant initiates the transfer through the
attendant console client application, which communicates the transfer to the attendant console
server application.
8.
The attendant console server application internally associates the call with a Service Queue and
sends a CTI redirect message to the CTI Manager service. This redirects the call from the attendant's
phone to a CTI port controlled by the attendant console server application.
Note
9.
A call transfer may also be initiated from the attendant's phone; however, this would remove the
attendant console server application from the call flow, and enhanced functionality (such as the
transfer recall feature) would no longer be possible.
At this stage, the Service Queue actually answers the call (there is a short connect) before issuing
the transfer, therefore the Cisco Media driver installed on the attendant console server application
is invoked. The region and location settings of this CTI port and the call-initiating gateway or phone
dictate the codec used for media. The configured Music on Hold (MoH) audio sources of the CTI
port also affect the MoH heard by the caller. Transfers are performed in this manner so that the
attendant console client application still maintains control of the call if there is no answer. Once the
call is received by the final party, the attendant console server application is removed from the call
flow.
18-43
Chapter 18
Figure 18-18
Attendant Console
Client Application
Attendant Phones
Cisco Unified CM
IP
Cisco CallManager
Service
IP
5
Cisco TAPI TSP
CTI Call Control
CT Gateway
Devices
CTI Route
Points
3
Service
Queues
CTI Ports
Call Park
Devices
AXL API
Cisco Unified CM
Database
252946
Queue DDIs
The attendant console server application's call park function does not use the inherent call park feature
of Unified CM. Instead, it uses its own call park facility using Call Park Devices. Call Park Devices work
very much like the Service Queues as outlined in steps 7 to 9 of Figure 18-18. Similar to transfers, Call
Park Devices allow the attendant console server application to maintain control of the call for the
duration of the parked call.
With Cisco Unified Attendant Console Standard, there is no attendant console server application
installed. The attendant console client application communicates directly with Unified CM.
Publisher The primary server used by the clients. If this server fails, all attendant operators are
switched to the subscriber server. Once the publisher is running again, the operators are prompted
to reconnect to the publisher.
You should consider providing redundancy on both sides of the integration for both CTI and AXL
communication.
Regarding CTI, the attendant console server application uses the Cisco TAPI Telephony Service Provider
(TSP) plug-in (downloaded from Unified CM) to communicate with the CTI Manager service. Cisco
TSP allows for the configuration of a primary and backup CTI Manager service. Cisco recommends
enabling the CTI Manager service on at least two Unified CM subscriber nodes in the cluster to gain
resilience in case the primary CTI Manager service goes offline. In the event of an attendant console
server failure, resilience can be achieved by configuring a Call Forward Unregistered (CFU) and Call
18-44
Chapter 18
Forward CTI failure destination on all of the CTI route points associated with Queue DDIs. If the
attendant console server application is offline, calls will automatically follow the Call Forward setting.
For example, when redundant attendant consoles are deployed, calls can be forwarded to the Cisco
Unified Attendant Console subscriber server. With a single attendant console server, the destination
could be a Hunt Pilot number or a Directory Number (DN) associated with a single IP phone.
AXL communication is enabled by activating the Cisco AXL Web Service on a Unified CM node.
Multiple Unified CM nodes can have the Cisco AXL Web Service enabled, but the attendant console
server application has only a single entry for Unified CM connectivity. In the event of a failure, an
administrator could update this entry to a backup Unified CM node running the Cisco AXL Web Service.
When redundant Cisco Unified Attendant Consoles are deployed, the attendant console servers can be
configured on different Unified CM nodes for the AXL Web Service.
The Unified CM also has a series of CTI route points and CTI ports configured for integration with Cisco
Unified Attendant Console. These devices have a device pool and therefore are assigned a Unified CM
group, which specifies a prioritized list of the Unified CM call processing nodes responsible for
maintaining registration. When the primary Unified CM in the Unified CM group is offline, the CTI
route points and CTI ports have the ability to register with a secondary Unified CM node, thus allowing
for high availability of the CTI route points and ports themselves.
The following general design guidance applies to the attendant console server application
components:
Queue DDI
One unique Queue DDI is required for each unique incoming directory number in the system
that should be routed specifically to the attendant consoles.
CT Gateway Device
Every incoming call into a Queue DDI is immediately redirected to a CT Gateway Device.
Design the system so that the number of CT Gateway Devices can handle the maximum
expected number of incoming calls at any given time.
18-45
Chapter 18
Service Queue
Each time an attendant transfers a call or places a call on hold, a Service Queue is required. The
system should be designed so that there are enough Service Queues to sustain the maximum
number of calls that all attendants in the system are in the process of transferring or putting on
hold at any given time. A general guideline is to provide 3 or 4 Service Queues per attendant,
but some scenarios might require more.
Call Park Device
Each time an attendant invokes the Call Park feature through the attendant console client
application, a Call Park Device is required. This feature does not use the inherent Call Park
capability of Unified CM. Design the system so that there are sufficient Call Park Devices to
handle the maximum number of calls parked by all attendants in the system at any given time.
Every Queue DDI, CT Gateway Device, Service Queue, and Call Park Device configured in the
attendant console server application creates a CTI route point or CTI port in Unified CM. The
number of CTI connections required to handle the Cisco Unified Attendant Console Advanced
integration also counts toward the CTI connection limits per cluster. (For more information on CTI
connection limits per cluster, see Capacity Planning for CTI, page 9-30.)
The attendant console server application provides busy lamp field (BLF) monitoring of end-user
devices, but it is important to note that this does not use the same facility in Unified CM that
provides BLF speed dial capability. Instead, the attendant console server application communicates
through CTI with Unified CM to obtain line state information on monitored devices. Once the
attendant console server application monitors an end-user device, it continues monitoring this
device through CTI until the number of devices monitored for BLF reaches a certain level (2,000).
Once this limit is reached, the BLF plug-in begins to drop devices from the list of monitored devices
in order to add newly requested devices to the list, thus ensuring that the number of devices
monitored by the attendant console server through CTI does not exceed the limit (2,000). These
devices monitored through CTI also count toward the CTI limits in Unified CM.
With respect to Quality of Service (QoS), the attendant console server application, the attendant
console client application, and the Cisco TSP all send their traffic marked as Best Effort (DSCP=0).
If this traffic traverses a WAN or a link that is typically congested, packets must be marked to receive
preferential treatment through the network. For a complete list of the TCP port numbers associated
with these applications, refer to the Cisco Unified Attendant Console installation guides, available
with appropriate login authentication at
http://www.cisco.com/c/en/us/support/unified-communications/unified-attendant-consoles/pro
ducts-installation-guides-list.html
Cisco TSP is not aware of partitions. Therefore, if the same directory number (DN) exists in multiple
partitions, the monitored device might not be the correct DN.
Cisco Unified Attendant Console Advanced can also integrate with the Cisco IM and Presence
Service through the SIP SIMPLE protocol. For more information about this type of integration, refer
to the appropriate Cisco Unified Attendant Console administration guide, available at
http://www.cisco.com/en/US/products/ps7282/prod_maintenance_guides_list.html
For design guidance on Cisco Unified Attendant Consoles, refer to the documentation available at
http://www.cisco.com/en/US/products/ps7282/products_implementation_design_guides_list.h
tml
18-46
Chapter 18
18-47
Chapter 18
Figure 18-19
Page #2 Group B
Page #1 Group A
Multicast RTP
Multicast RTP
HTTP signaling
HTTP signaling
Unified CM
SIP Trunk
CTI
Paging
Server
348723
Paging Originator
This sequence describes how the Cisco Paging Server initiates a broadcast to one or more IP phones as
illustrated in Figure 18-19:
1.
The caller dials a predefined number in Unified CM. This number routes the call to the Cisco Paging
Server over either a SIP trunk or CTI route point.
2.
3.
The caller hears a low stall tone. While the Cisco Paging Server plays this tone, it is sending a
command via HTTP to each phone in the recipient group. The command requests each phone to join
the multicast group.
4.
Once all phones have joined the multicast group, the Cisco Paging Server plays a high go-ahead
tone. When the caller hears this tone, it indicates that the Cisco Paging Server is transmitting the
RTP stream from the calling IP phone as a multicast RTP stream out to the receiving phones. When
the caller speaks, their voice is sent to the receiving phones.
5.
When the caller hangs up, the Cisco Paging Server sends another request to each IP phone, this time
to leave the multicast group, and the broadcast is over.
18-48
Chapter 18
Not all Cisco IP phones are compatible with the Cisco Paging Server. For a current list, refer to the
compatibility information available at
http://www.singlewire.com/compatibility-matrix.html
The Cisco Paging Server does not recognize Do Not Disturb (DND) and will send to phones with
DND enabled.
Multicast Considerations
If the Cisco Paging Server and IP phones are on separate IP subnets, the routers in between those
two subnets must be configured for multicast routing.
The Cisco Paging Server does not require any particular method of multicast routing (SM, DM,
S-DM, SSM, and so forth).
Some wide area network environments do not support multicast routing. For those environments,
GRE tunnels may be built between sites and used to transport multicast.
The multicast media streams always use the G.711 mu-law codec. No other codecs are allowed or
supported.
The RTP flow is always unicast from the initiator to the Cisco Paging Server, and then multicast
from the Cisco Paging Server to the receiving phones. If the Cisco Paging Server is deployed
centrally, then pages may cross WAN boundaries.
Cisco Paging Server multicast media streams are not calls. They do not count against either RSVP
or locations-based call admission control. WAN engineers should budget for these multicast media
streams in addition to other voice flowing across the enterprise network. However, the call between
the initiator and the Cisco Paging Server is a normal voice call. This call is subject to normal call
admission control restrictions.
If you have a multisite deployment, Cisco recommends configuring the Paging Server to use a range
of multicast addresses rather than the single default address. The reason for this is that Internet
Group Management Protocol (IGMP) multicast joins are effective for the multicast address only, not
the address and the port. If two broadcasts are going on simultaneously to two different sites, a
phone at either site will send out an IGMP join to the multicast address only. If both sites use the
same single address, the RTP streams for both broadcasts will be sent to both sites.
Different phone models and firmware versions may use different IGMP versions, which can impact
switch configuration.
18-49
Chapter 18
Other Considerations
Inbound calls to DialCast must be G.711 mu-law calls. Calls arriving to DialCast using other codecs
must be transcoded.
The Cisco Paging Server does maintain a CTI connection to Unified CM, but the load that this CTI
connection places on Unified CM is very low. The resources that this connection requires remain
constant regardless of cluster size.
The Cisco Paging Server requires that any firewalls between the server and the phones not be
configured to use Network Address Translation (NAT).
Beginning in Cisco Paging Server 8.4, QoS values are set to Unified CM default values (DSCP CS3
for signaling and DSCP EF for media). Signaling (DSCP CS3) applies to CTI and SIP traffic, while
media (DSCP EF) applies to SIP and CTI-initiated RTP streams as well as outbound multicast RTP
streams. Paging Server DSCP values cannot be changed in the field. Customers that wish to use
DSCP values different from these must re-mark Paging Server traffic in the network.
For more information, refer to the latest version of the InformaCast Virtual Appliance Basic Paging
Installation and User Guide, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-manager-c
allmanager/products-installation-guides-list.html
18-50
CH A P T E R
19
This chapter describes the voice messaging solutions available in the Cisco Unified Communications
System. It includes the Cisco voice messaging products Cisco Unity Connection and Cisco Unity
Express, and it covers the design guidelines and best practices for deploying these products together with
Cisco Unified Communications Manager (Unified CM). This chapter also covers aspects of integration
with third-party voicemail systems using industry standard protocols.
Although this guide focuses on the messaging deployment scenarios with regard to Unified CM, Cisco
Unified Communications Manager Express (Unified CME) is also noted where applicable, especially
when used with Survivable Remote Site Telephony (SRST) fallback support in a centralized Unified CM
deployment.
This chapter covers the following topics:
The chapter begins with a short description of each of the products in the Cisco messaging solutions
portfolio and provides a simple overview of where each product fits in an enterprise Unified
Communications solution. Next, messaging deployment models form the basis of discussion for
voicemail integrations, which start with a definition of the various messaging deployment models and
then explain how each of the messaging deployment models fits into the various Unified CM call
processing deployment models. Cisco Unity Connection is discussed in this section, while Cisco Unity
Express has a dedicated section for its supported deployment models. Key design guidelines are covered
for interoperability available within the Cisco Voice Messaging product portfolio. Virtualization is also
covered along with the important design factors to be considered while designing the virtual system.
Many system-level design considerations and best practices, including transcoding and various
integrations with Cisco Unified Communications Manager, are explained in this section. In addition, this
chapter provides details on third-party voicemail integration for supported industry-standard protocols.
19-1
Chapter 19
This chapter presents a high-level design discussion and is focused on how the voice messaging products
fit into a collaboration system with Unified CM. For detailed design guidelines for each product as well
as interoperability information for third-party messaging and telephony systems, refer to the Cisco Unity
Connection design guides, available at
http://www.cisco.com/en/US/products/ps6509/products_implementation_design_guides_list.html
Described in:
Revision Date
Voicemail-only refers to a telephony voicemail integration where there is no access to the voicemail
via any messaging client.
Integrated messaging refers to voicemail with telephony access as well as voicemail-only access via
a messaging client.
Unified messaging refers to voicemail with telephony access as well as voicemail, email, and fax
access via a messaging client.
Table 19-2 shows which Cisco products support these types of messaging.
Table 19-2
Messaging Type
Voicemail-only
Yes
Yes
Integrated messaging
Yes
Yes
Unified messaging
Yes
No
19-2
Chapter 19
Note
For further details on Unified Messaging with Cisco Unity Connection, see Single Inbox with Cisco
Unity Connection, page 19-43.
Based on the above messaging types and definitions, the two messaging product options are:
For a complete comparison of product feature, refer to the Cisco Messaging Products: Feature
Comparison, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps2237/products_data_sheets_list.html.
For more information on scalability of voice messaging products, refer to the section on Voice
Messaging, page 25-41, in the chapter on Collaboration Solution Sizing Guidance, page 25-1.
This chapter focuses on the design aspects of integrating Cisco Unity Connection and Cisco Unity
Express with Cisco Unified Communications Manager (Unified CM). Cisco Unified CM provides
functionality for Session Initiation Protocol (SIP) trunks, which support integration directly to Cisco
Unity Connection without the need for a SIP proxy server.
For information on earlier releases of Cisco Unity Connection, Unity Express, and Unified CM or
Unified CM Express, refer to the appropriate online product documentation available at
http://www.cisco.com.
As mentioned, the design topics covered in this chapter apply to voicemail-only, unified messaging, and
integrated messaging configurations. Additionally, this chapter discusses design aspects of deploying
Cisco Unity Connection with Microsoft Exchange (2003, 2007, or 2010). Cisco Unity Connection and
Unity Express have no dependencies on an external message store.
For additional design information about Cisco Unity Connection, including integrations with other
non-Cisco messaging systems, refer to the Design Guide for Cisco Unity Connection, available at
http://www.cisco.com/en/US/products/ps6509/products_implementation_design_guides_list.html
For additional design information about Cisco Unity Express, including integrations with other
non-Cisco messaging systems, refer to the applicable product documentation, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps5520/index.html
19-3
Chapter 19
Single-site messaging
Cisco Unity Express also supports three primary messaging deployment models:
Note
Single-site messaging
The Cisco Unity Express supports centralized voice messaging for up to 10 Unified CMEs. For more
information, refer to the Cisco Unified Communications Manager Express documentation at
http://www.cisco.com/en/US/products/sw/voicesw/ps4625/index.html.
Although the call processing deployment models for Cisco Unified CM and Unified CME are
independent of the messaging deployment models for Cisco Unity Connection and Unity Express, each
has implications toward the other that must be considered.
Cisco Unity Connection messaging redundancy is available in an active/active configuration. For more
information, refer to the Design Guide for Cisco Unity Connection available at
http://www.cisco.com/en/US/products/ps6509/products_implementation_design_guides_list.html
All messaging deployment models support voicemail, integrated messaging, and unified messaging
installations.
Single-Site Messaging
In this model, the messaging systems and messaging infrastructure components are all located at the
same site, on the same highly available LAN. The site can be either a single site or a campus site
interconnected via high-speed metropolitan area networks (MANs). All clients of the messaging system
are also located at the single (or campus) site. The key distinguishing feature of this model is that there
are no remote clients.
19-4
Chapter 19
Centralized Messaging
In this model, similar to the single-site model, all the messaging system and messaging infrastructure
components are located at the same site. The site can be one physical site or a campus site interconnected
via high-speed MANs. However, unlike the single-site model, centralized messaging clients can be
located both locally and remotely.
Distributed Messaging
A distributed messaging model consists of multiple single-site messaging systems distributed with a
common messaging backbone. There can be multiple locations, each with its own messaging system and
messaging infrastructure components. All client access is local to each messaging system, and the
messaging systems share a messaging backbone that spans all locations. Message delivery from the
distributed messaging systems occurs via the messaging backbone through a full-mesh or hub-and-spoke
type of message routing infrastructure.
Distributed messaging is essentially multiple, single-site messaging models with a common messaging
backbone. The exception to this rule is the PBX-IP Media Gateway (PIMG) and T1-IP Media Gateway
(TIMG) integrations. PIMG and TIMG integrations are not discussed in this design document. For
further information regarding PIMG or TIMG, refer to the Cisco Unity Connection integration guides
available at
http://www.cisco.com/en/US/products/ps6509/index.html
The distributed messaging model has the same design criteria as centralized messaging with regard to
local and remote GUI clients, TRaP, and message downloads.
Model Type
Cisco Unity
Connection
Cisco Unity
Express
Yes
Yes
Yes
No1
Yes
Yes
Yes
No1
Yes
Yes
Yes
No
Yes
Yes
1. Support for centralized voicemail messaging with Unified CME is available with Cisco Unity Express; however, this is not
applicable to Unified CM call processing deployment models.
19-5
Chapter 19
Each topic defines a messaging and Unified CM deployment model combination and then highlights
each Cisco voicemail messaging product applicable to that model as well as the design considerations
for that model combination. Not all combinations are discussed for each product. Some examples are
provided, with best practices and design considerations for each product. The intention is to provide an
understanding of the base messaging deployment models and the interaction with Unified CM without
detailing all possibilities.
For further details on site classification and a detailed analysis of supported combinations of messaging
and call processing deployment models, refer to the Design Guide for Cisco Unity Connection, available
at
http://www.cisco.com/en/US/products/ps6509/products_implementation_design_guides_list.html
19-6
Chapter 19
Figure 19-1
Central Site
Unified/Integrated
Messaging
Switch
101
Messaging System
Servers
100
PC
IP
IP
Transcoder
G.711
G.729
Unity
Connection
M
M
Region 1
IP WAN PSTN
Remote Site
Region 2
SRST or E-SRST
Unified/Integrated
Messaging
201
Switch
200
IP
IP
348724
PC
In Figure 19-1, regions 1 and 2 are configured to use G.711 for intra-region calls and G.729 for
inter-region calls.
As Figure 19-1 shows, when a call is made from extension 200 to the voicemail ports in Region 1, the
inter-region G.729 codec is used at the endpoint but the RTP stream is transcoded to use G.711 on the
voice ports. Unified CM transcoding resources must be located at the same site as the voicemail system.
Impact of Non-Delivery of RDNIS on Voicemail Calls Routed by AAR
In centralized messaging environments, automated alternate routing (AAR), a Unified CM feature, can
route calls over the PSTN to the messaging store at the central site when the WAN is oversubscribed.
However, when calls are rerouted over the PSTN, Redirected Dialed Number Information Service
(RDNIS) can be affected. Incorrect RDNIS information can impact voicemail calls that are rerouted over
the PSTN by AAR when Cisco Unity Connection is remote from its messaging clients. If the RDNIS
information is not correct, the call will not reach the voicemail box of the dialed user but will instead
receive the auto-attendant prompt, and the caller might be asked to re-enter the extension number of the
party they wish to reach. This behavior is primarily an issue when the telephone carrier is unable to
ensure RDNIS across the network. There are numerous reasons why the carrier might not be able to
ensure that RDNIS is properly sent. Check with your carrier to determine if they provide guaranteed
RDNIS deliver end-to-end for your circuits. The alternative to using AAR for oversubscribed WANs is
simply to let callers hear reorder tone in an oversubscribed condition.
19-7
Chapter 19
SRST or E-SRST at the Branch Site with Centralized Unified CM and Unity Connection, page 19-8
Multiple E-SRST or SRST Servers at the Branch Site with Centralized Unified CM and Unity
Connection, page 19-9
SRST or E-SRST at the Branch Site with Centralized Unified CM and Unity Connection
As shown in Figure 19-2, the central site contains Cisco Unified CM and Unity Connection to provide
primary call processing and voice messaging services under normal conditions. At the branch site, Cisco
Unified Enhanced Survivable Remote Site Telephony (E-SRST) and a Cisco Unity Connection SRSV
branch server are installed as a backup call agent and voice messaging server in the event of WAN
outage. Cisco Unity Connection installed at the central site uploads all phone and voice mailbox
information to the branch site SRSV server. SRST or E-SRST remains in the idle state until connectivity
to central site is lost. Once the branch site becomes isolated from central site and the keep-alive timer
between phones and Unified CM expires, branch phones are re-homed to the E-SRST or SRST router
which is preconfigured to send unanswered and busy calls to the Unity Connection SRSV branch server.
Subscribers can listen to voice messages left during a WAN outage by accessing voicemail. Upon WAN
restoration, all the voice messages are uploaded to a subscriber mailbox on the central Cisco Unity
Connection.
19-8
Chapter 19
Figure 19-2
SRST or E-SRST at Branch Site with Centralized Unified CM and Unity Connection
Branch site
IP
Central site
IP Phones
IP
M
M
M
SRSV
M
M
Unified CM
Local
Network
WAN
SRST
Branch site
Unity Connection
with native SRSV
IP
IP
IP Phones
Local
Network
IP
PSTN
Local
Network
SRSV
IP
IP Phones
IP
346375
SRST
Multiple E-SRST or SRST Servers at the Branch Site with Centralized Unified CM and Unity Connection
This deployment model is similar to first scenario, but multiple E-SRST and Cisco Unity Connection
SRSV branch servers are paired at the branch site for load balancing (see Figure 19-3). The administrator
must manually divide branch site users across two E-SRST servers using two different SRST references
in Unified CM to achieve load balancing. Cisco Unity Connection pushes the mailbox information to the
appropriate paired Cisco Unity Connection SRSV branch server. With this configuration, each Cisco
Unity Connection SRSV branch server contains the mailboxes for users on a single branch E-SRST.
Each Cisco Unity Connection SRSV branch server handles calls forwarded from its paired E-SRST
router in the event of WAN outage. Similar to the first scenario, the Cisco Unity Connection SRSV
branch server uploads all voicemail to a subscriber mailbox in the central Cisco Unity Connection upon
WAN restoration.
19-9
Chapter 19
Figure 19-3
Multiple E-SRST or SRST Servers at Branch Site with Centralized Unified CM and
Unity Connection
Central site
Branch site
IP
M
M
Local
Network
IP Phones
Unified CM
WAN
IP
SRST
SRSV
Pair 1
Gateway
Local
Network
IP
IP
IP
IP Phones
IP Phones
Note
PSTN
IP
IP
SRST
SRSV
Pair 2
346376
Unity Connection
with native SRSV
Pairing a single Cisco Unity Connection SRSV branch server with multiple E-SRST servers at a branch
site is not supported.
The maximum number of supported remote sites is 10 per central Cisco Unity Connection.
This solution supports both fallback methods, SRST and Enhanced SRST (E-SRST). The Cisco
Unity Connection SRSV branch server runs on the Cisco Services-Ready Engine (SRE) 900 and 910
blade servers and any supported Cisco Unity Connection platform such as Cisco Unified Computing
System (UCS) or UCS E-Series. Both the Cisco Unity Connection SRSV branch server and the
SRST or E-SRST router appear as a single logical unit, where the SRST router handles all control
signaling in the event the of a WAN outage.
The Cisco Unity Connection SRSV branch server becomes active if the WAN link goes down and
SRST is in active state. Otherwise it remains in the idle state.
Use HTTP over Secure Socket Layer (SSL) protocol to secure the connection between Cisco Unity
Connection and Cisco Unity Connection SRSV.
SRSV uses bandwidth from the WAN link during the following activities:
Configuration uploads from Cisco Unity Connection to Cisco Unity Connection SRSV
Uploading of voice messages from the branch Cisco Unity Connection SRSV server to the central
Cisco Unity Connection when the WAN link is restored
19-10
Chapter 19
Central
Site
Unified
Messaging
Switch
101
100
PC
IP
IP
Transcoder
M
M
Unity
Connection
Region 1
IP WAN PSTN
Remote
Site2
Messaging
System
Servers
Region 2
Cisco
Unity
Express,
SRST, or
E-SRST
Unity
Connection
Transcoder
Unified
Messaging
201
Switch
200
IP
IP
348725
PC
19-11
Chapter 19
For the configuration in Figure 19-4, transcoder resources must be local to each Cisco Unity Connection
message system site. Regions 1 and 2 are configured to use G.711 for intra-region calls and G.729 for
inter-region calls.
Voice messaging ports for both Cisco Unity Connection servers must be assigned the appropriate region
and location by means of calling search spaces and device pools configured on the Unified CM server.
In addition, to associate telephony users with a specific group of voicemail ports, you must configure
Unified CM voicemail profiles. For details on configuring calling search spaces, device pools, and
voicemail profiles, refer to the applicable version of the Cisco Unified Communications Manager
Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
Cisco Unity Connection supports digital and HTTPS networking, which enables multiple Unity
Connection clusters deployed over a WAN to communicate with each other. Using digital or HTTPS
networking, multiple Unity Connection clusters can share common directory information. This allows
users on multiple clusters to leave voicemail to each other. The Cisco Unity Connection cluster can
integrate with a corporate directory such as Microsoft Active Directory to synchronize user information
and can use digital or HTTPS networking to share the directory information at the same time.
Cisco Unity Connection with E-SRST
E-SRST offers the possibility for Cisco Unity Connection servers located in remote sites and registered
with a Unified CM at the central site to fall-back to E-SRST in the remote location. When the WAN link
is down and the phones fail-over to the E-SRST router, Cisco Unity Connection voicemail ports can also
fail-over to E-SRST mode to provide the remote site users with access to their voicemail with MWI
during the WAN outage.
Note
MWI has to be resynchronized from the Cisco Unity Connection server whenever a failover happens
from Unified CM to E-SRST mode, or vice versa.
19-12
Chapter 19
Central
Site
Unified
Messaging
Switch
101
100
PC
IP
IP
Transcoder
M
M
Unity
Connection_1
Remote
Site3 PC
300
301
Remote
Site2
Transcoder
Unified
Messaging
Region 1
Cisco Unity
Express, SRST, or
E-SRST
Switch
IP WAN PSTN
IP
IP
Region 3
Region 2
Cisco Unity
Express, SRST,
or E-SRST
Messaging
System
Servers
Unified
Messaging
Unity
Connection_2
201
Switch
200
IP
IP
348726
PC
Figure 19-5 shows the combination of two messaging models. Regions 1 and 3 use centralized
messaging with centralized call processing, while Region 2 uses distributed messaging with centralized
call processing. All regions are configured to use G.711 for intra-region calls and G.729 for inter-region
calls.
In Figure 19-5, centralized messaging and centralized call signaling are used between the Central Site
and Site3. The messaging system at the Central Site provides messaging services for clients at both the
Central Site and Site3. Site2 uses the distributed messaging model with centralized call processing. The
messaging system (Unity Connection 2) located at Site2 provides messaging services for only those
19-13
Chapter 19
users located within Site2. In this deployment, both models adhere to their respective design guidelines
as presented in this chapter. Transcoding resources are located locally to each messaging system site,
and they support clients who access messaging services from a remote site (relative to the messaging
system), as in the case of a Site2 user leaving a message for a Central Site user.
In addition, E-SRST mode is used for call processing backup of both IP phones and Cisco Unity
Connection voicemail ports. Deployed at the remote site (for example, Region 2 in Figure 19-5), this
fallback support provides backup call processing in the event that the phones lose connectivity with
Unified CM, such as during a WAN failure, while simultaneously providing users at the remote site with
access to the local Cisco Unity Connection server as well as MWI support during WAN failure. For
further details on E-SRST, refer to the product documentation available at
http://www.cisco.com/en/US/products/sw/voicesw/ps2169/index.html
19-14
Chapter 19
Figure 19-6
Cisco Unity Connection Centralized Messaging and Clustering Over the WAN with
Local Failover
Site 1
Unified/
Integrated
Messaging
Publisher/
TFTP
PC
101
Switch
IP
100
IP
Sub1
BKUP1
Transcoder
Messaging System
Servers
Region 1
Location 1
Unity
Connection
V
Primary
voice ports
IP WAN PSTN
Secondary
voice ports
Site 2
Sub2
BKUP2
Region 2
Location 2
Unified/Integrated
Messaging
201
Switch
200
IP
IP
292498
PC
For minimum bandwidth requirements between clustered servers see the section on Local Failover
Deployment Model, page 10-45.
Clustering over the WAN with Unified CM supports up to eight sites, as does Cisco Unity Connection.
The voicemail ports are configured only at the site where the Cisco Unity Connection messaging system
is located (see Figure 19-6). Voicemail ports do not register over the WAN to the remote site(s).
Messaging clients at the other site(s) access all voicemail resources from the primary site. There is no
benefit to configuring voice ports over the WAN to any of the remote sites because, in the event of a
WAN failure, remote sites would lose access to the centralized messaging system. Because of bandwidth
consideration, the voicemail ports should have TRaP disabled and all messaging clients should download
voicemail messages to their local PCs (unified messaging only).
19-15
Chapter 19
Cisco Unity Connection Distributed Messaging and Clustering over the WAN
Site 1
Unified
Messaging
101
PC
Switch
IP
100
IP
Publisher/TFTP
Transcoder
Sub1
BKUP1
Messaging
System
Servers
Unity
Connection
Region 1
Location 1
Primary
voice ports
IP WAN PSTN
Secondary
voice ports
Site 2
Sub2
BKUP2
Region 2
Location 2
Messaging
System
Servers
Transcoder
Unity
Connection
PC
201
IP
Switch
200
IP
292499
Unified
Messaging
In a purely distributed messaging implementation with clustering over the WAN, each site in the cluster
would have its own Cisco Unity Connection messaging server with messaging infrastructure
components. If not all of the sites have local Cisco Unity Connection messaging systems but some sites
have local messaging clients using a remote messaging server(s), this deployment would be a
combination model with both distributed messaging and centralized messaging. (See Combined
Messaging Deployment Models, page 19-13.) In the event of a WAN failure in this model, all remote
sites that use centralized messaging will lose voicemail capability until the WAN is restored.
19-16
Chapter 19
Each site that does not have a local messaging server must use a single messaging server for all of its
messaging clients, but all such sites do not have to use the same messaging server. For example, suppose
Site1 and Site2 each have a local messaging server. Site3 can then have all of its clients use (register
with) the messaging server at Site2, while Site4 can have all of its clients use the messaging server at
Site1. Transcoder resources are required at sites that have local Cisco Unity Connection messaging
server(s).
As with other distributed call processing deployments, calls going between these sites are transparent to
gatekeeper call admission control, therefore you must configure regions and locations in Unified CM to
provide call admission control. (See Managing Bandwidth, page 19-31.)
The distributed Cisco Unity Connection servers may also be networked using digital or HTTPS
networking.
Messaging Redundancy
Messaging redundancy is discussed in this section as it refers to Cisco Unity Connection. Cisco Unity
Express does not support messaging redundancy.
Unity
Connection
Active
Unity
Connection
Active
271573
Switch
Cisco Unity Connection SIP trunk implementation requires call forking for messaging redundancy
functionality. Cisco Unified Communications Manager supports the multi-destination SIP trunk feature.
With this multi-destination SIP trunk feature, administrators can define full-mesh trunking between
Cisco Unified CM and Cisco Unity Connection to achieve redundancy. Also, two separate SIP trunks
19-17
Chapter 19
can be configured, one for each server in a pair, and they can be added to the same route group associated
to the same route list.The route group should be configured in top-down order so that calls are sent to
the primary Unity Connection and overflow calls are sent to secondary Unity Connection server.
Note
SIP OPTIONS Ping should be enabled on the Cisco Unified CM SIP trunk for Cisco Unity Connection
failover to work properly.
19-18
Chapter 19
Figure 19-9
Site 1
Cisco Unity Connection Local Failover and Clustering Over the WAN
Unified
Messaging
Switch
101
PC
100
IP
IP
Publisher/TFTP
Transcoder
Sub1
BKUP1
Primary
voice ports
IP WAN PSTN
Sub2
BKUP2
Unified
Messaging
PC
201
Region 2
Location 2
Switch
IP
200
IP
292500
Secondary
voice ports
Site 2
Region 1
Location 1
Secondary
Primary
For information on configuring Cisco Unity Connection failover, refer to the Cisco Unity Connection
Failover Configuration and Administration Guide, available at
http://www.cisco.com/en/US/products/ps6509/index.html
19-19
Chapter 19
Figure 19-10
Unified CM
Cluster
Cisco Unity
Connection
Primary
Unified CM
Cluster
Cisco Unity
Connection
Secondary
IP
IP
IP
IP
IP
IP
MPLS/MAN
Data Center 2
Data Center 1
IP
Microsoft
Active
Directory
IP
Branch A
IP
IP
IP
IP
346378
Microsoft
Active
Directory
(150 ms RTT)
Branch B
Consider the following delay and bandwidth requirements when deploying Cisco Unity Connection
servers over different sites:
Note
Minimum of 7 Mbps bandwidth is required for every 50 ports. (For example, 250 ports require
35 Mbps.)
Bandwidth and latency requirements may differ for different versions of Cisco Unity Connection.
For a complete set of requirements, refer to the latest version of the System Requirements for Cisco Unity
Connection, available at
http://www.cisco.com/en/US/products/ps6509/prod_installation_guides_list.html
19-20
Chapter 19
Note
The Cisco Unity Connection cluster feature is also supported with Cisco Business Edition 6000.
Xcode
V
Messaging System
Servers
M
M
Unified CM Cluster
Unity
Connection
IP WAN/
PSTN
Xcode
Unified CM Cluster
292501
For the configuration in Figure 19-11, messaging clients at both Cluster 1 and Cluster 2 sites use the
Cisco Unity Connection messaging infrastructure physically located at Cluster 1.
19-21
Chapter 19
Deployment Models
Cisco Unity Express can be deployed as a single site or distributed voicemail and automated attendant
(AA) solution for Cisco Unified Communications Manager (Unified CM) or Unified Communications
Manager Express (Unified CME). However, Cisco Unity Express is supported with all of the Cisco
Unified CM deployment models, including:
Single-site deployments
Figure 19-12 shows a centralized call processing deployment incorporating Cisco Unity Express, and
Figure 19-13 shows a distributed call processing deployment.
Cisco Unity Express sites controlled by Unified CME, as well as other sites controlled by Unified CM,
can be interconnected with each other using SIP trunking protocol. Although Cisco Unity Express can
integrate with either Unified CM or Unified CME, it cannot integrate with both simultaneously.
Note
Cisco Unity Express supports a centralized deployment model with up to 10 Unified CMEs.
19-22
Chapter 19
Figure 19-12
Centralized Unity
Connection
Server
M
M
IP
IP
IP
VoIP
WAN
V
IP
IP
IP
IP
IP
IP
348727
PSTN
19-23
Chapter 19
Figure 19-13
Centralized Unity
Connection
Cisco
Server
Unified CM
Centralized Unity
Cisco
Connection
Unified CM
Server
M
M
M
IP
V
V
IP
IP
V
IP
VoIP
WAN
IP
IP
IP
IP
IP
IP
IP
IP
348728
PSTN
The most likely deployment model to use Cisco Unity Express is the multisite WAN model with
centralized call processing, where Cisco Unity Express provides distributed voicemail at the smaller
remote offices and a central Cisco Unity Connection system provides voicemail to the main campus and
larger remote sites.
Use Cisco Unity Express as a distributed voicemail solution if any of the following conditions apply to
your Unified CM network deployment:
Available WAN bandwidth is insufficient to support voicemail calls traversing the WAN to a central
voicemail server.
There is limited geographic coverage of the AA or branch site PSTN phone numbers published to
the local community, and these numbers cannot be dialed to reach a central AA server without
incurring toll charges.
The likelihood is high that a PSTN call into a branch office will be transferred from the branch AA
to a local extension in the same office.
Management philosophy allows remote locations to select their own voicemail and AA technology.
19-24
Chapter 19
The following characteristics and guidelines apply to Cisco Unity Express in either a centralized or
distributed Unified CM deployment:
A single Cisco Unity Express can be integrated with a single Unified CM cluster.
Cisco Unity Express integrates with Unified CM using a JTAPI application and Computer
Telephony Integration (CTI) Quick Buffer Encoding (QBE) protocol. CTI ports and CTI route points
control the Cisco Unity Express voicemail and automated attendant (AA) applications.
The following CTI route points are defined on Unified CM for Cisco Unity Express:
Automated attendant entry point (Cisco Unity Express can contain up to five distinct AAs and
The number of CTI ports and mailboxes supported for Cisco Unity Express on Unified CM depends
on the hardware platform. For details, refer to the Cisco Unity Express data sheet available at:
http://www.cisco.com/en/US/products/sw/voicesw/ps5520/products_data_sheets_list.html
For Cisco Unity Express deployments that require more than the maximum number of supported
mailboxes, consider using Cisco Unity Connection.
Each Cisco Unity Express mailbox can be associated with a maximum of two different extensions,
if needed.
The automated attendant function for any office deployed with Cisco Unity Express can be local to
the office (using the AA application in Cisco Unity Express) or centralized (using Cisco Unity
Express for voicemail only).
Cisco Unity Express can be networked with other Cisco Unity Expresses or with Cisco Unity
Connection via Voice Profile for Internet Mail (VPIM) version 2. Thus, a Cisco Unity Express
subscriber can send, receive, or forward messages to or from another remote Cisco Unity Express
or Cisco Unity Connection subscriber.
Cisco Unity Express allows you to specify up to three Unified CMs for failover. If IP connectivity
to all three Unified CMs is lost, Cisco Unity Express switches to Survivable Remote Site Telephony
(SRST) call signaling, thus providing AA call answering service as well as mailbox access to IP
phones and PSTN calls coming into the branch office.
Cisco Unity Express automated attendant supports dial-by-extension and dial-by-name functions.
The dial-by-extension operation enables a caller to transfer a call to any user endpoint in the
network. The dial-by-name operation uses the directory database internal to Cisco Unity Express
and does not interact with external LDAP or Active Directory databases.
Cisco Unity Express is not supported in pure SIP networks that do not have either Cisco Unified CM
or Unified CME controlling the SIP phones.
Cisco Unity Express can be deployed on a separate Unified CME or SRST router or a separate PSTN
gateway.
When Cisco Unity Express is deployed on a router separate from Unified CME or SRST, configure
the command allow-connections h323 to sip for H.323-to-SIP routing.
Figure 19-14 shows the protocols involved in the call flow between Unified CM and Cisco Unity
Express.
19-25
Chapter 19
Figure 19-14
Cisco
Unified CM
Cisco Unity
Express
M
M
VoIP
WAN
PSTN
TDM
IP
JTAPI (CTI-QBE)
IP
114709
H.323 or MGCP
SCCP or SIP
RTP
Cisco Unity Express is controlled via JTAPI (CTI-QBE) from Unified CM.
The Message Waiting Indicator (MWI) on the phone is affected by Cisco Unity Express
communicating a change of mailbox content to Unified CM via CTI-QBE, and by Unified CM in
turn sending a MWI message to the phone to change the state of the lamp.
The voice gateway communicates via H.323, SIP, or MGCP to Unified CM.
Real-Time Transport Protocol (RTP) stream flows carry the voice traffic between endpoints.
Figure 19-15 shows the protocols involved in the call flow between the router for SRST or E-SRST mode
and Cisco Unity Express when the WAN link is down.
19-26
Chapter 19
Figure 19-15
Protocols Used Between Cisco Unity Express and the Router for SRST or E-SRST
Cisco Unity
Express
Cisco
Unified CM
M
M
SRST
VoIP
WAN
TDM
IP
IP
114710
PSTN
SCCP or SIP
RTP
Note
Phones are controlled via SCCP or SIP from the router for SRST or E-SRST mode.
Cisco Unity Express communicates with the SRST router via an internal SIP interface.
Although MWI changes are not supported in SRST mode with previous releases of Cisco Unity
Express, voice messages can be sent and retrieved as during normal operation, but the MWI lamp
state on the phone remains unchanged until the phone registers again with Unified CM. At that time,
all MWI lamp states are automatically resynchronized with the current state of the users' Cisco
Unity Express voicemail boxes. Cisco Unity Express also supports MWI for SRST mode.
Cisco Unity Express supports SIP Subscriber/Notify and Unsolicited Notify to generate MWI
notifications, in both Unified CME and SRST modes.
SRST subscribes to Cisco Unity Express for MWI for each of the ephone-dns registered to receive
MWI notifications.
Voicemail Networking
This section covers specific considerations for voicemail networking, including Cisco Unity Connection
and Cisco Unity Express.
Voicemail networking is the ability to allow subscribers (voicemail users) to send, receive, reply to, and
forward voicemail messages between systems such as Cisco Unity Connection and Cisco Unity Express
using an embedded Simple Mail Transfer Protocol (SMTP) server and a subset of the Voice Profile for
Internet Mail (VPIM) version 2 protocol. Both voicemail messaging products support interoperability
between one another using VPIM messaging.
19-27
Chapter 19
Voicemail Networking
Subscribers can receive, send, and forward messages to or from another remote Cisco Unity Express
or Cisco Unity Connection for locations configured on the originating system.
Subscribers can also reply to a remote message received from a remote system.
Subscribers can be recipients of a distribution list or individual message originating from Cisco
Unity Connection.
For more information on voicemail networking with a specific product, refer to the corresponding
voicemail product documentation available at
http://www.cisco.com/en/US/products/ps6509/index.html
Digital Networking
Digitally networked systems use Simple Mail Transfer Protocol (SMTP) for both directory replication
and message transport. As shown in Figure 19-16, multiple Unity Connection nodes are joined in
full-mesh topology for sharing directory information. Only full-mesh topology is supported with Cisco
Unity Connection digital networking.
Using full-mesh topology for networking requires only a single hop for transport of information between
nodes, but the number of links increases with the number of nodes.
Figure 19-16
Digital Networking
Cluster A
Cluster B
Cluster C
348729
Directory replication
via SMTP
19-28
Chapter 19
Consider the following guidelines when deploying Cisco Unity Connection digital networking:
A single Cisco Unity Connection digital network supports a maximum of 100,000 users, but
multiple digital networks can be joined using Voice Profile for Internet Mail (VPIM) networking to
support more users. If any of the Cisco Unity Connection nodes in the digital network system is
running Cisco Unity Connection 7.0, then the maximum number of users supported is 50,000.
One Cisco Unity Connection can be a member of only one Cisco Unity Connection digital network.
Multiple Cisco Unity Connection digital networks can be joined using VPIM. Each Cisco Unity
Connection digital network must have one server defined as the bridgehead or site gateway. The
bridgehead or site gateway is used to communicate with other digital networks.
HTTPS Networking
HTTPS networking uses hub and spoke topology, which enables data replication in a tree structure. The
hub is a single point of communication for all the leaf spokes. All the directory replication occurs
through the hub using HTTPS protocol. Each spoke is a leaf node that gathers directory information from
the hub, and single spoke can connect to only one hub. As shown in Figure 19-17, spoke clusters A and
B are connected to hub cluster H. If cluster A needs to fetch any directory information, it sends a query
to node H. Hub node H replicates its own as well as node Bs directory information to node A.
Figure 19-17
HTTPS Networking
Cisco Unity Connection
Cluster
Cluster H (Hub)
Cluster A (Spoke)
Cluster B (Spoke)
348730
Directory replication
via HTTPS
Consider the following guidelines when deploying Cisco Unity Connection HTTPS networking:
A single Unity Connection node or cluster can be member of only one HTTP(S) network.
A single HTTPS network supports a maximum of 100,000 users and 150,000 contacts, but multiple
digital or HTTPS networks can be joined together using Voice Profile for Internet Mail (VPIM)
networking to support more than 100,000 users and/or contacts.
A single HTTPS network system supports a single site, and each site can have a maximum of
25 nodes; however, multiple HTTPS network systems can be joined using VPIM.
All the Cisco Unity Connection servers must be version 10.0 or higher to support HTTPS
networking.
19-29
Chapter 19
Voicemail Networking
In an HTTPS network, Cisco Unity Connection locations are joined together using a hub and spoke
topology. The number of direct HTTPS links to any location must be less than or equal to 5.
HTTPS networking cannot be used with digital networking at the same site; however, a single
HTTPS network can communicate with a digital network by using VPIM. Each Cisco Unity
Connection digital or HTTPS network must have one server defined as the bridgehead or site
gateway. The bridgehead or site gateway is used to communicate with other digital or HTTP(S)
networks.
Full synchronization occurs after any node or cluster is added to the HTTPS network. If any
discrepancy in directory data exists, then resynchronization occurs. HTTPS networking supports
both manual and automatic full synchronization and resynchronization. The periodic interval for
automatic synchronization is configurable.
Directory replication occurs through the publisher node in Unity Connection. If the publisher goes
down, then directory replication through this publisher node stops and a subscriber node provides
directory replication.
For more information on these interoperability options, refer to the latest version of the Networking
Guide for Cisco Unity Connection, available at
http://www.cisco.com/en/US/products/ps6509/prod_maintenance_guides_list.html
Note
The Tested Reference Configurations include selected Cisco Unified Computing System (UCS)
platforms. Other platforms may be supported with the specifications-based hardware support policy.
For VMware vSphere ESXi 5.1 and earlier, at least one processor core must be reserved for the VMware
ESXi hypervisor/scheduler. For VMware vSphere ESXi 5.5 and later, the Latency Sensitivity function
is included to reduce virtual machine latency. When the Latency Sensitivity is set to a high value, you
do not need to reserve any unused processor core for the ESXi hypervisor/scheduler.
For more information on deploying Cisco Unified Communications and Cisco Unity Connection in a
virtualized system, refer to the documentation available at
http://www.cisco.com/go/uc-virtualized
General information about deploying Unified Communications on virtualized servers is also available in
the section on Deploying Unified Communications on Virtualized Servers, page 10-53.
For Cisco Unity Connection virtualization, also refer to the latest version of the Design Guide for Cisco
Unity Connection available at
http://www.cisco.com/en/US/products/ps6509/products_implementation_design_guides_list.html
19-30
Chapter 19
Managing Bandwidth
Unified CM provides a variety of features for managing bandwidth. Through the use of regions,
locations, and even gatekeepers, Unified CM can ensure that the number of voice calls going over a WAN
link does not oversubscribe the existing bandwidth and cause poor voice quality. Cisco Unity Connection
relies on Unified CM to manage bandwidth and to route calls. If you deploy Cisco Unity Connection in
an environment where calls or voice ports might cross WAN links, these calls will be transparent to
gatekeeper-based call admission control. This situation occurs any time the Cisco Unity Connection
server is servicing either distributed clients (distributed messaging or distributed call processing) or
when Unified CM is remotely located (distributed messaging or centralized call processing).
Unified CM provides regions and locations for call admission control.
Figure 19-18 uses a small centralized messaging and centralized call processing site to illustrate how
regions and locations work together to manage available bandwidth. For a more detailed discussion of
regions and locations, refer to the chapter on Bandwidth Management, page 13-1.
19-31
Chapter 19
Figure 19-18
Central Site
Unified/Integrated
Messaging
Switch
101
Messaging System
Servers
100
PC
IP
IP
Unity
Connection
M
M
Region 1
Location 1
IP WAN PSTN
Remote Site
Region 2
Location 2
SRST
Unified/Integrated
Messaging
201
Switch
200
IP
IP
292504
PC
In Figure 19-18, regions 1 and 2 are configured to use G.711 for intra-region calls and G.729 for
inter-region calls. Locations 1 and 2 are both set to 24 kbps. Location bandwidth is budgeted only in the
case of inter-location calls.
An intra-region (G.711) call would not be budgeted against the available bandwidth for the location. For
example, when extension 100 calls extension 101, this call is not budgeted against the 24 kbps of
available bandwidth for Location 1. However, an inter-region call using G.729 is budgeted against both
bandwidth allocations of 24 kbps for Location 1 and Location 2. For example, when extension 100 calls
extension 200, this call would be connected but any additional (simultaneous) inter-region calls would
receive reorder (busy) tone.
19-32
Chapter 19
Which codecs will be negotiated between the majority of the endpoints and Cisco Unity
Connection? This will help you decide on which codecs need to be advertised by Cisco Unity
Connection and which do not. You can then decide on when you need Unified CM to provide
hardware transcoding resources in lieu of doing computationally significant native transcoding in
Cisco Unity Connection, such as when requiring a large number of clients connected to Cisco Unity
Connection using G.722 or iLBC.
Which types of graphical user interface (GUI) clients (web browsers, email clients, media players,
and so forth) will be fetching the recordings, and which codecs do the GUI clients support?
What quality of the sound is produced by the selected codec? Some codecs are higher quality than
others. For example, G.711 has a higher quality than G.729a, and it is a better choice if higher audio
quality is necessary.
How much disk space does the codec use per second of recording time?
Table 19-4 summarizes the characteristics of the codec formats supported by Cisco Unity Connection.
Table 19-4
Codec Characteristics
Recording Format
(Codec)
Audio Quality
Supportability
Linear PCM
Highest
Widely supported
16 KBps
Moderate
Widely supported
8 KBps
G.729a
Lowest
Poorly supported
1 KBps
G.726
Moderate
Moderately supported
3 KBps
GSM 6.10
Moderate
Moderately supported
1.6 KBps
19-33
Chapter 19
Refer to the System Administration Guide for Cisco Unity Connection for details on changing the codec
advertised by Cisco Unity Connection. The choices for advertised codecs are G.711 mu-law, G.711
a-law, G.729, iLBC and G.722. There is also a list of preferences according to how they are ordered in
the list (top-down). For SCCP integrations, the order of the codecs has no bearing because codecs are
advertised and Unified CM negotiates the codec based on the location of the port and device in the
negotiated call. For SIP integrations, however, the order list is significant. If the codec is preferred, then
Cisco Unity Connection will advertise that it supports both protocols but will prefer to use the one
specified over the other.
For information on how to change the system-level recording format in Cisco Unity Connection
Administration, refer to the System Administration Guide for Cisco Unity Connection.
19-34
Chapter 19
Unity
Connection
Cluster
H.323
MGCP
SIP
IP
IP
IP
IP
IP
IP
IP
IP
IP
IP
IP
IP
292509
Session
Management
Edition Cluster
The following information must be sent on the intercluster trunks between Unified Communications leaf
clusters and Unified CM Session Management Edition, and on the SIP trunk to Cisco Unity Connection:
Inbound and outbound redirecting number information element (IE) delivery on MGCP and H.323
gateways and H.323 trunks
Diversion information that is sent on non-Q.SIG MGCP, H.323, or SIP trunks picks up only the calling
party transformations that are defined by the voice mailbox mask of the voicemail profile that is assigned
to the redirecting DN. Any calling party transformations that are defined in a route pattern or route list,
or through outbound calling party transformation calling search spaces (CSSs), are not applied to
diversion information.
19-35
Chapter 19
Q.SIG-Enabled Trunk
For Q.SIG-enabled SIP, MGCP and H.323 trunks, the original called party number is sent in Q.SIG
diverting leg information application protocol data units (APDUs).
On Q.SIG-enabled H.323, MGCP, and SIP trunks all calling, called, and redirecting number information
is always sent in the encapsulated Q.SIG message and not in the outer H.323 message or SIP headers.
The sent diversion information does not pick up any calling party transformation and does not honor any
voicemail mask setting. Q.SIG tunneling-enabled trunks do not support transport of the + character in
Q.SIG APDUs. Because of this limitation, the users voice mailbox number should be of the same format
as the directory number used in the leaf Unified Communications system. For example:
Users with directory numbers of the format 4YYYY should have a corresponding voice mailbox
number of the same 4YYYY format.
Users with directory numbers of the E.164 format +XX4YYY should have a corresponding voice
mailbox number of the same E.164 +XX4YYYY format.
Cisco Unity Connection allows an alternate extension to be associated with the voice mailbox of the user.
For example:
Redirected Dialed Number Information Service (RDNIS) is not supported with Q.SIG-enabled H.323 or
SIP trunks. The original called party or redirecting number is sent in a Q.SIG DivertingLegInformation2
APDU instead of via RDNIS.
System contact phone numbers for the Cisco Unity Connection System
Personal call transfer rule (PCTR) phone numbers for the Cisco Unity Connection System
Message waiting indicator (MWI) extensions for the Cisco Unity Connection System
When importing users from LDAP with E.164-formatted primary phone numbers, use the regular
expression and replacement pattern that together convert phone numbers into extensions. For more
information on this, refer to the sections on converting phone numbers into extensions in the latest
version of the System Administration Guide for Cisco Unity Connection, available at
http://www.cisco.com/en/US/products/ps6509/prod_maintenance_guides_list.html
19-36
Chapter 19
If you want to import users from Cisco Unified Communications Manager (Unified CM) with E.164
formatted extensions through AXL integration, you will have to export the E.164 extensions from
Unified CM into a comma-separated values (CSV) file and perform the necessary translations on the
alternate extensions (in Excel, for example) prior to using the Bulk Administration Tool (BAT) to import
them into Unity Connection. For more details on using the Cisco Unity Connection Bulk Administration
Tool, refer to the latest version of the User Moves, Adds, and Changes Guide for Cisco Unity
Connection, available at
http://www.cisco.com/en/US/products/ps6509/prod_maintenance_guides_list.html
Attempt sign-in
Ring-no-answer (RNA)
An administrator can import URIs into Unity Connection from an LDAP directory or AXL integration
with Cisco Unified CM.
Note
The administrator cannot delete or edit an alternate extension with the Directory URI phone type. This
alternate extension can be edited or deleted only from its original source (LDAP directory or Cisco
Unified Communications Manager from which the user was imported).
19-37
Chapter 19
eMWI also works in a distributed call processing environment with centralized messaging over an
intercluster trunk (H.323 or SIP).
Note
SIP
SIP Trunk or
QSIG Intercluster Trunk
IP
292505
Figure 19-20
SIP Trunk or
SCCP Connection
Phone
Unity Connection
Unified CM Cluster 1
Unified CM Cluster 2
Figure 19-21 illustrates eMWI over an intercluster trunk (H.323 or SIP) in a distributed call processing
environment with centralized voice messaging.
Figure 19-21
Message-Waiting: yes
Message-Waiting: yes
Message-Waiting: yes
With voice message counts
292506
SIP
IP
Phone
Unity Connection
Unified CM Cluster 2
Unified CM Cluster 1
As shown in Figure 19-21, Cluster 2 and its voice messaging solution support eMWI, but Cluster 1 does
not. If an eMWI update with a voice message count is sent from the voice messaging solution intended
for the Cluster 2 phone, Cluster 1 will forward only a standard MWI to Cluster 2 without the voice
message count.
The following guidelines apply to eMWI:
All clusters should support eMWI. If an intermediate cluster does not support eMWI, then the
terminating cluster will receive a standard MWI only without voicemail counts.
Standard MWI does not generate much traffic because it sends only a change of lamp state (ON or
OFF). However, enabling eMWI can increase the amount of traffic because it also sends message
counts from the messaging system. The amount of traffic depends on the number of messages and
change notifications.
19-38
Chapter 19
Cisco Unity Connection Server(s) Integrated with a Unified CM Cluster (No Dedicated
Backup Servers)
Single Site
Publisher/TFTP
M
Sub1
Sub2
Messaging System
Servers
Unity
Connection
Secondary
voice ports
292507
Primary
voice ports
The Unified CM cluster in Figure 19-22 employs 1:1 server redundancy and 50/50 load balancing.
During normal operations, each subscriber server is active and handles up to 50% of the total server call
processing load. In the event of a subscriber server failure, the remaining subscriber server takes up the
load of the failed server.
This configuration uses two groups of voicemail ports, with each group containing one-half of the total
number of licensed voice ports. One group is configured so that its primary server is Sub1 and its
secondary (backup) server is Sub2. The second group is configured so that Sub2 is the primary server
and Sub1 is the backup.
Make sure that MWI-only ports or any other special ports are equally distributed between the two
groups. During the configuration of the voice ports, pay special attention to the naming convention.
When configuring the two groups of ports in Cisco Unity Connection, make sure that the device name
prefix is unique for each group and that you use the same device name when configuring the voicemail
ports in Unified CM Administration. The device name prefix is unique for each group of ports in this
example, with group Sub1 using CiscoUM1 as the device name prefix and Sub2 using CiscoUM2 in this
example.
19-39
Chapter 19
For additional design information on the ratio of inbound to outbound voicemail ports (for MWI,
message notification, and TRaP), refer to the Cisco Unity Connection System Administration Guide,
available at
http://www.cisco.com/en/US/products/ps6509/index.html
Note
The device name prefix is unique for each group of ports and must match the same naming convention
for the voicemail ports configured in Unified CM Administration.
In Unified CM Administration, half of the ports in this example are configured to register using the
unique device name prefix of CiscoUM1, and the other half are configured to register using the unique
device prefix CiscoUM2. (See Table 19-5.) When the ports register with Unified CM, half will be
registered with subscriber server Sub1, and the other half will be registered with Sub2, as shown in
Table 19-5.
Table 19-5
Device Name
Description
Device Pool
Status
IP Address
CiscoUM1-VI1
Unity Connection 1
Default
Standard Profile
1.1.2.9
CiscoUM1-VI2
Unity Connection 1
Default
Standard Profile
1.1.2.9
CiscoUM1-VI3
Unity Connection 1
Default
Standard Profile
1.1.2.9
CiscoUM1-VI4
Unity Connection 1
Default
Standard Profile
1.1.2.9
CiscoUM2-VI1
Unity Connection 1
Default
Standard Profile
1.1.2.9
CiscoUM2-VI2
Unity Connection 1
Default
Standard Profile
1.1.2.9
CiscoUM2-VI3
Unity Connection 1
Default
Standard Profile
1.1.2.9
CiscoUM2-VI4
Unity Connection 1
Default
Standard Profile
1.1.2.9
Note
The naming convention used for the voicemail ports in Unified CM Administration must match the
device name prefix used in Cisco UTIM, otherwise the ports will fail to register.
19-40
Chapter 19
Cisco Unity Connection Server(s) Integrated with a Single Unified CM Cluster with
Backup Subscriber Server(s)
Publisher/TFTP
Sub1
Publisher/TFTP
Sub2
Sub1
Sub2
M
M
BKUP1 BKUP2
BKUP
Messaging System
Servers
Messaging System
Servers
Primary
voice ports
Unity
Connection
Unity
Connection
292508
Secondary
voice ports
Configuration of the voicemail ports in this case is similar to the 50/50 load-balanced cluster. However,
instead of configuring the voice ports to use the opposite subscriber server as the secondary server, the
individual shared or dedicated backup server is used. In the Unified CM cluster with a shared backup
server, both of the secondary ports for the subscriber servers are configured to use the single backup
server.
The voice port names (device name prefix) must be unique for each Cisco UTIM group and must be the
same as the device names used on the Unified CM server.
To configure the voicemail ports on Cisco Unity Connection, use the Telephony Integration section of
the Unity Connection Administration console. For details, refer to the Cisco Unity Connection
administration guides available at
http://www.cisco.com/en/US/products/ps6509/index.html
19-41
Chapter 19
Note
Global Address
Voice messages are stored as .wav files and are independent of IPv6 or IPv4.
IPv6 support is disabled by default, but system administrators can enable IPv6 and configure IPv6
address settings either in Cisco Unified Operating System Administration or in the command line
interface (CLI). Cisco Unity Connection can obtain an IPv6 address either through router advertisement,
through DHCP, or from addresses configured manually either in Cisco Unified Operating System
Administration or through the CLI. Cisco Unity Connection Administration and Cisco Personal
Communications Assistant can be accessed using IPv6 addresses.
Note
IPv6 addressing cannot be enabled during installation or upgrade of Cisco Unity Connection. Cisco
Unity Connection does not support "IPv6 ONLY" server configuration. Cisco Unity Connection supports
Unicast only for IPv6.
Cisco Unity Connection over IPv6 supports following functionality:
Cisco Unity Connection offers auto-discovery functionality over IPv6, which allows Unity
Connection to search for Microsoft Exchange servers to communicate with.
Cisco Unity Connection can be integrated with an IPv6 Microsoft Exchange 2007 or 2010 server to
enable the Single Inbox feature.
Cisco ViewMail for Outlook (VMO) supports communication between Outlook and Cisco Unity
Connection over IPv6.
Voice messages received on Cisco Unity Connection can be accessed using any IMAP client such
as Outlook over IPv6.
Cisco Unity Connection can be integrated with LDAP over IPv6 to import the user information.
Cisco Unity Connection also offers Telephone Record and Playback (TRaP) functionality over IPv6,
which enables users to record or play back messages over an IPv6-enabled phone so that signaling
can happen over IPv6.
19-42
Chapter 19
Note
All voice messages remain on the Cisco Unity Connection server regardless of the type of messaging
deployment. Cisco Unity Connection is the authoritative source of voice messaging traffic, notifications,
and synchronizations.
The amount of space a single voicemail message can acquire is configured on the Cisco Unity
Connection server and is similar to message aging. The maximum size for a voicemail message is also
configured on the Microsoft Exchange Server. Typically, the Microsoft Exchange Server maintains a
larger size than Cisco Unity Connection that is synchronized to the mailbox. Hence, the minimum size
of the message in Microsoft Exchange should be bigger than the maximum size in Cisco Unity
Connection.
19-43
Chapter 19
From a security aspect for communications between Cisco Unity Connection and Microsoft Exchange,
HTTPS is chosen as the default option. HTTP is also supported but not recommended because it reduces
security and might also need further configuration on Microsoft Exchange. At the same time, there is an
option to validate the Microsoft Exchange certificate, provided that access to the certificate server is
available.
Cisco Unity Connection can be integrated with IBM Lotus Domino using software from Cisco partners
Esnatech and Donoma to enable the Single Inbox feature. Esnatech Office-LinXTM Cloud Connect
Edition and Donoma Unify for Lotus Notes both enable integration between Unity Connection and IBM
Lotus Domino. For more detail information on deployment, installation, and configuration of these
products, refer to the documentation available at http://www.esnatech.com/landing/cisco.htm and
http://donomasoftware.com/donoma-unify-for-lotus-notes/.
Cisco Unity Connection also supports integrated voice and fax services with cloud-based applications
such as Google Apps Gmail and VMware Zimbra. Unity Connection can be integrated with these email
applications using Esnatech Office-LinXTM Cloud Connect Edition. This solution allows for
bidirectional synchronization of voice messages between Cisco Unity Connection and these email
applications. (See Figure 19-24.)
Cisco Unity Connection Deployment with Cloud-Based Google Email Application
Esnatech Cloud
Connect
Cisco Unified
Communications
IP
VoIP (SIP)
IP
CTI (TAPI)
Unified CM Cluster
CUMI/CUPI
(REST API)
Esnatech
Office-LinX
Voice Engine
Rich
Presence
and Call Control
Synchronization
Services
Client
Application
Gateway
Unity Connection
Cluster
UC
Mobile
UC Client Desktop
(Windows and Mac)
Google Apps
Services
Mail
Sync
Calendar
Sync
Contact
Sync
UC Client
Web Gadget
292510
Figure 19-24
Esnatech Cloud Connect Edition uses the Cisco Unity Connection Messaging Interface (CUMI) API and
the Cisco Unity Connection Provisioning Interface (CUPI) API for synchronization of subscriber and
messaging information. Users can also initiate calls through Gtalk using the Click-to-Dial feature.This
call control information gets exchanged with Cisco Unified Communications Manager through the CTI
(TAPI) interface. For more detail information on deployment, installation and configuration, refer to the
documentation available at http://www.esnatech.com/landing/cisco.htm.
19-44
Chapter 19
Ensure that the IP phones having Cisco Unity Express as their voicemail destination are located on
the same LAN segment as the router hosting Cisco Unity Express.
If uninterrupted automated attendant (AA) and voicemail access is required for a site deployed with
Cisco Unity Express, ensure that Cisco Unity Express, SRST, and the PSTN voice gateway are all
located at the same physical site. Hot Standby Router Protocol (HSRP) or other redundant router
configurations are not currently supported with Cisco Unity Express.
Each mailbox can be associated with a primary extension number and a primary E.164 number.
Typically, this number is the direct-inward-dial (DID) number that PSTN callers use. If the primary
E.164 number is configured to any other number, use Cisco IOS translation patterns to match either
the primary extension number or primary E.164 number so that the correct mailbox can be reached
during SRST mode.
Each Cisco Unity Express site must be associated with a CTI route point for voicemail and one for
AA (if licensed and purchased), and you must configure the same number of CTI ports as Cisco
Unity Express ports licensed. Ensure that the number of sites with Cisco Unity Express does not
exceed the CTI scalability guidelines presented in the chapter on Call Processing, page 9-1.
Cisco Unity Express is associated with a JTAPI user on Unified CM. Although a single JTAPI user
can be associated with multiple instances of Cisco Unity Express in a system, Cisco recommends
associating each dedicated JTAPI user in Unified CM with a single Cisco Unity Express.
If Unified CM is upgraded from a previous version, the password of the JTAPI user automatically
gets reset on Unified CM. Therefore, after the upgrade, the administrator must make sure that the
JTAPI password is synchronized between Cisco Unity Express and Unified CM so that Cisco Unity
Express can register with Unified CM.
The CTI ports and CTI route points can be defined in specific locations. Cisco recommends using
locations-based call admission control between Unified CM and Cisco Unity Express. RSVP may
also be used.
Ensure proper Quality of Service (QoS) and bandwidth for signaling traffic that traverses the WAN
between Cisco Unity Express and Unified CM. Provision 20 kbps of bandwidth for CTI-QBE
signaling for each Cisco Unity Express site. See the chapter on Network Infrastructure, page 3-1,
for more details.
The CTI-QBE signaling packets from Unified CM to Cisco Unity Express are marked with a DSCP
value of AF31 (0x68). Unified CM uses TCP port 2748 for CTI-QBE signaling.
The Unified CM JTAPI library sets the proper IP Precedence bits in all outgoing QBE signaling
packets. As a result, all signaling between Cisco Unity Express and Unified CM will have the proper
QoS bits set.
19-45
Chapter 19
Note
Configure a SIP trunk for SRST and Unified CM for support of SIP phones (through JTAPI).
Cisco Unity Express supports G.729 SIP calls via a transcoder, with the ability added in Cisco IOS
Release 12.3(11)XW for RFC 2833 to pass through a transcoder.
Cisco Unity Express supports delayed media (no SDP in the INVITE message) for call setup in case
of a slow-start call from Unified CM.
Cisco Unity Express supports both blind and consultative transfer, but the default transfer mode is
consultative transfer (semi-attended) using REFER in SIP calls. Use the Cisco Unity Express
command line interface to explicitly change the transfer mode to consultative transfer using REFER
or blind transfer using BYE/ALSO. If REFER is not supported by the remote end, BYE/ALSO will
be used.
Cisco Unity Express supports outcall for voice message notifications. It also supports consultative
transfers. During both of these call setups, Cisco Unity Express can receive 3xx responses to the
INVITE. Cisco Unity Express processes only 301 (Moved Permanently) and 302 (Moved
Temporarily) responses to the INVITE. This requires the URL from the Contact header from the 3xx
response to be used to send a new INVITE. 305 (Use Proxy) responses are not supported.
For compatibility between Cisco Unified CM and Cisco Unity Express, refer to the Cisco Unity Express
Compatibility Matrix, available at
http://www.cisco.com/en/US/docs/voice_ip_comm/unity_exp/compatibility/cuecomp.htm.
For more information about Cisco Unity Express, refer to the product documentation available at
http://www.cisco.com/en/US/products/sw/voicesw/ps5520/index.html
19-46
Chapter 19
Note
This section does not discuss how to size a third-party voicemail system for ports and/or storage. For
this type of information, contact your voicemail vendor, who should be better able to discuss the
individual requirements of their own system, based upon specific traffic patterns.
Integration
Integration is defined as the physical connection between a voicemail system and its associated PBX or
call processing agent, and it also provides for the feature set between the two. There are many voicemail
vendors, and it is not uncommon for customers to want to continue to use an existing voicemail system
when deploying Cisco Unified CM.
Note
Cisco does not test or certify any third-party voicemail systems. Within the industry, it is generally
considered to be the responsibility of the voicemail vendor to test and/or certify their products with
various PBX systems. Cisco does, of course, test its interfaces to such equipment and will support these
interfaces regardless of which third-party voicemail system is connected.
Cisco Unified CM can be integrated with a third-party PBX by using QSIG, which also allows a
third-party PBX to connect to Unified CM through a Primary Rate Interface (PRI) T1/E1 trunk. Each
method has its own advantages and disadvantages, and the method you employ will largely depend on
how your voicemail system is integrated to your current PBX.
Today there are other potential methods of voicemail integration, such as H.323 or SIP. However, due to
the varying methods of vendor implementation, features supported, and other factors, these third-party
voicemail integrations will have to be evaluated on a per-customer basis. Customers are advised to
contact their Cisco Account Team and/or Cisco Partner to discuss these options further.
Messaging
Messaging is defined as the exchange of messages between voicemail systems, and there are several
open standards for this purpose.
The most common protocol deployed to allow messaging between dissimilar systems is Voice Profile for
Internet Mail (VPIM). VPIM has seen several updates to its specification, and although Version 2 is not
the latest, it still appears to be the most widely adopted. The messaging protocol prior to VPIM is Audio
Messaging Interchange Specification - Analog (AMIS-A), and it is fairly rare in its adoption due mainly
to its cumbersome user interface as well as the analog technology it employs and its lack of features.
19-47
Chapter 19
19-48
CH A P T E R
20
Cisco Unified Communications Manager IM and Presence Service provides native standards-based,
dual-protocol, enterprise instant messaging (IM) and network-based presence as part of Cisco Unified
Communications. This secure, scalable, and easy-to-manage service within Cisco Unified
Communications Manager offers users feature-rich communications capabilities both within and
external to the enterprise.
The Cisco Unified Communications Manager IM and Presence Service is one of the options for IM and
Presence, which enhances the value of a Cisco Unified Communications and Collaboration system. The
main presence component of the solution is the Cisco IM and Presence Service for all on-premises
deployment needs, which incorporates the Extensible Communications Platform (XCP) and supports
SIP/SIMPLE and Extensible Messaging and Presence Protocol (XMPP) for collecting information
regarding a user's availability status and communications capabilities. The user's availability status
indicates whether or not the user is actively using a particular communications device such as a phone.
The user's communications capabilities indicate the types of communications that user is capable of
using, such as video conferencing, web collaboration, instant messaging, or basic audio.
The IM and Presence Service is tightly integrated with Cisco and third-party compatible desktop and
mobile presence and instant messaging clients, which also include Cisco Jabber SDK. It enables the
clients to perform various functions such as instant messaging, presence, click-to-call, phone control,
voice, video, visual voicemail, and web collaboration. The IM and Presence Service offers the flexibility
of rich, open interfaces that enable implementation for IM and Cisco network-based presence, as well as
IM and presence federation for a wide variety of business applications.
The aggregated user information captured by the Cisco IM and Presence Service enables Cisco Jabber,
Cisco Unified Communications Manager applications, and third-party applications to increase user
productivity. These applications help connect colleagues more efficiently by determining the most
effective form of communication.
Note
The Cisco IM and Presence Service must be deployed with the equivalent version of Cisco Unified
Communications Manager (Unified CM) using Cisco provided VM configuration OPTION user
configuration templates on virtual servers on Cisco Unified Computing System (UCS). Cisco does not
support over-subscription of VM resources between virtual servers, and system resources should be
dedicated per virtual server being deployed.
This chapter explains the basic concepts of presence and instant messaging within the Cisco Unified
Communications System for on-premises, cloud, and hybrid options, and it provides guidelines for how
best to deploy the various components of the presence and instant messaging solution.
20-1
Chapter 20
Described in:
Revision Date
Split subclusters
Presence
Presence refers to the ability and willingness of a user to communicate across a set of devices. It involves
the following phases or activities:
Status combines the capabilities of what the device or user can do (voice, video, instant messaging, web
collaboration, and so forth) and the attributes showing the state of the device or user (available, busy, on
a call, and so forth). Presence status can be derived from automatic events such as client login and
telephone off-hook, or it can be derived from explicit notification events for changing status such as the
user selecting Do Not Disturb from a change-status pick list.
Terminology surrounding presence refers to a watcher, presence entity (presentity), and presence server.
The presence entity publishes its current status to the presence server by using a PUBLISH or
REGISTER message for SIP/SIMPLE clients, or by using an XML Presence Stanza for XMPP clients.
It can be a directory number (DN) or a SIP uniform resource identifier (URI) that resides within or
outside the communications cluster. A watcher (device or user) requests presence status about a presence
entity by sending a message to the presence server. The presence server responds to the watcher with a
message containing the current status of the presence entity.
20-2
Chapter 20
20-3
Chapter 20
Presence
Note
For voice-only, video-only, or IM-only deployments, the IM and Presence Service in Cisco
Collaboration System Release (CSR) 11.x has increased the user cluster capacity for Full UC Mode to
75,000 users.
The concept of a presence user appears throughout this chapter; therefore, keep in mind the meaning of
a user as defined for Cisco IM and Presence. By default an IM and Presence Service user is defined in a
Unified Communications deployment as user@default_domain (the basis for the Jabber Identifier, or
JID), where user is what is configured manually or in the Unified CM LDAP synchronization agreement
(sAMAccountName, email, employeenumber, telephonenumber, or UserPrincipalName) and
default_domain is the domain configured in the IM and Presence Service administration.
UserID@Default_Domain is the default IM addressing scheme when you install the IM and
Presence Service.
DirectoryURI IM addressing scheme supports multiple domains, alignment with the user's email
address, and alignment with Microsoft SIP URI.
The default setting of user@default_domain allows for only a single domain, whereas DirectoryURI
allows for greater flexibility in handling multiple domains and email addresses as the contact identifier.
A user can log into Jabber with their sAMAccountName attribute, while the Jabber ID is mapped to the
DirectoryURI field. The Fleixble JID structure makes it independent of UID for authentication.
Note
DirectoryURI is a global administrative setting on Unified CM. If DirectoryURI is selected for IM and
Presence Service addressing, all clients in the deployment must be able to handle and support the
DirectoryURI option.
While UserID can be mapped to the email address, that does not mean the IM URI equals the email
address. Instead it becomes <email-address>@Default_Domain. For example,
amckenzie@example.com@sales-example.com. The Active Directory (AD) mapping setting that you
choose is global to all users within that IM and Presence Service cluster. It is not possible to set different
mappings for individual users.
Unlike the UserID@Default_Domain IM addressing scheme, which is limited to a single IM domain, the
DirectoryURI IM addressing scheme supports multiple IM domains. Any domain specified in the
DirectoryURI is treated as hosted by the IM and Presence Service. The user's IM address is used to align
with their DirectoryURI, as configured on Cisco Unified Communications Manager.
20-4
Chapter 20
Cisco Unified CM
WebEx iOS
WebEx Android
WebEx Connect
WebEx Messenger
Jabber iOS
For information about SAML SSO, see the section on On-Premises Cisco IM and Presence Service
SAML SSO for Jabber, page 20-36, or the latest version of the SAML SSO Deployment Guide for Cisco
Unified Communications Applications available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
20-5
Chapter 20
The following Collaboration clients are supported by the Cisco Unified Communications System:
Cisco Jabber desktop clients for both Macintosh and Windows provide robust and feature-rich
collaboration capabilities, including standards-based IM and presence, audio and video, visual
voicemail, desktop sharing, desk phone control, Microsoft Office integration, and contact management.
Cisco Jabber desktop clients can be deployed to use on-premises services in which Cisco IM and
Presence and Cisco Unified Communications Manager provide client configuration, instant messaging
and presence, and user and device management. Cisco Jabber for Windows and Cisco Jabber for Mac
can also be deployed to use cloud-based services through integration with Cisco WebEx Messenger
service.
Deskphone Control Mode Using a Cisco IP Phone for audio (and video, if supported)
When a Jabber desktop client is in deskphone control mode, it does not register with Unified CM
using SIP, but instead it uses CTI/JTAPI to initiate, monitor, and terminate calls, monitor line state,
and provide call history, while controlling a Cisco IP Phone.
Cisco provides collaboration clients for the following mobile devices: Android, BlackBerry, and Apple
iOS devices such as iPhone and iPad. For more information on Cisco Jabber for mobile devices, see the
chapter on Mobile Collaboration, page 21-1.
20-6
Chapter 20
Cisco UC IntegrationTM for Microsoft Lync allows for integrated Cisco Unified Communications
services with Microsoft Lync and Microsoft Office Communications Server (OCS) R2, while delivering
a consistent user experience. The solution extends the presence and instant-messaging capabilities of
Microsoft Lync by providing access to a broad set of Cisco Unified Communications services, including
standards-based audio and video, unified messaging, web conferencing, deskphone control, and
telephony presence.
Third-Party XMPP Clients and Applications
Cisco IM and Presence, with support for SIP/SIMPLE and Extensible Messaging and Presence Protocol
(XMPP), provides support of third-party clients and applications to communicate presence and instant
messaging updates between multiple clients. Third-party XMPP clients allow for enhanced
interoperability across various desktop operating systems. In addition, web-based applications can
obtain presence updates, instant messaging, and roster updates using the HTTP interface with SOAP,
REST, or BOSH (based on the Cisco AJAX XMPP Library API). For additional information on the
third-party open interfaces, see the section on Third-Party Presence Server Integration, page 20-52.
(SSO or not)
20-7
Chapter 20
Note
Ambiguous Name Resolution (ANR) is a search algorithm associated with Lightweight Directory
Access Protocol (LDAP) clients, and it allows for objects to be bound without complex search filters.
ANR is useful when you are locating objects and attributes that may or may not be known by the client.
UDS usage considerations for on-premise deployments:
UDS is a set of HTTP-based services provided by Unified CM. The UDS contact source is one UDS
service, which provides contact and number resolution.
Using the UDS contact source to resolve contacts does put additional load on the server. When using
the UDS contacts service, the capacity of the subscriber node to register endpoints is reduced by
50%. This reduction is to allow processing capacity to handle contact resolution requests from
Jabber clients.
If the UDS contact source (on-net) is being used, then your design must comply with the cluster
sizing recommendations.
Jabber on-premises (off-net) always uses UDS as a contact source such as off Edge deployments.
Again, your design must comply with cluster sizing recommendations.
20-8
Chapter 20
LDAP Directory
You can configure a corporate LDAP directory to satisfy a number of different requirements, including
the following:
User provisioning You can provision users automatically from the LDAP directory into the Cisco
Unified Communications Manager database using directory integration. Cisco Unified CM
synchronizes with the LDAP directory content so that you avoid having to add, remove, or modify
user information manually each time a change occurs in the LDAP directory.
User authentication You can authenticate users using the LDAP directory credentials. Cisco IM
and Presence synchronizes all the user information from Cisco Unified Communications Manager
to provide authentication for client users.
User lookup You can enable LDAP directory lookups to allow Cisco clients or third-party XMPP
clients to search for contacts in the LDAP directory.
Display Name
Phone number
Mail ID
Group filters must be configured and validated by the administrator of the Cisco CallManager
Application.
The administrator should ensure that the User filters and Group filters are named appropriately and
assigned to the right filter fields meant for User and Group filters.
DirSync service does not validate the filter string format for syntax or logical accuracy; the
administrator must verify the accuracy of the filters.
For every filter string there should be one string added for ignoring security groups while performing
synchronization.
20-9
Chapter 20
When a contact cannot be found in the local Jabber desktop client cache or contact list, a search for
contacts can be made. The WebEx Messenger user can utilize a predictive search whereby the cache,
contact list, and local Outlook contact list are queried as the contact name is being entered. If no matches
are found, the search continues to query the corporate directory (WebEx Messenger database).
Your choice of deployment will depend primarily upon your product choice for IM and presence and the
requirement for additional services such as voice and video, voicemail, and deskphone control.
20-10
Chapter 20
Voice and
Video Media
IMAP
HTTP
RTP
REST
RTP
Call Control
SIP, TFTP,
CTI, HTTP
Contact
Search
IM and
Presence
REST
LDAP
SSL
AES
XMPP
SOAP
Audio, Video
SIP
RTP
TFTP
HTTP
CTI Mgr
Unied CM
Unity
Connecon
LDAP
Directory
IM and
Presence
348742
CCMCIP
The on-premises deployment model for Cisco Jabber for Windows relies on the following components:
Cisco Unified Communications Manager provides all user and device configuration capabilities.
Cisco Unified Communications Manager and Cisco conferencing devices provide audio and video
conferencing capabilities.
Microsoft Active Directory or another supported LDAP directory provides contact sources.
These components are the essential requirements to achieve a base deployment of Cisco Jabber clients.
After you set up and configure a base deployment, you can set up and configure additional deployment
options such as:
Video Provides capabilities to enable users to transmit and receive video calls.
Voicemail Provides voicemail capabilities that users can retrieve directly in the Cisco Jabber
client user interface or when users dial their voicemail number.
Desktop sharing Enables users to share their desktops via Binary Flow Control Protocol (BFCP).
Microsoft Office integration Provides user availability status and messaging capabilities directly
through the user interface of Microsoft Office applications such as Microsoft Outlook.
20-11
Chapter 20
Call Control
Contact
Search
IM and
Presence
Audio, Video,
Web Conferencing
HTTPS
XMPP
HTTPS
WebEx
User List
WebEx
Messenger
Service
WebEx
Meeng
Center
348743
Voice Mail
The cloud-based deployment model for Cisco Jabber for Windows relies on Cisco WebEx Messenger
service for the following services:
These services are the essential components required to achieve a base deployment of Cisco Jabber for
Windows. After you set up and configure a base deployment, you can set up and configure additional
deployment options such as:
Cisco WebEx Meeting Center Offers hosted collaboration features such as online meetings and
events.
Microsoft Office integration Provides user availability status and messaging capabilities directly
through the user interface of Microsoft Office applications such as Microsoft Outlook. This
integration is set up by default.
Calendar integration Calendar integration with WebEx Meeting Center, Outlook, and IBM Lotus
Notes is also supported.
20-12
Chapter 20
For information on WebEx Messenger service configuration for Jabber Clients, refer to the Cisco WebEx
Messenger Administrators Guide, available at
http://www.webex.com/webexconnect/orgadmin/help/index.htm
Figure 20-4
Voice and
Video Media
RTP
Call Control
SIP, TFTP,
CTI, HTTP
Contact
Search
IM and
Presence
HTTPS
XMPP
Audio, Video,
Web Conferencing
HTTPS
RTP
SIP
TFTP
HTTP
CTI Mgr
Unity
Connecon
Unied CM
WebEx
User List
WebEx
Messenger
Service
WebEx
Meeng
Center
348744
CCMCIP
20-13
Chapter 20
20-14
Chapter 20
Figure 20-5
10.1.1.2
10.1.1.4
IP
1000
SUBSCRIBE sip:2000@10.1.1.4
Event: presence
Contact: sip:unique_id@10.1.1.2
As long as the watcher
subscription state is still
active, a NOTIFY will be
sent any time there is a
status change of the
presence entity.
200 OK
190463
NOTIFY sip:unique_id@10.1.1.2
Event: presence
<xml message body containing status>
200 OK
If the presence entity resides outside the Unified CM cluster, Unified CM routes a SUBSCRIBE request
out the appropriate SIP trunk, based on the SUBSCRIBE calling search space, presence group, and SIP
route pattern. When Unified CM receives a SIP NOTIFY response on the trunk, indicating the presence
entity status, it responds to the SIP line-side presence request by sending a SIP NOTIFY message to the
presence watcher, indicating the current status of the presence entity. (See Figure 20-6.)
SIP Trunk SUBSCRIBE/NOTIFY Exchange
10.1.1.2
rtp.xyz.com
IP
1000
sjc.xyz.com
SUBSCRIBE sip:3000@sjc.xyz.com
Event: presence
Contact: sip:unique_id@10.1.1.2
200 OK
NOTIFY sip:unique_id@10.1.1.2
Event: presence
<xml message body containing status>
200 OK
SUBSCRIBE sip:3000@sjc.xyz.com
Event: presence
Contact: sip:unique_id@rtp.xyz.com
200 OK
NOTIFY sip:unique_id@rtp.xyz.com
Event: presence
<xml message body containing status>
200 OK
190464
Figure 20-6
SUBSCRIBE messages for any directory number or SIP URI residing outside the Unified CM cluster
are sent or received on a SIP trunk in Unified CM. The SIP trunk could be an interface to another
Unified CM or it could be an interface to the Cisco IM and Presence Service.
20-15
Chapter 20
Indicators for Speed Dial Presence on Cisco Unified IP Phones 7900 Series
State
Icon
LED
Idle
Unknown
190465
Busy
20-16
Chapter 20
Note
Cisco strongly recommends that you do not leave any calling search space defined as <None>. Leaving
a calling search space set to <None> can introduce presence status or dialing plan behavior that is
difficult to predict.
20-17
Chapter 20
When multiple presence groups are defined, the Inter-Presence Group Subscribe Policy service
parameter is used. If one group has a relationship to another group via the Use System Default setting
rather than being allowed or disallowed, this service parameter's value will take effect. If the
Inter-Presence Group Subscribe Policy service parameter is set to Disallowed, Unified CM will block
the request even if the SUBSCRIBE calling search space allows it. The Inter-Presence Group Subscribe
Policy service parameter applies only for presence status with call history lists and is not used for BLF
speed dials.
Presence groups can list all associated directory numbers, users, and devices if you enable dependency
records. Dependency records allow the administrator to find specific information about group-level
settings. However, use caution when enabling the Dependency Record Enterprise parameter because it
could lead to high CPU usage.
Select the appropriate model of Cisco Unified IP Phones that have the ability to display user phone
state presence status.
Note
Call history list presence capabilities are enabled on a global basis; however, user status can be
secured by using a presence policy.
BLF speed dials are administratively controlled and are not impacted by the presence policy
configuration.
Cisco Business Edition can be used in ways similar to Unified CM to configure and control user presence
capabilities. For more information, refer to the chapter on Call Processing, page 9-1.
20-18
Chapter 20
routing, policy, and federation management; the Rich Presence Service, which handles presence state
gathering, network-based rich presence composition, and presence-enabled routing functionality; and
support for ad-hoc group chat storage with persistent chat and message archiving handled to an external
database. If persistent chat is enabled, ad-hoc rooms are stored to the external PostgreSQL or Oracle
database for the duration of the ad-hoc chat. If persistent chat is disabled, ad-hoc chats are stored in
volatile memory for the duration of the chat.
Applications (either Cisco or third-party) can integrate presence and provide services that improve the
end user experience and efficiency. In addition, Cisco Jabber is a supported client of the Cisco IM and
Presence Service that also integrates instant messaging and presence status.
The Cisco IM and Presence Service also contains support for interoperability with Microsoft Lync
Server 2010 and 2013 and the Microsoft Lync client for any Cisco Unified IP Phone connected to a
Unified CM. The Microsoft Lync client interoperability includes click-to-dial functionality, phone
control capability through Remote Call Control (RCC), and presence status of Cisco Unified IP Phones.
Cisco Unified
Communications
Manager Publisher
1A
2A
3A
292382
Database
Sync
20-19
Chapter 20
Figure 20-9
1A
2A
3A
1B
2B
3B
292383
Cisco Unified
Communications
Manager Publisher
Figure 20-10
1A
2A
Cisco Unified
Communications
Manager Publisher
3A
Sub-cluster 3
1B
2B
4A
292384
Database
Sync
The on-premises Cisco IM and Presence Service utilizes and builds upon the database used by the
Unified CM publisher by sharing the user and device information.
Note
A single Unified CM cluster supports only a single IM and Presence Service cluster; therefore, a separate
IM and Presence Service cluster is required for each Unified CM cluster.
Intracluster traffic participates at a very low level between the Cisco IM and Presence Service and
Unified CM, and between the Cisco IM and Presence Service publisher and subscriber servers. Both
clusters share a common hosts file and have a strong trust relationship using IPTables. At the level of the
database and services, the clusters are separate and distinct; however, the configuration and
administration is primarily done on the Unified CM cluster, with limited configuration and
administration done on the IM and Presence Service cluster. There is currently no Transport Layer
Security (TLS) or IPSec utilization for intracluster traffic.
20-20
Chapter 20
The Cisco IM and Presence Service publisher communicates directly with the Unified CM publisher via
the AVVID XML Layer Application Program Interface (AXL API) using the Simple Object Access
Protocol (SOAP) interface. When first configured, the Cisco IM and Presence Service publisher
performs an initial synchronization of the entire Unified CM user and device database. All Cisco IM and
Presence Service users are configured in the Unified CM End User configuration. During the
synchronization, Cisco IM and Presence Service populates these users in its database from the
Unified CM database and does not provide end-user configuration from its administration interface.
After synchronization, users must be enabled for IM and Presence Service through the Cisco Unified
Communications Manager administrator interface before the Cisco IM and Presence Service can mange
them.
Note
Note
When Cisco IM and Presence Service is performing the initial database synchronization from
Unified CM, do not perform any administrative activities on Unified CM while the synchronization
agent is active.
The IM and Presence Service cluster supports a maximum of 75,000 users in Full UC mode, across
3 single IM and Presence nodes deployed with the 25,000-user VM configuration template but with no
high availability.
A high availability deployment for 75,000 users would require 3 IM and Presence subcluster pairs
deployed with the 25,000-user VM configuration template option.
20-21
Chapter 20
Note
The 25,000-user VM configuration should be used only for megacluster deployments, which must be
reviewed and approved by Cisco prior to deployment. If your deployment is not a megacluster, then use
a lower capacity VM configuration template, such as the 15,000-user or 5,000-user VM configuration
template.
Single site
Cisco IM and Presence Service is supported with all the Unified CM deployment models. However,
Cisco recommends locating the Cisco IM and Presence Service publisher in the same physical datacenter
as the Unified CM publisher due to the initial user database synchronization. All on-premises Cisco IM
and Presence servers should be physically located in the same datacenter within the Cisco IM and
Presence Service cluster, with the exception of geographic datacenter redundancy and clustering over
the WAN (for details, see Clustering Over the WAN, page 20-30).
For more information on Unified CM deployment models, see the chapter on Collaboration Deployment
Models, page 10-1.
Cisco IM and Presence Service deployment depends on high-availability requirements, the total number
of users, and the server being used. Detailed configuration and deployment steps can be found in the
Deployment Guide for Cisco IM and Presence, available at
http://www.cisco.com/en/US/products/ps6837/products_installation_and_configuration_guides_lis
t.html
A highly available Cisco IM and Presence Service cluster requires two servers per subcluster. This
allows for users to fail-over between the servers within the subcluster; however, the total number of users
supported and the time to failover vary based on which features are enabled, the average size of contact
lists, the rate of traffic placed on the servers, and the placement of the servers if deployed across a WAN.
Once a Cisco IM and Presence Service subcluster is configured for two servers, it always operates as
highly available if High Availability is configured in the Unified CM administration System > Presence
Redundancy Group. High availability can be deployed using an Active/Standby model or an
Active/Active model, and these modes are controlled by the Enterprise Parameter User Assignment
Mode for Presence Server. By default all users are balanced across all servers in the cluster, and Cisco
recommends leaving this parameter set to its default value.
Note
20-22
Chapter 20
Cisco IM and Presence Active/Active mode (Setting User Assignment Mode for Presence Server to
balanced), which is the default and recommended setting for load distribution, will automatically assign
users equally across all servers in the subclusters. Each server is synchronized and ready for a failover
if the other server in the subcluster fails. For example, in Figure 20-9 the first user would be assigned to
server 1A, the second user to server 2A, the third user to server 3A, the fourth user to server 1B, the fifth
user to server 2B, the sixth user to server 3B, and so forth. The users are assigned equally across all the
servers in the cluster.
Cisco IM and Presence Active/Active deployments with a balanced User Assignment Mode for
Presence Server allows for redundancy flexibility based on the features being used, the size of user
contact lists, and the traffic (user data profiles) being generated. A Cisco IM and Presence Active/Active
deployment with a fully redundant mode, regardless of features, requires the total number of supported
users to be reduced in half (for example, a deployment of 15,000 Users OVAs in a balanced
high-availability redundant configuration supports up to 15,000 users per subcluster). A Cisco IM and
Presence Active/Active deployment with a non-redundant mode requires a more detailed look at the
Cisco IM and Presence Service features being utilized, the average size of the users contact lists, as well
as the traffic being generated. For example, for a deployment with presence and instant messaging
enabled and calendaring and mobility integration disabled, with an average contact list of 30 users and
a user data profile of a few presence and instant message updates, it is possible to support more than
15,000 users per subcluster.
A Cisco IM and Presence Service cluster deployment that is not highly available allows each server in
the subcluster to support up to the maximum number of users for the server, and the total number of
supported users for all servers in the cluster can be up to the maximum number of users for the IM and
Presence Service cluster. Once a second server is added in a subcluster, the subcluster will still act as if
in a high-available deployment; however, if a server failure occurs, an attempt to fail-over might not
result in success if the online server reaches its capacity limit based on the Cisco IM and Presence
Service features enabled, the average user contact list size, and the traffic being generated by the users.
Deployment requirements:
Cisco UCS virtual machine for one 15,000-user VM configuration Full UC template
Deployment:
One single-server subcluster using User Assignment Mode for Presence Server = balanced
20-23
Chapter 20
Example 20-2 Two Unified CM Clusters with Cisco IM and Presence Service
Deployment requirements:
Cisco UCS virtual machines for two 15,000-user VM configuration Full UC templates
Deployment:
Two Cisco IM and Presence Service clusters (one per Cisco Unified Communications Manager
cluster), each with one server using User Assignment Mode for Presence Server = balanced
Example 20-3 Single Unified CM Cluster with Cisco IM and Presence Service
Deployment requirements:
Hardware:
Cisco UCS virtual machines for two 5,000-user VM configuration Full UC templates
Deployment:
One two-server subcluster using User Assignment Mode for Presence Server set to balanced, with
a PostgreSQL database instance for the cluster
Example 20-4 Single Cisco Business Edition Cluster with Cisco IM and Presence Service
Deployment requirements:
Hardware:
Cisco Business Edition 6000H using the 1,000-user VM configuration Full UC template
Deployment:
One two-server subcluster using User Assignment Mode for Presence Server set to balanced, with
a unique PostgreSQL database instance per server in the cluster for persistent chat functionality
20-24
Chapter 20
This is just an example of what megacluster deployment might consist of, if needed to deploy more than
40,000 users and/or devices.
Note
All deployments requiring more than 4 UCM subscriber pairs must be reviewed and approved by the
Cisco Megacluster team prior to deployment. For information about the megacluster approval process,
refer to http://wikicentral.cisco.com/display/CCM/Megacluster.
IM and Presence deployment requirements:
Hardware platform:
Cisco UCS B-Series with Unified CM 10,000-user VM configuration template and 25,000-user IM
and Presence VM configuration template
Example 20-6 Multiple Unified CM Clusters with Cisco IM and Presence Service on a Single UCS
B-Series Server
Deployment requirements:
Deployment
Five Cisco IM and Presence Service clusters in a single platform UCS B-Series, each one serving
one of the five Unified CM clusters with 15,000 users each.
20-25
Chapter 20
and not exceeding the maximum number of supported devices for cluster deployments. The maximum
number of IM-only users supported is 75,000. For a complete list of platform requirements for the Cisco
IM and Presence Service, as well as the maximum number of users supported per platform, refer to the
documentation available at
http://docwiki.cisco.com/wiki/Virtualization_for_Unified_CM_IM_and_Presence
The Cisco IM and Presence Service supports deployments using only VM configuration templates on
virtualized servers; physical server support is not available for the Cisco IM and Presence Service.
Cisco recommends using identical VM configurations for all IM and Presence nodes in a cluster.
However, mixing VM configurations of different capacities within a cluster is allowed as long as the VM
configuration used for the IM and Presence publisher node is of equal capacity or larger than the VM
configuration used on any of the subscriber nodes in the same cluster.
Similar guidelines apply to redundancy and high availability models. Within a cluster, the VM
configuration used on the IM and Presence standby and/or backup virtual subscriber node must not be
larger than the VM configuration of the IM and Presence publisher or its own respective active and/or
primary subscriber.
If you are mixing VM configurations within the same IM and Presence cluster, consider the possible
impact to performance and capacity due to the various VM configurations being used. The overall cluster
capacity might ultimately be dictated by the capacity of the smallest VM configuration within the cluster.
Note
IM and Presence integration with Unified CM does not imply that they are part of the same cluster but
rather two separate clusters.
Single-Cluster Deployment
Figure 20-11 represents the communication protocols between the Cisco IM and Presence Service, the
LDAP server, and Cisco Unified Communications Manager for basic functionality. For complete
information on Cisco IM and Presence Service administration and configuration, refer to the Cisco IM
and Presence installation, administration, and configuration guides, available at
http://www.cisco.com/en/US/products/ps6837/tsd_products_support_series_home.html
20-26
Chapter 20
Figure 20-11
LDAPv3
1
2
M
M
3
M
Cisco IM and
Presence Service
SIP/SIMPLE
CTI/QBE
LDAP
SOAP
348731
Cisco Unified CM
Figure 20-11 depicts the following interactions between Cisco IM and Presence Service components:
1.
The SIP connection between the Cisco IM and Presence Service and Unified CM handles all the
phone state presence information exchange.
Unified CM configuration requires the Cisco IM and Presence Service to be added as application
servers on Unified CM and also requires a SIP trunk pointing to the Cisco IM and Presence Service.
The address configured on the SIP trunk could be a Domain Name System (DNS) server (SRV) fully
qualified domain name (FQDN) that resolves to the Cisco IM and Presence Services, or it could
simply be an IP address of an individual Cisco IM and Presence Service. The Cisco IM and Presence
Service handles the configuration of the Cisco Unified Communications Manager application server
entry automatically through AXL/SOAP once the administrator adds a node in the system topology
page through Cisco IM and Presence Service administration.
If DNS is highly available within your network and DNS SRV is an option, configure the SIP trunk
on Unified CM with a DNS SRV FQDN of the Cisco IM and Presence Service publisher and
subscriber. Also configure the Presence Gateway on the Cisco IM and Presence Service with a DNS
SRV FQDN of the Unified CM subscribers, equally weighted. This configuration will allow for
presence messaging to be shared equally among all the servers used for presence information
exchange.
If DNS is not highly available or not a viable option within your network, use IP addressing. When
using an IP address, presence messaging traffic cannot be equally shared across multiple
Unified CM subscribers because it points to a single subscriber.
Unified CM provides the ability to further streamline communications and reduce bandwidth
utilization by means of the service parameter IMP PUBLISH Trunk, which allows for the PUBLISH
method (rather than SUBSCRIBE/NOTIFY) to be configured and used on the SIP trunk interface to
Cisco IM and Presence Service. Once the IMP PUBLISH Trunk service parameter has been enabled,
the users must be associated with a line appearance and not just a primary extension.
20-27
Chapter 20
2.
The Computer Telephony Integration Quick Buffer Encoding (CTI-QBE) connection between Cisco
IM and Presence Service and Unified CM is the protocol used by presence-enabled users in Cisco
IM and Presence Service to control their associated phones registered to Unified CM. This CTI
communication occurs when Cisco Jabber is using Desk Phone mode to do Click to Call or when
Microsoft Office Communicator is doing Click to Call through Microsoft Office Communications
Server 2007 or Microsoft Lync.
a. Unified CM configuration requires the user to be associated with a CTI Enabled Group, and the
primary extension assigned to that user must be enabled for CTI control (checkbox on the
Directory Number page). The CTI Manager Service must also be activated on each of the
Unified CM subscribers used for communication with the Cisco IM and Presence Service
publisher and subscriber. Integration with Microsoft Office Communications Server 2007 or
Microsoft Lync requires that you configure an Application User, with CTI Enabled Group and
Role, on Unified CM.
b. Cisco IM and Presence Service CTI configuration (CTI Server and Profile) for use with Cisco
Jabber is automatically created during the database synchronization with Unified CM. All Cisco
Jabber CTI communication occurs directly with Unified CM and not through the Cisco IM and
Presence Service.
Cisco IM and Presence Service CTI configuration (Desktop Control Gateway) for use with
Microsoft Office Communications Server 2007 or Microsoft Lync requires you to set the
Desktop Control Gateway address (Cisco Unified Communications Manager Address) and a
provider, which is the application user configured previously in Unified CM. Up to eight Cisco
Unified Communications Manager Addresses can be provisioned for increased scalability. Only
IP addresses can be used for Desktop Control Gateway configuration in the Cisco IM and
Presence Service. Administrators should ensure that any configuration and assignment of Cisco
Unified Communications Manager addresses is evenly distributed for the purpose of load
balancing.
3.
The AXL/SOAP interface handles the database synchronization from Unified CM to populate the
Cisco IM and Presence Service database.
a. No additional configuration is required on Unified CM.
b. Cisco IM and Presence Service security configuration requires you to set a user and password
The LDAP interface is used for LDAP authentication of users. For more information regarding
LDAP synchronization and authentication, see the chapter on Directory Integration and Identity
Management, page 16-1.
Unified CM is responsible for all user entries via manual configuration or synchronization directly
from LDAP, and Cisco IM and Presence Service then synchronizes all the user information from
Unified CM. If a user logs into the Cisco IM and Presence Service and LDAP authentication is
enabled on Unified CM, Cisco IM and Presence Service will go directly to LDAP for the user
authentication using the Bind operation.
When using Microsoft Active Directory, consider the choice of parameters carefully. Performance
of Cisco IM and Presence Service might be unacceptable when a large Active Directory
implementation exists and the configuration uses a Domain Controller. To improve the response
time of Active Directory, it might be necessary to promote the Domain Controller to a Global
Catalog and configure the LDAP port as 3268.
20-28
Chapter 20
Intercluster Deployment
The deployment topology in previous sections is for a single Cisco IM and Presence Service cluster
communicating with a single Unified CM cluster. Presence and instant messaging functionality is
limited by having communications within a single cluster only. Therefore, to extend presence and instant
messaging capability and functionality, these standalone clusters can be configured for peer
relationships for communication between clusters within the same domain. This functionality provides
the ability for users in one cluster to communicate and subscribe to the presence of users in a different
cluster within the same domain.
To create a fully meshed presence topology, each Cisco IM and Presence Service cluster requires a
separate peer relationship for each of the other Cisco IM and Presence Service clusters within the same
domain. The address configured in this intercluster peer could be a DNS FQDN that resolves to the
remote Cisco IM and Presence Service cluster servers, or it could also simply be the IP address of the
Cisco IM and Presence Service cluster servers.
The interface between each Cisco IM and Presence Service cluster is two-fold, an AXL/SOAP interface
and a signaling protocol interface (SIP or XMPP). The AXL/SOAP interface, between publisher-only
servers of an IM and Presence Service cluster, handles the synchronization of user information for home
cluster association, but it is not a full user synchronization. The signaling protocol interface (SIP or
XMPP) is a full mesh between all servers within the deployment. It handles the subscription and
notification traffic, and it rewrites the host portion of the URI before forwarding if the user is detected
to be on a remote Cisco IM and Presence Service cluster within the same domain.
When Cisco IM and Presence Service is deployed in an intercluster environment, a presence user profile
should be determined. The presence user profile helps determine the scale and performance of an
inter-cluster presence deployment and the number of users that can be supported. The presence user
profile helps establish the number of contacts (or buddies) a typical user has, as well as whether those
contacts are mostly local cluster users or users of remote clusters.
The traffic generated between Cisco IM and Presence Service clusters is directly proportional to the
characteristics of the presence user profile. For example, assume presence user profile A has 30 contacts
with 20% of the users on a local Cisco IM and Presence Service cluster and 80% of the users on a remote
Cisco IM and Presence Service cluster, while presence user profile B has 30 contacts with 50% of the
users on a local Cisco IM and Presence Service cluster and 50% of the users on a remote Cisco IM and
Presence Service cluster. In this case, presence user profile B will provide for slightly better network
performance and less bandwidth utilization due to a smaller amount of remote cluster traffic.
20-29
Chapter 20
A Cisco IM and Presence Service cluster can be deployed between two sites with a single subcluster
topology, where one server of the subcluster is in one geographic site and the other server of the
subcluster is in another site. This deployment must have a minimum bandwidth of 5 Mbps, a maximum
latency of 80 ms round-trip time (RTT), and TCP method event routing.
Intra-Cluster Bandwidth and Delay
Unified CM
IM and Presence
Cluster-1
Unified CM
IM and Presence
Cluster-2
348978
Figure 20-12
40ms
40ms
WAN
40ms
40ms
AD
AD
IM and Presence Pair
Cisco IM and Presence Service high availability allows for users on one node within a subcluster to
automatically fail-over to the other node within the subcluster. With a Cisco IM and Presence Service
subcluster containing a maximum of two nodes, remote failover is essentially between two sites, one site
for each node. A scalable highly available capacity for a Cisco IM and Presence Service cluster is up to
three subclusters; therefore, a scalable highly available remote failover topology would consist of the
following two sites:
This deployment must have a minimum bandwidth of 5 Mbps per subcluster, a maximum latency of
80 ms round-trip time (RTT), and TCP method event routing. Each new subcluster added to the
deployment requires an additional 5 Mbps of dedicated bandwidth to handle the database and state
replication.
In the split subcluster model, the subcluster pairs are split across the WAN and deployed at two separate
locations. The split subcluster model supports a maximum of 6 nodes deployed at 6 different locations,
assuming the bandwidth and delay requirements are followed, with a maximum round-trip time (RTT)
of 80 ms and 5 Mbps of bandwidth for each of the 6 nodes.
20-30
Chapter 20
The Unified CM publisher and IM and Presence publisher must reside on the same side of the WAN,
otherwise issues can arise with upgrades, especially with refresh (dual) upgrades.
All IM and Presence nodes must have a minimum bandwidth of 5 Mbps, and an additional 5 Mbps
for the cluster for database synchronization.
Every IM and Presence node must also reside within and not exceed 80 ms RTT delay.
All users must be distributed evenly across all split cluster nodes. For example, assuming a
subcluster supports 15,000 users, then each node of the split subcluster would support 7,500 on the
15k-user VM configuration template.
Local Failover
A Cisco IM and Presence Service cluster deployment between two sites may also contain a subcluster
topology per site (single node or dual node for high availability), where one subcluster is in one
geographic site and the other subcluster is in another geographic site. This topology allows for the users
to remain at their local site (highly available or not) without the requirement or need to fail-over to a
different site or location. This deployment must have a minimum bandwidth of 5 Mbps dedicated
bandwidth between each subcluster in the respective sites, a maximum latency of 80 ms round-trip time
(RTT), and TCP method event routing.
Bandwidth and latency considerations
With a Cisco IM and Presence Service cluster that has a topology of nodes split across a WAN, the
number of contacts within a user's client can impact the bandwidth needs and criteria for the deployment.
The traffic generated within and between Cisco IM and Presence Service clusters is directly proportional
to the characteristics of the presence user profile, and thus the amount of bandwidth required for
deployment. Cisco recommends 25% or fewer remote contacts for a client in environments where the
bandwidth is low (10 Mbps or less), and at all times the maximum round-trip latency must be 80 ms or
less.
Persistent Chat and Compliance logging considerations
When Cisco IM and Presence Service is enabled for persistent chat, message archiving, or compliance
logging and a sublcuster is split across a WAN, the external database server(s) must reside on the same
side of the WAN as the Cisco IM and Presence Services that use them. With the ability to support
multiple database instances on a single server and the requirement for an external database server to
reside on the same side of the WAN, if a Cisco IM and Presence Service cluster is split across a WAN,
then two external database servers will be required.
Federated Deployment
Cisco IM and Presence Service allows for business-to-business communications by enabling
inter-domain federation, which provides the ability to share presence and instant messaging
communications between different domains. Inter-domain federation requires explicit DNS domains to
be configured, as well as a security appliance (Cisco Adaptive Security Appliance) in the DMZ to
terminate federated connections with the enterprise.
Multi-Domain Support
The IM and Presence Service provides the ability to configure more than a single domain for federation.
The domains are automatically discovered by the system when using DirectoryURI, or the administrator
can add the domains manually. When a federated deployment involves multiple domains, then DNS SRV
20-31
Chapter 20
records need to be published for each email domain. Each DNS SRV record should resolve to an
identical set of results where XMPP federation is a list of all XMPP federation nodes and SIP federation
is the Public FQDN of the Routing IM & Presence node.
Federation with multiple email domains also requires regeneration of the security certificates cup-xmpp
(certificate presented to XMPP clients) and cup-xmpp-s2s (certificate presented to federated systems).
For both of these certificates, all the domains must be included as Subject Alt Name (SAN) entries. The
manual administrative configuration gives the administrator the option to pre-populate the domains so
that it is not necessary to regenerate the certificates every time a new domain automatically gets
discovered.
If all the federated domains are within the same trust boundary, where a deployment has all components
within a single datacenter, then the use of the Adaptive Security Appliance is not required. For
information on inter-domain federation, refer to the latest version of Interdomain Federation for IM and
Presence Service on Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-installation-and-configuration-guides-list.html
Figure 20-13 shows the basic inter-domain federation deployment between two different domains,
indicated by Domain A and Domain B. The Adaptive Security Appliance (ASA) in the DMZ is used as
a demarcation into the enterprise. XMPP traffic is passed through, whereas SIP traffic is inspected. All
federated incoming traffic is routed through the Cisco IM and Presence Service that is enabled as a
federation node, and is routed internally to the appropriate server in the cluster where the user resides.
For intercluster deployments, intercluster peers propagate the traffic to the appropriate home cluster
within the domain. All federated outgoing traffic is directed outward through any node in the IM and
Presence Service cluster that has XMPP federation enabled. Multiple nodes can be enabled as federation
nodes within large enterprise deployments, where each request is routed based on a round-robin
implementation of the data returned from the DNS SRV lookup.
20-32
Chapter 20
Figure 20-13
Unified CM
Unified CM
IM &
Presence
IM &
Presence
DMZ
DMZ
Unified
CM
Internet
IBM Lotus Sametime
Connect Clients
Cisco Jabber
Adaptive
Security
Appliance
Cisco Jabber
Domain A
SIP/SIMPLE
IM &
Presence
Domain B
348738
IM &
Presence
XMPP
Intercluster Peer
Cisco IM and Presence Service also provides configuration through SIP to allow for inter-domain
federation with Microsoft and AOL, as depicted in Figure 20-14. Cisco IM and Presence Service
inter-domain federation with Microsoft Lync Server provides basic presence (available, away, busy,
offline) and point-to-point instant messaging. Rich presence capability (On the Phone, In a Meeting, On
Vacation, and so forth), as well as advanced instant messaging features, are not supported. Cisco IM and
Presence Service inter-domain federation with AOL allows federation with users of AOL public
communities (aim.com, aol.com), with users of domains hosted by AOL, and with users of a far-end
enterprise that federates with AOL (that is, AOL is being used as a clearing house).
Note
A SIP federation (inter-domain to AOL) on Cisco IM and Presence Service must be configured for each
domain of the AOL network, which can consist of both hosted networks and public communities. Each
unique hosted domain must be configured; however, only a single aol.com public community needs to
be configured because the AOL network allows a user to be addressed as user@aol.com or
user@aim.com
20-33
Chapter 20
Figure 20-14
Unified CM
Unified CM
DMZ
DMZ
User Communities
aol.com, icq.com, aim.com
IM &
Presence
IM &
Presence
Unified
CM
Hosted Domains
user@hosteddomain.com
Adaptive
Security
Appliance
AOL SIP
Access
Gateway
Clearing House
user@federatedcompany.com
Internet
Microsoft
Lync Server
IM &
Presence
Cisco Jabber
Access
Proxy
and
Front-end
Server
Cisco Jabber
Microsoft Lync
Clients
Domain A
Domain B
348739
SIP/SIMPLE
XMPP
Intercluster Peer
Table 20-2 lists the state mappings between Cisco IM and Presence Service and Microsoft Lync Server.
Table 20-2
Cisco Status
Cisco Color
Status to AOL
Do not disturb
Red
Busy
Away
Busy
Yellow
Busy
Away
On the phone
Yellow
Busy
Away
In a meeting
Yellow
Busy
Away
Yellow
Away
Away
Available
Green
Available
Available
Unavailable/offline
Grey
Offline
Offline
20-34
Chapter 20
Note
Cisco IM and Presence Service must publish a DNS SRV record (SIP, XMPP, and each text conferencing
node) for the domain to allow for other domains to discover the Cisco IM and Presence Services through
DNS SRV. With a Microsoft Lync Server deployment, this is required because Cisco IM and Presence
Service is configured as a Public IM Provider on the Access Edge server. If the Cisco IM and Presence
Service cannot discover the Microsoft domain using DNS SRV, you must configure a static route on
Cisco IM and Presence Service for the external domain.
The Cisco IM and Presence Service SIP federation deployment can be configured with redundancy using
a load balancer between the Adaptive Security Appliance and the Cisco IM and Presence Service, or
redundancy can also be achieved with a redundant Adaptive Security Appliance configuration. For
XMPP federation, redundancy can be achieved using DNS SRV records.
For additional configuration and deployment considerations regarding a federated deployment, refer to
the latest version of Interdomain Federation for IM and Presence Service on Cisco Unified
Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-installation-and-configuration-guides-list.html
An intra-domain partitioned federated deployment, shown in Figure 20-15, is a secondary option that
allows for Cisco IM and Presence Service and Microsoft Lync Server to federate presence and instant
messaging within the same presence domain. The users are partitioned across both deployments, within
the single presence domain, and are licensed either on Cisco IM and Presence Service or on the
Microsoft Lync Server.
Note
The user cannot be licensed on both the Cisco and Microsoft platforms at the same time.
Figure 20-15
Domain A
M
M
Cisco
Jabber
IM and
Presence
Cisco Unified CM
Shared
Active
Directory
Microsoft Office
Communicator
or Lync
External XMPP
Federated
Enterprise
IM and Presence Users
SIP/SIMPLE
XMPP
Microsoft Office
Communications
Server 2007 R2 or
Lync Server
External SIP
Federated
Enterprise
IM and Presence and
Communications Server Users
348732
20-35
Chapter 20
The partitioned intra-domain federation between the Cisco and Microsoft platforms is based on the
SIP/SIMPLE protocol and allows for basic presence and instant messaging exchange, as supported with
the Cisco IM and Presence Service inter-domain federation support for Microsoft. Rich presence and
group chat functionality are not supported with the partitioned intra-domain presence federation.
Inter-domain federation and partitioned intra-domain federation can be supported simultaneously with
the following qualifications:
XMPP federation may be enabled on the Cisco IM and Presence Service deployment but is available
only to Cisco IM and Presence Service licensed users.
SIP federation may be enabled either on Cisco IM and Presence Service or on Microsoft Office
Communications Server 2007 R2 or Lync Server; however, for SIP Federation to be available to
both Cisco and Microsoft users, it must be enabled on Microsoft Office Communications
Server 2007 R2 or Lync Server.
The supported end point for SAML SSO deployment is Cisco Jabber.
The Security Assertion Markup Language Single Sign-On (SAML SSO) feature enhances the end user
experience by avoiding the need to log in multiple times to multiple applications within the collaboration
solution.
SAML SSO provides a secure mechanism to use credentials and relevant information of the end user
across multiple Unified Communications applications such as Unified CM, Cisco Unity Connection, IM
and Presence, Jabber clients, and so on.
For the SAML SSO feature to work correctly, ensure that the network architecture scales to meet the
number of users for each cluster, assuming that each user may have as many as five or more services that
require authentication and a minimum of two devices associated to each user. For a deployment across
multiple Unified Communications applications, all SAML requests must authenticate with the IdP for
Cisco Jabber clients to log in successfully.
Note
SSO is supported by Unified Communications services with SAML and OAuth only.
Cisco Jabber with SAML SSO does impact performance during logins, and the current maximum login
rate for 5,000 users is within half an hour. This is assuming that you have distributed devices and users
evenly across all nodes and that Cisco Jabber is in softphone mode.
Cisco Jabber is the only supported client and/or endpoint for IM and Presence deployments that supports
SAML SSO.
For sizing information and examples, see the section on SAML SSO Cisco Jabber Client, page 25-19, in
the chapter on Collaboration Solution Sizing Guidance, page 25-1.
20-36
Chapter 20
20-37
Chapter 20
Figure 20-16
1A
ODBC
External Database
5A
4A
6A
348734
Cisco Unified
Communications
Manager Publisher
3A
The IM and Presence Service supports point-to-point file transfer between XMPP clients. Table 20-3
lists the chat room limits for the IM and Presence Service
Table 20-3
Number of
Maximum
1,500 rooms
16,500 rooms
1,000 occupants
100 messages
15 messages
Text conferencing, sometimes referred to as multi-user chat, is defined as ad-hoc group chat and
persistent group chat and is supported as part of the XCP feature set. In addition, offline instant
messaging (storing instant messages for users who are currently offline) is also supported as part of the
XCP feature set. Cisco IM and Presence Service handles storage for each of these instant messaging
features in different locations. Offline instant messaging is stored locally in the Cisco IM and Presence
Service IDS database.
Ad-hoc group chat is stored locally in memory on the Cisco IM and Presence Service. Persistent group
chat requires an external database to store chat rooms and conversations. The external databases
supported are PostgreSQL (see http://www.postgresql.org/) and Oracle (see http://www.oracle.com).
Note
Cisco does not provide any database best practices or any data extraction tools. Those tasks and tools are
expected to be provided by a database administrator.
20-38
Chapter 20
Note
When Oracle is used as the external database, and tablespace information must be configured.
Note
There is no high availability solution for MFT or persistent chat because there is only one connection
from the IM and Presence node to each external third-party database.
Software
Note
If you are required to have an encrypted connection to the external database, you must use Oracle 11g.
No other supported database version allows encrypted connections.
You can install the database on either a Linux or a Windows operating system. See the PostgreSQL and
Oracle documentation for details on the supported operating systems and platform requirements.
IPv4 and IPv6 are supported, as is dual-stack mode.
Transfer Process
The flow for transferring a file to a single recipient involves the following steps:
1.
The sender's client uploads the file via HTTP, and the server responds with a URI for the file.
2.
3.
An entry is written to the external database log table to record the upload.
4.
The senders client sends an IM to the recipient; the IM includes the URI of the file.
5.
6.
7.
8.
The flow for transferring a file to a group chat or persistent chat room is similar, except that the sender
sends the IM to the chat room, and each chat room participant sends a separate request to download the
file.
20-39
Chapter 20
If you deploy any combination of the persistent group chat, message archiver, or managed file
transfer features on an IM and Presence Service node, you can assign the same physical external
database installation and file server to all of these features. However, you should consider the
potential IM traffic and file transfers (size and number) when you determine the server capacity.
The node public key is invalidated if the node's assignment is removed. If the node is reassigned at
a later date, a new node public key is automatically regenerated. The external file server will also
need to be reconfigured.
The Cisco XCP File Transfer Manager service must be active on each node where managed file
transfer is required.
750 Jabber users sending a single 500 KB file to 750 other Jabber users
Usage Level
Transfers per
Hour
CPU % Total
CPU % AFT
JM Table
Additional Size
Low usage
1,500
35%
25%
3,000
0.6 MB
1.5 MB
Medium usage
4,500
50%
40%
9,000
2.8 MB
4.5 MB
Maximum usage
12,000
65%+
55%
24,000
7.8 MB
12.0 MB
The File Transfer window has the following File Transfer Type controls:
Peer-to-Peer One-to-one file transfers are allowed, but files are not archived or stored on a server.
Group chat file transfer is not supported.
Managed File Transfer and Persistent Group Chat both require an external database instance per IM and
Presence node in an IM and Presence cluster.
Note
A node that has Managed File Transfer enabled should not be deployed in a cluster with a node that has
Peer-to-Peer enabled. The recommended migration path is to configure the Peer-to-Peer nodes as
Managed and Peer-to-Peer File Transfer nodes, and then change them to Managed File Transfer nodes.
20-40
Chapter 20
ODBC
External Database
3A
5A
6A
348735
Cisco Unified
Communications
Manager Publisher
2A
A blocking third-party compliance solution, which not only allows logging of messages but also applies
policy to message delivery and message content, is provided through a third-party compliance server
solution. Cisco IM and Presence Service third-party compliance can be deployed with multiple
compliance servers for each server in the cluster, multiple servers per compliance server, or some other
combination. All Cisco IM and Presence Service servers in the cluster are subject to compliance.
Figure 20-18 shows a deployment with a compliance server for each server in the IM and Presence
Service cluster; whereas Figure 20-19 shows a mapping of a single compliance server to multiple IM and
Presence Service servers, or multiple compliance servers to a single IM and Presence Service server. The
various deployment options allow for greater flexibility in compliance policy routing and cluster
deployment.
20-41
Chapter 20
Figure 20-18
Cisco Unified
Communications
Manager Publisher
2A
1A
3A
XDB
XDB
Figure 20-19
292392
Database Sync
1A
2
2A
3
3A
Sub-cluster 1
Sub-cluster 2
Sub-cluster 3
Load Balancer
Load Balancer
Load Balancer
XDB
XDB
XDB
Note
348736
Once the administrator selects the cluster-wide compliance and moves away from the 1:1 compliance
mapping, it is not possible to revert back to 1:1 compliance mapping. Therefore, it is important to map
out an appropriate third-party compliance server deployment when selecting cluster-wide compliance.
20-42
Chapter 20
Cluster-wide compliance allows for configuration of compliance profiles based on particular events,
while allowing those events to be prioritized and routed to the appropriate compliance servers when there
are overlapping events in the compliance profiles. Every compliance server must have a compliance
profile assigned, and multiple compliance servers can share the same compliance profile.
Collaboration Client Message Logging Storage Requirements
The message archiving and Persistent Chat functionality use an external database to store messages
offline. There are a number of factors to consider for the storage requirements of a deployment, such as
the customer topology, how the database is tuned, and how messaging is used within the organization.
The following calculations provide guidelines for these inputs to be used in estimating the raw database
storage requirements of a deployment for external database storage. These calculations presume
single-byte character data encoding; therefore, additional storage may be needed if internationalized
character sets are used.
Cisco IM and Presence Service supports both SIP and XMPP clients, and there are slightly different
amounts of overhead per message based on the protocol. The overhead per message for message
archiving could actually be larger or smaller depending on deployment, Jabber Identifier/UserID size,
client type, and thread ID; therefore, an average overhead amount is used. For SIP-based messages the
average overhead is 800 bytes and for XMPP messages the average overhead is 600 bytes.
The minimum storage requirements (in bytes) for message archiving per month for Cisco Jabber users
can be calculated as follows:
(Number of users) (Number of messages/hour) (Number of busy hours/month)
(600 + (3 Number of characters/message))
The message archiving requirements above must be doubled if Enable Outbound Message Logging is
enabled on Cisco IM and Presence Service compliance configuration.
The minimum storage requirements (in bytes) for persistent chat per month for Cisco Jabber users can
be calculated as follows:
(Number of users) (Number of Persistent Chat messages/hour) (Number of busy hours/month)
(700 + (3 Number of characters/message))
Note
Persistent Chat is supported only with XMPP clients and uses an average overhead of 700 bytes.
These message archive and Persistent Chat numbers are the minimum storage requirements based on an
average over time; therefore, a buffer multiplier of 1.5 (150%) should be used to account for very large
UserIDs, larger than expected instant message lengths, and other factors that tend to increase the storage
requirements. Table 20-5 lists some examples of storage requirements for Cisco Collaboration Clients.
Table 20-5
Profile
Number of
Users
Number of
Messages per
Hours
Number of
Message Archive
Busy Hours per Average Size of Storage
Month
Requirement
Message
Persistent Chat
Storage
Requirement
Light
1,500
10
200
100
2.7 GB
3.0 GB
Medium
2,500
15
200
250
10.2 GB
10.9 GB
High
2,500
25
200
500
26.3 GB
27.5 GB
20-43
Chapter 20
Outlook Web Access Communication Between Cisco IM and Presence Service and Microsoft Exchange
Cisco IM and
Presence Calendar
Microsoft Exchange
Notification Port
Cisco IM and
Presence Calendar
Exchange
HTTP Search
SEARCH
SUBSCRIBE
HTTP 200 OK
HTTP Notify
NOTIFY
200 OK
HTTP Poll
POLL
UNSUBSCRIBE
292393
HTTP Unsubscribe
20-44
Chapter 20
Note
Cisco IM and Presence Service can be deployed with a single Microsoft Exchange Server or with
multiple Microsoft Exchange Servers, in a single forest only. Microsoft Exchange deployment allows for
clustering of multiple Exchange servers; therefore, Cisco IM and Presence Service will honor the
REDIRECT message to the exchange server that is hosting the user for which Cisco IM and Presence
Service is requesting status.
Cisco IM and Presence Service, as well as Cisco Unified Communications Manager, must have the
appropriate locales installed for the users to select their locale.
Cisco IM and Presence Service supports all the standard Unified Communications locales for
calendar integration.
Users must be configured for the locale that is desired, either through the end user pages or
administratively through the Bulk Administration Tool.
Cisco IM and Presence Service sends the appropriate locale folder with the initial query. Queries are
redirected, if required, through the response of the initial Front-End or Client Access Microsoft
Exchange server.
20-45
Chapter 20
Cisco IM and Presence Exchange Web Services calendar integration supports both a polling of calendar
information as shown in Figure 20-21 as well as a subscription/notification for calendar information as
shown in Figure 20-22. Various configuration parameters control the rate of polling intervals, the
frequency of subscriptions, and the fault tolerance of timers. For additional configuration details, refer
to the Integration Note for Configuring Cisco IM and Presence with Microsoft Exchange, available at
http://www.cisco.com/en/US/products/ps6837/products_installation_and_configuration_guides_lis
t.html
Figure 20-21
Exchange Web Services Polling with Cisco IM and Presence Service Calendar
Microsoft
Exchange
Client Access
Cisco IM and
Presence Calendar
HTTP Post -EWS CalendarInfo FindItem
POST
POST
POST
POST
292395
20-46
Chapter 20
Figure 20-22
Microsoft
Exchange
Client Access
Cisco IM and
Presence Calendar
HTTP Subscribe -EWS CalendarInfo
SUBSCRIBE
NOTIFY
User adds/modifies/deletes a calendar event
NOTIFY
HTTP Post SendNotification
UNSUBSCRIBE
292396
Exchange Web Services Auto Discover is also supported by Cisco IM and Presence Service if a service
connection point (SCP) Active Directory object has been created for each server where the Client Access
Server (CAS) role is installed. The calendar gateway is configured with Auto Discover using the domain
and optionally the site instead of a host and port. Cisco IM and Presence Service uses the auto-discover
algorithm to determine which Exchange Web Services URL to use in contacting the correct Client
Access Server Exchange Server.
20-47
Chapter 20
The real-time eventing model uses an application user on Cisco IM and Presence Service to establish an
administrative session, which allows for end users to log in with that session key. Once the end user has
logged in, the user registers and subscribes for presence updates using Representational State Transfer
(REST). Figure 20-23 highlights the Third-Party Open API real-time eventing model interaction with
Cisco IM and Presence Service.
Figure 20-23
notify(subid)
4
271402
getPresencesubscription (subid)
The call flow in Figure 20-23 illustrates the following sequence of events:
1.
The application initiates a SOAP login request to Cisco IM and Presence Service via the super-user
application user (APIUser), and Cisco IM and Presence Service returns a session key. The
application can then log in the end-user with this session key (essentially, the end-user logs in via
the application).
2.
The end user registers the endpoint using the application-user session key.
3.
The application initiates a subscribe request (using the session key) on behalf of the end user to
retrieve user information, contact list, and presence rules.
20-48
Chapter 20
4.
5.
Polling Model
The polling model uses an application user on Cisco IM and Presence Service to establish an
administrative session, which allows for end users to log in with that session key. Once the end user has
logged in, the application requests presence updates periodically, also using Representational State
Transfer (REST). Figure 20-24 highlights the Third-Party Open API polling model interaction with
Cisco IM and Presence Service.
Figure 20-24
271403
The call flow in Figure 20-24 illustrates the following sequence of events:
1.
The application initiates a SOAP login request to Cisco IM and Presence Service via the super-user
application user (APIUser), and Cisco IM and Presence Service returns a session key. The
application can then log in the end-user with this session key (essentially, the end-user logs in via
the application).
2.
The application requests presence state and bypasses the eventing model.
3.
The application requests presence state and bypasses the eventing model.
Note
Both Basic presence and Rich presence can be retrieved; however, the polling model puts an
additional load on the presence server.
20-49
Chapter 20
The XCP architecture allows for two additional open interfaces for presence, instant messaging, and
roster management: a client XMPP interface and a Cisco AJAX XMPP Library interface. The client
XMPP functionality enables third-party XMPP clients to integrate presence, instant messaging, and
roster management, and it is a complementary interface to the SIP/SIMPLE interface on Cisco IM and
Presence Service. The client XMPP interface is treated as a normal XMPP client within Cisco IM and
Presence Service; therefore, sizing of the interface should be treated as a normal XMPP client.
The Cisco AJAX XMPP Library API provides a Web 2.0 style of interface to integrate XCP features into
web applications and widgets, and it is made directly available from Cisco IM and Presence Service. The
Cisco AJAX XMPP Library API is exclusively a client-side JavaScript library that communicates to the
Bidirectional-streams Over Synchronous HTTP (BOSH) interface, which is essentially an XMPP over
HTTP interface that allows the server to push data to a web browser through a long-polling technique.
Observe the following requirements when integrating either model of the Third-Party Open API with
Cisco IM and Presence Service:
Certificates are required for the presence interface (sipproxy.der) and the configuration interface
(tomcat_cert.der).
No more than 1000 Third-Party Open API users can be integrated per Cisco IM and Presence Service
deployment.
To improve performance, balance the Third-Party Open API users across all servers in the Cisco IM
and Presence Service cluster.
You can obtain additional information and support for use of the Cisco IM and Presence Service
Third-Party Open API through Cisco Developer Services, available at:
http://developer.cisco.com/web/cupapi
Information and assistance for developers is also available from the Cisco Developer Community, which
is accessible through valid Cisco login authentication at:
http://developer.cisco.com/
If LDAP integration is possible, LDAP synchronization with Unified CM should be used to pull all
user information (number, ID, and so forth) from a single source. However, if the deployment
includes both an LDAP server and Unified CM that does not have LDAP synchronization enabled,
then the administrator should ensure consistent configuration across Unified CM and LDAP when
configuring user directory number associations.
Cisco IM and Presence Service marks Layer 3 IP packets via Differentiated Services Code Point
(DSCP). Cisco IM and Presence Service marks all call signaling traffic based on the Differential
Service Value service parameter under SIP Proxy, which defaults to a value of DSCP 24 (PHB CS3).
Presence Policy for Cisco IM and Presence Service is controlled strictly by a defined set of rules
created by the user.
Use the service parameter IMP PUBLISH Trunk to streamline SIP communication traffic with the
Cisco IM and Presence Service.
Associate presence users in Unified CM with a line appearance, rather than just a primary extension,
to allow for increased granularity of device and user presence status. When using the service
parameter IMP PUBLISH Trunk, you must associate presence users in Unified CM with a line
appearance.
20-50
Chapter 20
A Presence User Profile (the user activity and contact list contacts and size) must be taken into
consideration for determining the server hardware and cluster topology characteristics.
Use the User Assignment Mode for Presence Server enterprise parameter default of balanced for
best overall cluster performance.
Cisco IM and Presence Service requires an external database instance for each server in the cluster
for persistent chat, and one database instance per cluster for message archiving. Native compliance
supports mapping of all or a sub-set of servers in an IM and Presence Service cluster to one external
compliance database. Third-party compliance supports flexible deployments where there can be
multiple compliance servers per IM and Presence Service server, multiple IM and Presence Service
servers per compliance server, or some combination. The external databases supported are
PostgreSQL and Oracle, and all IM and Presence Service servers in the cluster are subject to
compliance.
Cisco IM and Presence Service supports a total of 75,000 users per cluster for full Unified
Communications mode. The sizing for users must take into account the number of SIP/SIMPLE
users and the number of XMPP users. XMPP users have slightly better performance because
SIP/SIMPLE users employ the IM Gateway functionality into the XCP architecture.
All eXtensible Communications Platform (XCP) communications and logging are stored in GMT
and not localized to the installed location.
For ease of user migration and contact list migration, Cisco IM and Presence Bulk Administration
Tool supports bulk contact list importation using a comma-separated value (csv) file as input for this
bulk importation.
For a complete listing of ports used by Cisco IM and Presence Service, refer to Port Usage Information
for Cisco IM and Presence, available at
http://www.cisco.com/en/US/products/ps6837/products_device_support_tables_list.html
20-51
Chapter 20
When using mobile and remote access, the following Jabber features are not supported:
File transfer
The following guidelines apply when integrating the Cisco IM and Presence Service and Microsoft Lync:
Communications between Cisco IM and Presence Service and Microsoft Lync uses the SIP/SIMPLE
interface. However, Microsoft Lync tunnels Computer-Supported Telecommunications Applications
(CSTA) traffic over SIP. Therefore, the CTI gateway on the Cisco IM and Presence Service must be
configured to handle the CSTA-to-CTI conversion for Click to Call phone control.
Cisco IM and Presence Service deployment with Microsoft Lync for Remote Call Control, should
consist of a single subcluster pair of servers that make up the Cisco IM and Presence Service cluster.
20-52
Chapter 20
The following table lists the number of users supported per platform. The user count is based solely
on the Unified CM platform equivalent, regardless of the IM and Presence Service platform.
1,000 User
1,000
4,000
2,500 User
2,500
10,000
7,500 User
7,500
30,000
10,000 User
10,000
40,000
You must configure the same end-user ID in LDAP, Unified CM, and Microsoft Lync. This practice
avoids any conflicts between Microsoft Lync authentication with Active Directory (AD) and the
end-user configuration on Unified CM, as well as conflicts with user phone control on Unified CM.
For Active Directory, Cisco recommends that the user properties of General, Account, and
Communications all have the same ID. To ensure the Cisco IM and Presence Service users are
consistent, LDAP Synchronization and Authentication should be enabled with Unified CM.
You must configure Microsoft Lync Host Authentication to contain the Cisco IM and Presence
Service publisher and subscriber.
You can configure routing of the SIP messages to Cisco IM and Presence Service by means of Static
Routes in the Microsoft Lync properties.
You must configure an incoming and outgoing access control list (ACL) on the Cisco IM and
Presence Service to allow for communications with Microsoft Lync.
You must enable each user for use of Microsoft Lync in the Cisco IM and Presence Service
configuration, in addition to enabling each user for presence in Unified CM.
Take into account bandwidth considerations for Microsoft Lync login due to the exchange of
configuration information between Microsoft Lync and the Microsoft Communications Server, and
due to initial communication with the Cisco IM and Presence Service CTI gateway.
To address the issue of a reverse look-up of a directory number that corresponds to a user, use the
guidelines documented in the Release Notes for Cisco IM and Presence, available at
http://www.cisco.com/en/US/products/ps6837/prod_release_notes_list.html
20-53
Chapter 20
A secure connection (SSL and AES) to the Cisco WebEx Messenger XMPP cloud platform for
presence, instant messaging, VoIP, PC-to-PC video, media transfer (screen capture and file transfer),
and desktop sharing
XMPP federation with other WebEx Messenger organizations and third-party XMPP clients and
XMPP instant messaging (IM) networks
Cisco Unified Communications integration for call control, voice messaging, and call history
20-54
Chapter 20
Figure 20-25
Developer APIs
Inter Domain
Meeng
Center
Cisco
webex
Messenger Service
Meeng Center
Escalaon
Integrated:
VoIP
Video
File Transfer
Desk Sharing
Oce
&
SharePoint
API
API
Notes
Calendar
SSL Signalling
Desk Phone
Control
CSF
SIP
SCCP/SIP
MeengPlace
Conferencing
SIP/CTI/TFTP/CCMIP
CUCM
MAC
Web
Enterprise Components
Unity
344755
IMAP
Centralized Management
Cisco WebEx Messenger service provides a web-based administrative tool to manage the solution across
the organization. Cisco WebEx Messenger service users are configured and managed through the Cisco
WebEx Administration Tool, which enables administrators to set up basic security and policy controls
for features and services. These policies can be applied enterprise-wide, by group, or individually. There
are various methods to provision the user database that are further described in the Cisco WebEx
administrator's guide available at
http://www.webex.com/webexconnect/orgadmin/help/index.htm
20-55
Chapter 20
Single Sign On
Single Sign On (SSO) enables companies to use their on-premises SSO system, including Security
Assertion Markup Language (SAML) support, to simplify the management of Cisco WebEx Messenger
or IM and Presence Service by allowing users to securely log into any of the Unified Communications
applications in the solution using their corporate login credentials. The user's login credentials are not
sent to Cisco, thus protecting the user's corporate login information. Figure 20-26 shows the credential
handshake that occurs on user login to Cisco WebEx Messenger as well as Unified CM.
Note
If Cisco Jabber is deployed with Cisco WebEx Meeting Server, Cisco Unified CM and WebEx Meeting
Server must be in the same domain.
Figure 20-26
IdP
acme.com
(X.509)
3) Challenge
IDMS
During provisioning
upload enterprise
cercate
4) Send Credenals
5)Generate Asseron
SAML
Asseron
7) Post Asseron
8) Validate User
SP
SAML Auth
Service
1) FederatedSSO.do?org=acme.com
209634
6) Sign Asseron
9) Logon Response
A user account can be configured to be created automatically the first time a user logs into Cisco IM
client. Users are prevented from accessing the Cisco WebEx Messenger service if their corporate login
account is deactivated.
For more information on Single Sign On with WebEx Messenger service, refer to the documentation
available at
http://developer.cisco.com/web/webex-developer/sso-reference
20-56
Chapter 20
Security
The Cisco WebEx security model consists of functional layers of security. Figure 20-27 illustrates the
separate but interrelated elements that compose each layer.
Figure 20-27
Policy
Management
Access Control:
Set policy for individuals, groups,
organizations
Physical
Security
Third-Party Audits
Authentication
SSAE 16
Type II
348737
Encryption
The bottom layer represents the physical security in the Cisco WebEx data centers. All employees go
through an extensive background check and must provide dual-factor authentication to enter the
datacenter.
The next level is policy management, where the WebEx Messenger organization administrator can set
and manage access control levels by setting different policies for individual users, groups, or the entire
Cisco WebEx Messenger organization. White-list policies, specific to external users or domains, can be
created to allow instant messaging exchanges. The Cisco WebEx Messenger organizational model also
allows for the creation of specific roles and groups across the entire user base, which allows the
administrator to assign certain privileges to roles or groups as well as to set policies, including access
control, for the entire organization.
Access to the Cisco WebEx Messenger service is controlled at the authentication layer. Every user has
a unique login and password. Passwords are never stored or sent over email in clear text. Passwords can
be changed only by the end-users themselves. The administrator can choose to reset a password, forcing
the end-user to change his or her password upon the next login. Alternatively, an administrator may
choose to use the Single Sign On (SSO) integration between Cisco WebEx Messenger service and the
company's directory to simplify end-user access management. The Single Sign On integration is
achieved through the use of an Identity Management System (IDMS).
The encryption layer ensures that all instant messaging communications between Cisco WebEx
Messenger users is encrypted. All instant messaging communication between Cisco WebEx Messenger
users and the server in the Messenger Collaboration cloud is encrypted by default using SSL encryption.
An additional level of security is available whereby IM communication can be encrypted end-to-end
using 256-bit AES level encryption.
20-57
Chapter 20
The Cisco WebEx Messenger platform uses third-party audits such as the SSAE 16 Type II audit to
provide customers with an independent semi-annual security report. This report can be reviewed by any
customer upon request with the Cisco Security organization. For additional Cisco WebEx Messenger
service security, refer to the Cisco WebEx Connect Security White Paper, available at
http://www.cisco.com/en/US/products/ps10528/prod_white_papers_list.html
20-58
Chapter 20
High Availability
With the use of the multi-tenant Software-as-a-Service architecture, if any individual server in a group
fails for any reason, requests can be rerouted to another available server in the Cisco WebEx Messenger
Platform.
The Cisco WebEx Network Operations Team provides 24x7 active monitoring of the Cisco WebEx
Collaboration Cloud from the Cisco WebEx Network Operations Center (NOC). For a comprehensive
overview of the Cisco WebEx technology, refer to the information at
http://www.cisco.com/en/US/solutions/ns1007/collaboration_cloud.html
Global Site Service Is responsible for monitoring and switching traffic at the network level.
Database Replication Ensures that the data transactions occurring on the primary site are
transferred to the backup site.
File Replication Ensures that any file changes are maintained in synchronization between the
primary and the backup site.
20-59
Chapter 20
Organization policies cannot be enforced on third-party XMPP clients, and features such as end-to-end
encryption, desktop share, video calls, PC-to-PC calls, and teleconferences are not supported with
third-party clients. To allow non-WebEx Messenger service XMPP IM clients to authenticate to your
WebEx Messenger service domain(s), DNS SRV records must be updated. The specific DNS SRV entry
can be found in Cisco WebEx administration, under Configuration and IM Federation.
The use of non-Messenger service XMPP clients in Cisco WebEx administration, under Configuration
and XMPP IM Clients, must be explicitly allowed.
For additional information on enabling third-party XMPP clients to connect to the WebEx Messenger
platform, refer to the Cisco WebEx administrator's guide available at
http://www.webex.com/webexconnect/orgadmin/help/index.htm
Inter-Domain Federation
SRV Record
_xmpp-server._tcp.<acme.com>
Port 5269
acme.com
SRV Record
_xmpp-server._tcp.<partner.com>
Port 5269
partner.com
Public DNS
Public DNS
P
Public
DNS
XMPP Gateway
X
XMPP Gateway
Public DNS
Public DNS
customer.com
SRV Record
_xmpp-server._tcp.<customer.com>
Port 5269
209637
acme.com.webex.com
Currently the WebEx Messenger service does not interoperate with Yahoo! Messenger and Windows
Live Messenger, but it can federate with AIM through a federation gateway.
20-60
Chapter 20
20-61
Chapter 20
20-62
CH A P T E R
21
Mobile Collaboration
Revised: June 15, 2015
Mobile collaboration solutions and applications provide the ability to deliver features and functionality
of the enterprise IP communications environment to mobile workers wherever they might be. With
mobile collaboration solutions, mobile users can handle business calls on a multitude of devices and
access enterprise applications whether moving around the office building, between office buildings, or
between geographic locations outside the enterprise. Mobile collaboration solutions provide mobile
workers with persistent reachability and improved productivity as they move between, and work at, a
variety of locations.
Mobile collaboration solutions can be divide into two main categories:
Mobility within the enterprise is limited to utilization within the network boundaries of the enterprise,
whether those boundaries span only a single physical building, multiple physical buildings in close
proximity or separated by long distances, or even home offices where network infrastructure is still
controlled and managed by the enterprise when it is extended to the home office.
On the other hand, mobility beyond the enterprise involves a bridging of the enterprise infrastructure to
the Internet or mobile provider infrastructures and finds users leveraging public and private networks for
connectivity to enterprise services. In some cases the lines between these two types of mobility are
somewhat blurred, especially in scenarios where mobile devices are connecting back to the enterprise
for collaboration services over the Internet or mobile data and mobile voice networks.
Mobility within the enterprise can be divided into three main areas based on feature sets and solutions:
21-1
Chapter 21
Mobile Collaboration
Multisite mobility
With this type of mobility, users move within the enterprise from one physical location to another,
and this movement typically involves crossing IP address spaces as well as PSTN egress/ingress
boundaries. This type of mobility involves the same types of operations and features as with campus
mobility (physical hardware moves, WLAN roaming, and Cisco Extension Mobility) but replicated
at each site within the enterprise. In addition, the Device Mobility feature can be leveraged to ensure
that, as user's move devices between sites, phone calls are routed through the local site egress
gateway, media codecs are negotiated appropriately, and call admission control mechanisms are
aware of the device's location.
Mobility beyond the enterprise can be divided into two high-level Cisco solution sets:
The various applications and features discussed in this chapter apply to all Cisco Unified
Communications deployment models unless otherwise noted.
This chapter begins with a discussion of mobility features and solutions available within the enterprise
infrastructure. It includes an examination of functionality and design considerations for campus or
single-site deployments, multisite deployments, and even remote site deployments. This comprehensive
set of solutions provides many benefits for mobile workers within the enterprise, including
enterprise-class communications and improved productivity regardless of physical location. This
discussion of mobility within the enterprise paves the way for examination of mobility solutions beyond
the enterprise that leverage the mobile provider and Internet provider infrastructure and capabilities.
21-2
Chapter 21
Mobile Collaboration
Whats New in This Chapter
These solutions enable a bridging of the enterprise network infrastructure and mobile functionality to
the provider network infrastructure in order to leverage advanced mobile features and communication
flows that can be built on the solid enterprise mobility infrastructure.
This chapter provides a comprehensive examination of mobility architectures, functionality, and design
and deployment implications for enterprise collaboration mobility solutions. The analysis and
discussions contained within this chapter are organized at a high level as follows:
Described in
Revision Date
Cisco Spark client and Cisco Collaboration Cloud Cisco Spark, page 21-93
services
21-3
Chapter 21
Mobile Collaboration
Physical wired
phone moves
Wireless IP
device roaming
Extension
PSTN
348813
Mobility
21-4
Chapter 21
Mobile Collaboration
Mobility Within the Enterprise
21-5
Chapter 21
Mobile Collaboration
making and receiving voice calls. Therefore, when deploying wireless phones and clients, it is
imperative to conduct a WLAN radio frequency (RF) site survey before, during, and after the
deployment to determine appropriate cell boundaries, configuration and feature settings, capacity, and
redundancy to ensure a successful voice and video over WLAN (VVoWLAN) deployment.
APs can be deployed autonomously within the network so that each AP is configured, managed, and
operated independently from all other APs, or they can be deployed in a managed mode in which all APs
are configured, managed, and controlled by a WLAN controller. In the latter mode, the WLAN controller
is responsible for managing the APs as well as handling AP configuration and inter-AP roaming. In
either case, to ensure successful VVoWLAN deployment, APs should be deployed using the following
general guidelines:
Channel 1
Channel 140
Channel 11
Channel 36
Channel 1
Channel 52
2.4 GHz
GH
Hz ch
channel
hannel cells
5 GHz channel cells
Ch
Channel
hanne
el 6
Channel 120
Channel 6
Channel 108
Channel 11
1
Channel 64
Channel 11
Channel 56
Channel 1
Channel 100
348679
Figure 21-2
As shown in Figure 21-2, non-adjacent WLAN AP channel cells should overlap by a minimum of
20%. This overlap ensures that a wireless device can successfully roam from one AP to the next as
the device moves around within the campus location while still maintaining voice and data network
connectivity. A device that successfully roams between two APs is able to maintain an active voice
call without any noticeable change in the voice quality or path.
21-6
Chapter 21
Mobile Collaboration
Mobility Within the Enterprise
As shown in Figure 21-3, WLAN AP channel cells should be deployed with cell power-level
boundaries (or channel cell radius) of -67 decibels per milliwatt (dBm). Additionally, the
same-channel cell boundary separation should be approximately 19 dBm.
A cell radius of approximately -67 dBm (or less) minimizes packet loss, which can be problematic
for real-time voice and video traffic. A same-channel cell separation of 19 dBm is critical to ensure
that APs or clients do not cause co-channel interference to other devices associated to the same
channel, which would likely result in poor voice quality. The cell radius guideline of -67 dBm
applies for both 2.4 GHz (802.11b/g/n) and 5 GHz (802.11a/n/ac) deployments.
Figure 21-3
The RADIUS
of the cell
should be:
67 dBm
The separation of
same channel cells
should be:
19 dBm
86 dBm
253894
67 dBm
Note
The 19 dBm same-channel cell separation is simplified and is considered ideal. It is very unlikely that
this 19 dBm of separation can be achieved in most deployments. The most important RF design criteria
are the -67 dBm cell radius and the minimum 20% recommended overlap between cells. Designing to
these constraints optimizes channel separation.
Wireless roaming is not limited to wireless phones but also applies to software-based phones running on
wireless personal computers. For example, a user can roam wirelessly throughout the campus with a
laptop computer running Cisco IP Communicator or Cisco Jabber.
Most wireless APs, wireless phones, and wireless PC clients provide a variety of security options for
providing secure access to the enterprise WLAN. In all cases, select a security method supported by both
the WLAN infrastructure and the wireless devices that matches the security policies and requirements
of the enterprise.
For more information on the Cisco Unified Wireless Network Infrastructure, see Wireless LAN
Infrastructure, page 3-61. For more details on real-time traffic over WLAN design, including voice and
video over WLAN, refer to the Real-Time Traffic over Wireless LAN Solution Reference Network Design
Guide, available at
http://www.cisco.com/c/en/us/td/docs/solutions/Enterprise/Mobility/RToWLAN/CCVP_BK_R780
5F20_00_rtowlan-srnd.html
21-7
Chapter 21
Mobile Collaboration
Note
EM is supported only with Unified CM call control and only on EM-capable endpoint devices.
For more information about EM, see Extension Mobility, page 18-7.
21-8
Chapter 21
Mobile Collaboration
Mobility Within the Enterprise
Maximum of 27 simultaneous voice over WLAN (VoWLAN) bidirectional streams per 802.11g/n
(2.4 GHz) channel cell with Bluetooth disabled and 24 Mbps or higher data rates.
Assuming a video resolution of 720p (high-definition) and a video bit rate of up to 1 Mbps, a
maximum of 8 simultaneous VVoWLAN bidirectional streams per 802.11 g/n (2.4 GHz) with
Bluetooth disabled or 802.11 a/n/ac (5 GHz) channel cell.
These voice and video call capacity values are highly dependent upon the RF environment, the
configured or supported video resolution and bit rates, the wireless endpoint and its specific capabilities,
and the underlying WLAN system features. Actual capacities for a particular deployment could be less.
Note
A single call between two wireless endpoints associated to the same AP is considered to be two
simultaneous bidirectional streams.
Scalability of EM is dependent almost completely on the login/logout rate of the feature within
Unified CM. It is important to know the number of extension mobility users enabled within the
Unified CM cluster as well as how many users are moving around the campus and exercising this feature
at any given time to ensure that sufficient EM login/logout capacity can be provided to these mobile
users. For more information on EM capacity planning, see the chapter on Collaboration Solution Sizing
Guidance, page 25-1.
In all cases, the Unified CM cluster(s) within the campus must have sufficient device registration
capacity to handle device registration for moved devices, regardless of whether they are wired or
wireless devices. Of course, assuming all devices being moved throughout the campus are already
deployed within the campus network, then sufficient capacity within the call control platform should
already be in place prior to the movement of devices. If new devices are added to the deployment for
mobility purposes, however, device registration capacity should be considered and, if necessary,
additional capacity should be added.
Finally, given the many features and functions provided by Unified CM, configuration and deployment
of these mobility solutions does have sizing implications for the overall system. Determining actual
system capacity is based on considerations such as number of endpoint devices, EM users, and busy hour
21-9
Chapter 21
Mobile Collaboration
call attempt (BHCA) rates to number of CTI applications deployed. For more information on general
system sizing, capacity planning, and deployment considerations, see the chapter on Collaboration
Solution Sizing Guidance, page 25-1.
To accommodate physical device mobility within the campus ensure that the network connection
used at a new location has the same type of IP connectivity (VLANs, inter-VLAN routing, and so
forth), connection speed, quality of service, security, and network services (in-line power, dynamic
host control protocol (DHCP), and so forth) as provided by the previous network connection. Failure
to replicate these connection parameters, services, and features will lead to diminished functionality
and in some case complete loss of functionality.
When deploying wireless IP devices and software-based clients, it is imperative to conduct a WLAN
radio frequency (RF) site survey before, during, and periodically after the deployment to determine
appropriate cell boundaries, configuration and feature settings, capacity, and redundancy to ensure
a successful voice and video over WLAN (VVoWLAN) deployment.
APs should be deployed with a minimum cell overlap of 20%. This overlap ensures that a dual-mode
device can successfully roam from one AP to the next as the device moves around within a location,
while still maintaining voice and data network connectivity.
APs should be deployed with cell power level boundaries (or channel cell radius) of -67 dBm in
order to minimize packet loss. Furthermore, the same-channel cell boundary separation should be
approximately 19 dBm. A same-channel cell separation of 19 dBm is critical for ensuring that APs
or clients do not cause co-channel interference to other devices associated to the same channel,
which would likely result in poor voice and video quality.
Deploy EM services in a highly redundant manner so that the loss of a single Unified CM node does
not have adverse effects on the feature operation. If EM services are critical, consider deploying a
server load balancing solution to route around Unified CM node failures and provide highly
available functionality. For more information on EM high availability, see High Availability for
Extension Mobility, page 18-15.
Provide sufficient wireless voice and video call capacity on the campus network by deploying the
appropriate number of wireless APs to handle the desired call capacity based on wireless user BHCA
rates. Each 802.11g/n (2.4 GHz) or 802.11a/n/ac (5 GHz) channel cell can support a maximum of
27 simultaneous voice-only calls with 24 Mbps or higher data rates. Each 802.11g/n (2.4 GHz) or
802.11a/n/ac (5 GHz) channel cell can support a maximum of 8 simultaneous video calls assuming
720p video resolution at up to 1 Mbps bit rate. For 2.4 GHz WLAN deployments, Bluetooth must
be disabled to achieve this capacity. Actual call capacity could be lower depending on RF
environment, wireless endpoint type, and WLAN infrastructure.
21-10
Chapter 21
Mobile Collaboration
Multisite Enterprise Mobility
Central Site
Wireless IP
device roaming
Device
Mobility
Extension
Mobility
PSTN
IP
WAN
Regional Site
Wireless IP
device roaming
348814
Branch Site
B
21-11
Chapter 21
Mobile Collaboration
Note
While Figure 21-4 depicts a multisite deployment with centralized call processing (as evidenced by a
single Unified CM cluster within the central site), the same design and deployment considerations for
multisite enterprise mobility deployments apply to distributed call processing environments. Differences
in mobility feature operation when deployed in distributed call processing environments are described
in the following discussions.
21-12
Chapter 21
Mobile Collaboration
Multisite Enterprise Mobility
Note
EM and EMCC are supported only with Unified CM call control and only on EM-capable endpoint
devices.
Device Mobility
With Cisco Unified CM, a site or a physical location is identified using various settings such as locations,
regions, calling search spaces, and media resources. Cisco Unified IP Phones residing in a particular site
are statically configured with these settings. Unified CM uses these settings for proper call
establishment, call routing, media resource selection, and so forth. However, when dual-mode phones
and other mobile client devices such as Cisco Unified Wireless IP Phones are moved from their home
site to a remote site, they retain the home settings that are statically configured on the phones.
Unified CM then uses these home settings on the phones in the remote site. This situation is undesirable
because it can cause problems with call routing, codec selection, media resource selection, and other call
processing functions.
Cisco Unified CM uses a feature called Device Mobility, which enables Unified CM to determine if the
IP phone is at its home location or at a roaming location. Unified CM uses the device's IP subnets to
determine the exact location of the IP phone. By enabling device mobility within a cluster, mobile users
can roam from one site to another, thus acquiring the site-specific settings. Unified CM then uses these
dynamically allocated settings for call routing, codec selection, media resource selection, and so forth.
This section begins with a discussion surrounding the main purpose for the Device Mobility feature,
followed by an in-depth discussion of the Device Mobility feature itself. This discussion covers the
various components and configuration constructs of the Device Mobility feature. This section also
presents an in-depth discussion of the impact of the Device Mobility feature on the enterprise dial plan,
including the implication for various dial plan models.
Note
21-13
Chapter 21
Mobile Collaboration
Headquarters
Denver
(303)
555-1234
IP
M
M
IP
M
IP
PSTN
V
G.711
IP
IP
Dials
9-1-303555-1234
IP
IP
Branch 2
IP
345893
IP
When a user in Branch1 moves to Branch2 and calls a PSTN user in Denver, the following behavior
occurs:
Unified CM is not aware that the user has moved from Branch1 to Branch2. An external call to the
PSTN is sent over the WAN to the Branch1 gateway and then out to the PSTN. Thus, the mobile user
continues to use its home gateway for all PSTN calls.
The mobile user and Branch1 gateway are in the same Unified CM region and location.
Location-based call admission control is applicable only for devices in different locations, and an
intra-region call uses the G.711 voice codec. Thus, the call over the IP WAN to the Branch1 gateway
uses the G.711 codec and is not tracked by Unified CM for purposes of call admission control. This
behavior can result in over-subscription of the IP WAN bandwidth if all the remote links are
low-speed links.
The mobile user creates a conference by adding multiple Branch2 users to the existing call with the
PSTN user in Denver. The mobile user uses the conferencing resource that is on the Branch1
gateway, therefore all conference streams flow over the IP WAN.
21-14
Chapter 21
Mobile Collaboration
Multisite Enterprise Mobility
Note
Device Mobility is an intra-cluster feature and does not span multiple Unified CM clusters. In distributed
call processing environments, Device Mobility must be enabled and configured on each Unified CM
cluster within the deployment.
Note
In deployments where Device Mobility is not configured, administrators may wish to over-provision
WAN bandwidth between site locations to ensure that physical movement of devices across the WAN
and between sites does not over-subscribe the WAN. The amount of bandwidth to over-provision on each
WAN link depends on the anticipated rate at which users will move devices between two locations.
Device Mobility Info Configures IP subnets and associates device pools to the IP subnets.
Device Mobility Group Defines a logical group of sites with similar dialing patterns (for example,
US_dmg and EUR_dmg in Figure 21-6).
Physical Location Defines the physical location of a device pool. In other words, this element
defines the geographic location of IP phones and other devices associated with the device pool. (For
example, all San Jose IP phones in Figure 21-6 are defined by physical location SJ_phyloc.)
21-15
Chapter 21
Mobile Collaboration
Figure 21-6
Device Mobility
Info
Device Pool
Physical
Location
Device Mobility
Group
SJ1_dmi
10.1.1.0/24
SJ-A_dp
(building A)
SJ2_dmi
10.1.2.0/24
SJ-B1_dp
SJ_phyloc
(building B)
(SJ campus)
SJ3_dmi
10.1.3.0/24
SJ-B2_dp
US_dmg
(building B)
10.2.1.0/24
LON_dmi
10.10.10.0/24
RTP_dp
RTP_phyloc
LON_dp
LON_phyloc
(RTP campus)
EUR_dmg
(LON campus)
190179
RTP_dmi
Unified CM assigns a device pool to an IP phone based on the device's IP subnet. The following steps,
illustrated in Figure 21-7, describe the behavior:
1.
The IP phone tries to register to Unified CM by sending its IP address in the Skinny Client Control
Protocol (SCCP) or Session Initiation Protocol (SIP) registration message.
2.
Unified CM derives the device's IP subnet and matches it with the subnet configured in the Device
Mobility Info.
3.
If the subnet matches, Unified CM provides the device with a new configuration based on the device
pool configuration.
Figure 21-7
1. Register me with
10.10.23.10
IP subnet
Device
Device Pool
Mobility Info
10.10.1.x
Dallas_DMI
Dallas_DP
10.10.22.x
RTP_DMI
RTP_DP
10.10.23.x
SJC_DMI
SJC_DP
SJC
3. Here is your
configuration
Unified CM
2. Hes in SJC!
190178
21-16
Chapter 21
Mobile Collaboration
Multisite Enterprise Mobility
Unified CM uses a set of parameters under the device pool configuration to accommodate Device
Mobility. These parameters are of the following two main types:
The parameters under these settings will override the device-level settings when the device is roaming
within or outside a Device Mobility Group. The parameters included in these settings are:
Date/time Group
Region
Location
Network Locale
SRST Reference
Physical Location
The roaming sensitive settings primarily help in achieving proper call admission control and voice codec
selection because the location and region configurations are used based on the device's roaming device
pool.
For more details on various call admission control techniques, see the chapter on Bandwidth
Management, page 13-1.
The roaming sensitive settings also update the media resource group list (MRGL) so that appropriate
remote media resources are used for music on hold, conferencing, transcoding, and so forth, thus
utilizing the network efficiently.
The roaming sensitive settings also update the Survivable Remote Site Telephony (SRST) gateway.
Mobile users register to a different SRST gateway while roaming. This registration can affect the dialing
behavior when the roaming phones are in SRST mode.
For example, if a user moves with their phone to a new location that loses connectivity to Unified CM,
then based on the roaming sensitive Device Mobility settings, a new SRST reference is configured for
the moved phone and the moved phone will now be under control of the local roaming location SRST
router. When this occurs, not only would the user's phone be unreachable from the PSTN or other sites
because the devices DID will not have changed and will still be anchored at their home location, but in
addition reachabililty from devices within the local failed site might be difficult without the use of
abbreviated dialing as implemented within SRST.
As an example, assume that a user moves a phone from their home location in San Jose, which has a
directory number of 51234 and an associated DID of 408 555 1234 to a remote location in New York,
and that the link between the New York site and San Jose fails shortly after the user roams to the New
York location. In this scenario the phones in the New York site will all fail-over to the SRST router in
that site. The roaming/moved phone will also register to the New York SRST router because its SRST
reference was updated based on the device mobility roaming sensitive settings. In this scenario, the local
New York devices will register to the SRST router with five-digit extensions just as they do to
Unified CM, and as a result the roaming phone still has a directory number of 51234. To reach the
roaming phone from all other sites and from the PSTN, the number 408 555 1234 will be routed to the
San Jose PSTN gateway to which this particular DID is anchored. Because the New York site is
disconnected from the San Jose site, any such calls will be routed to the users voicemail boxes since
they will be unreachable at their desk phones. Likewise, calls internally within the local failed site will
21-17
Chapter 21
Mobile Collaboration
have to be dialed using five-digit abbreviated dialing or based on the configured digit prefixing as
defined by the dialplan-pattern and extension-length commands within the SRST router. In either case,
local callers will have to be understand the required dialing behavior for reaching the local roaming
device by abbreviated dialing. In some cases this may be simply five-digit dialing or it may be that users
have to dial a special digit prefix to reach the local roaming phone. The same logic applies to outbound
dialing from the moved or roaming phone in New York because its dialing behavior might have to be
altered in order to reach local extensions using abbreviated dialing. Outbound dialing to the PSTN from
the local roaming device should remain the same, however.
Device Mobility Related Settings
The parameters under these settings will override the device-level settings only when the device is
roaming within a Device Mobility Group. The parameters included in these settings are:
AAR Group
The device mobility related settings affect the dial plan because the calling search space dictates the
patterns that can be dialed or the devices that can be reached.
Device Mobility Group
Device Mobility Group, as explained earlier, defines a logical group of sites with similar dialing patterns
(for example, sites having the same PSTN access codes and so forth). With this guideline, all sites have
similar dialing patterns in the site-specific calling search spaces. Sites having different dialing behavior
are in a different Device Mobility Group. As illustrated in Figure 21-6, the San Jose and RTP sites
Device Mobility Info, Device Pools, and Physical Locations are different; however, all of these have
been assigned to the same Device Mobility Group US_dmg because the required dialing patterns and
PSTN access codes are the same between the two locations. On the other hand, the London site is
assigned to a separate Device Mobility Group EUR_dmg due to the fact that the required dialing patterns
and PSTN access codes there are different than those of the US sites. A user roaming within a Device
Mobility Group may preserve his dialing behavior at the remote location even after receiving a new
calling search space. A user roaming outside the Device Mobility Group may still preserve his dialing
behavior at the remote location because he uses his home calling search space.
However, if a Device Mobility Group is defined with sites having different dialing patterns (for example,
one site requires users to dial 9 to get an outside line while another site requires users to dial 8 to get an
outside line), then a user roaming within that Device Mobility Group might not preserve his same dialing
behavior at all locations. A user might have to dial digits differently at different locations after receiving
a new calling search space at each location. This behavior can be confusing for users, therefore Cisco
recommends against assigning sites with different dialing patterns to the same Device Mobility Group.
21-18
Chapter 21
Mobile Collaboration
Multisite Enterprise Mobility
The flowchart in Figure 21-8 represents the operation of the Device Mobility feature.
Figure 21-8
Device Registers
Y
Device Mobility
Mode ON?
Devices IP
Subnet matches
DMI
Y
Select roaming DP from the DMI
(round robin if more than one DP)
Is the DMG
different?
Compare the PL of DP
associated with DMI with PL of
home DP associated to Device
Is the PL
different?
Is any overlapping
parameter set to
NONE on device?
LEGEND
DMI:Device Mobility Info
PL:Physical Location
DP: Device Pool
DMG:Device Mobility Group
190166
21-19
Chapter 21
Mobile Collaboration
If the overlapping parameters listed in Figure 21-8 have the same configurations on the device as
well as the device pool, then these parameters may be set to NONE on the device. These parameters
must then be configured on the device pool. This practice can greatly reduce the amount of
configuration because the devices do not have to be configured individually with all the parameters.
Define one physical location per site. A site may have more than one device pool.
Define sites with similar dialing patterns for PSTN or external/off-net access with the same Device
Mobility Group.
A "catch-all" Device Mobility Info with IP subnet 0.0.0.0 may be defined for all non-defined
subnets, depending on the company policy. This Device Mobility Info may be used to assign a device
pool that can restrict access or usage of the network resources. (For example, the device pool may
be configured with a calling search space NONE that will block any calls from the device associated
with this device pool while roaming.) However, by doing so, administrators must be aware of the
fact that this will block all calls, even 911 or other emergency calls. The calling search space may
be configured with partitions that will give access only to 911 or other emergency calls.
Typically the desired egress gateway selection behavior of roaming devices is to use gateways local to
the visited site. The recommended way to implement egress gateway selection that is specific to the
calling device is to use PSTN route patterns pointing to route lists that use Standard Local Route Group.
Using Standard Local Route Group in a route list effectively means that Standard Local Route Group,
when routing an actual call, will be replaced with the Local Route Group configured in the device pool
of the calling endpoint. This schema ensures that site-unspecific route patterns and route lists are used;
site-specific egress gateway selection completely relies on device pool-level Local Route Group
configuration.
For roaming devices (whether roaming inside or between device mobility groups), the device mobility
feature always ensures that the Local Route Group of the roaming device pool is used as Standard Local
Route Group. This guarantees that, with Local Route Group egress gateway selection, a visited
site-specific route group (and thus gateways local to the visited site) will typically be used. This behavior
ensures that, for example, emergency calls routed via route patterns that use a Standard Local Route
Group route list will always use egress gateways local to the visited site.
Local Route Group egress gateway selection can be used with all dial plan approaches explained in the
chapter on Dial Plan, page 14-1.
If certain calls from roaming endpoints need to be routed through gateways local to the home site of the
roaming phone, then routing for these calls has to be implemented through route patterns pointing to
route lists that use fixed site-specific route groups instead of Standard Local Group.
21-20
Chapter 21
Mobile Collaboration
Multisite Enterprise Mobility
In a line/device dial plan approach, these route patterns would be addressed by the device CSS
configured on the endpoint. When roaming but not leaving the device mobility group, the calling
endpoints device CSS is replaced by the Device Mobility CSS configured on the roaming device pool.
If fixed egress gateway selection is required for some calls and the route patterns for those calls are
addressed by the device CSS, you have to make sure that roaming devices always roam across device
mobility groups. This will guarantee that roaming endpoints always use the device CSS configure on the
endpoint.
When using the +E.164 dial plan approach explained in the chapter on Dial Plan, page 14-1, all PSTN
route patterns are accessible by the line CSS, which is not changed or updated for roaming devices. In
this dial plan, site-specific route patterns tying specific PSTN destinations to fixed gateways (for
example, in the home location of the roaming device) are not affected by device mobility operation.
21-21
Chapter 21
Mobile Collaboration
Variable Length On-Net Dialing with Flat Addressing Using the Line/Device Approach without Local Route Group
Figure 21-9 shows a variable-length on-net dial plan with flat addressing for Device Mobility.
Figure 21-9
Variable-Length On-Net Dial Plan with Flat Addressing for Device Mobility
Device
Mobility
Info
Line
Calling
Search
Space
Device
Pool
Device
Calling
Search
Space
Patterns in
Partitions(pt)
Route Lists
Delivers 89191XXX
Site1_Translation_pt
1XXX [Prefix 8919]
Site1_PSTN_pt
911
9.911
Site1_DP
Site1_D_ CSS
9.[2-9]XXXXXX
9.1[2-9]XX[2-9]XXXXXX
Site1_DMI
Site1_
PSTNRL
To Site 1
Route
Groups &
Devices
Site2_
PSTNRL
To Site 2
Route
Groups &
Devices
9.011!
9.011!#
Site 1 Phones
Extension:
89191XXXDID:
(919)234-1XXX
Internal_pt
89191000
89191001
84081000
Block_Local_pt
Internal_CSS
9.[2-9]XXXXXX
Block_National_pt
9.1[2-9]XX[2-9]XXXXXX
Local_ CSS
BlocK_INTL_pt
9.011!
National_CSS
9.011!#
INTL_CSS
(None)
Site 2 Phones
Extension:
84081XXXDID:
(408)555-1XXX
Site2_PSTN_pt
911
Site2_DMI
9.911
Site2_DP
Site2_D_ CSS
9.[2-9]XXXXXX
9.1[2-9]XX[2-9]XXXXXX
9.011!
9.011!#
Delivers 84081XXX
190171
Site2_Translation_pt
1XXX [Prefix 8408]
21-22
Chapter 21
Mobile Collaboration
Multisite Enterprise Mobility
The following design considerations apply to the dial plan model in Figure 21-9:
In this dial plan the translation patterns implementing 4-digit intra-site dialing are addressed by the
device CSS. This is done to avoid the requirement to have site-specific line CSSs. Mobile users
inherit the intra-site dialing of the visited site because the device CSS is updated with the roaming
device pools device mobility CSS (assuming the user is roaming inside the device mobility group).
If this behavior is not desired, consider defining each site as a Device Mobility Group. However,
users must be aware that, for any external PSTN calls, the mobile phone continues to use the home
gateway and therefore consumes WAN bandwidth. This can be avoided by using Standard Local
Route Group (see Egress Gateway Selection for Roaming Devices, page 21-20).
Additional device calling search spaces may be configured for roaming users with access only to the
PSTN and internal phones partitions. This configuration will need at least one additional device pool
and calling search space per site. Thus, N sites will need N device pools and N calling search spaces.
However, this configuration will not require defining each site as a Device Mobility Group. With
this configuration mobile users, when roaming, will not have access to dialing habits through
translation patterns in their device CSS.
Mobile users registered with a remote SRST gateway have unique extensions. However, mobile
users must be aware that no PSTN user can call them when they are registered to a remote SRST
gateway.
+E.164 Dial Plan with Traditional Approach and Local Route Group
As described in the chapter on Dial Plan, page 14-1, the line/device approach has some specific issues,
and creating a +E-164 dial plan based on the line/device approach is not recommended. The
recommended approach for +E.164 dial plans is to combine class of service selection and dialing
normalization on the line CSS and use the Local Route Group feature to address the requirement for
site-specific egress gateway selection. In this approach the device CSS on the phone is not used at all. If
you combine this approach with device mobility, the only roaming sensitive component of the design is
the device pools local route group. For a roaming phone (whether roaming inside or between device
mobility groups), the local route group defined on the phones home device pool will always be updated
with the local route group defined on the roaming device pool. This guarantees that all calls always
egress through a gateway local to the visited site.
Note
Cisco TelePresence System endpoints do not support registration redundancy with Cisco IOS SRST.
21-23
Chapter 21
Mobile Collaboration
Ensure that all critical services (device registration, PSTN connectivity, DNS, DHCP, and so forth)
are deployed at each site in a multisite deployment so that failure of the connection between the site
and other sites does not disrupt critical operations. In addition, ensure that a sufficient number of
physical network ports and wireless LAN APs are available at each site to support movement of
devices and required call capacity.
In situations in which sites with different dialing patterns (for example, sites having different PSTN
access codes) are configured in the same Device Mobility Group, roaming users might have to dial
numbers differently based on their location, which can be confusing. For this reason, Cisco
recommends assigning sites with similar dialing patterns (for example, sites having the same PSTN
access codes) to the same Device Mobility Group. Doing so ensures that roaming users can dial
numbers the same way at all sites within the Device Mobility Group.
The Device Mobility settings from the "roaming" device pool are applied only when users roam
within the same Device Mobility Group; therefore, avoid roaming between different Device
Mobility Groups because the resulting call routing behavior will cause originated calls from the
moved phone to be routed using the "home" or device-configured calling search space. This can lead
to unnecessary consumption of WAN bandwidth because the call might be routed through a different
site's gateway rather than the local "roaming" gateway.
Define only one physical location per site. This ensures that device mobility is engaged only in
scenarios in which a user is roaming between sites. For roaming within the same site, the concerns
that mandate Device Mobility (for example, WAN bandwidth consumption, codec selection, and call
admission control) are not present because low-speed links typically are not deployed within a
single site.
In failover scenarios, "roaming" phones will utilize the SRST reference/gateway as dictated by the
"roaming" device pool's roaming sensitive settings. Therefore, in these situations the "roaming"
phone is unreachable from the PSTN due to the fact that the DID for this phone is anchored in
another location's PSTN gateway. Furthermore, for outbound calls from the "roaming" phone,
dialing behavior might have to be altered for things such as PSTN access codes, and speed dials
configured on the phone might not be usable.
If your system requires the ability to use abbreviated dialing or to use speed dials that rely on
abbreviated dialing, Cisco recommends using a Uniform On-net dial plan model because it will
ensure that abbreviated dialing (direct or through speed dials) continues to work even when the
mobile user's phone is in a roaming location. Abbreviated dialing is still possible with this dial plan
model because all extensions or directory numbers are unique across all sites, and therefore
abbreviated dialing can be used universally due to the fact that there are no overlapping extensions.
21-24
Chapter 21
Mobile Collaboration
Remote Enterprise Mobility
If your system uses a Variable Length On-net dial plan model (using either the line/device or the
line-CSS-only +E.164 dial plan approach), Cisco recommends configuring speed dials in a universal
way so that a single unique extension can be reached when called. By configuring speed dials using
full +E.164 numbers or using site or access codes, you can enable roaming users to use the same
speed dials at any location.
If Device Mobility is enabled for users who on occasion access the enterprise network through a
VPN connection, Device Mobility Info (DMI) for VPN attached phones should contain IP subnets
distributed or owned by the VPN concentrators to ensure that "roaming" to a VPN location results
in appropriate dynamic Device Mobility configuration changes. Be sure to associate the DMI with
the same device pool that is used for any devices co-located with the VPN concentrators.
If Device Mobility is enabled for users who access the enterprise network through Cisco Expressway
mobile and remote access, Device Mobility Info (DMI) for Expressway attached devices should
contain IP subnets used by the Expressway-C node(s) to ensure that "roaming" to an Expressway
location results in appropriate dynamic Device Mobility configuration changes. Be sure to associate
the DMI with the same device pool that is used for any devices co-located with the Expressway-C
node.
21-25
Chapter 21
Mobile Collaboration
Figure 21-10
Central Site
Wireless IP
device roaming
Cisco
Expressway
Remote or Home
Office
ce
VPN-less
Cisco
ASA
PSTN
VPN
Internet
Mobile
Data
Public/Private
802.11
Wi-Fi Hotspot
Regional Site
Wireless IP
device roaming
348815
Mobile
U
Users
21-26
Chapter 21
Mobile Collaboration
Remote Enterprise Mobility
21-27
Chapter 21
Mobile Collaboration
Cisco
Jabber
Cisco
ASA
Internet
Off
f ice
Home Office
Unified
ed CM
IP WAN
Remote
motte B
Branch
ran
348761
Central
ntral Site
The following design guidelines pertain to enabling the Device Mobility feature for user devices at a
remote site connected to the enterprise through a client or router-based VPN connection:
Configure Device Mobility Info (DMI) with the IP subnets distributed or owned by the VPN
concentrators.
Associate the DMI with the same device pool that is used for devices co-located with the VPN
concentrators. However, parameters such as calling privileges, network locale, and so forth, must be
taken into consideration.
Educate the remote site users to point to the geographically nearest enterprise VPN concentrator
when making client-based or router-based VPN connections.
These guidelines ensure that call admission control is correctly applied on the enterprise WAN and over
the connection to the remote site.
21-28
Chapter 21
Mobile Collaboration
Remote Enterprise Mobility
For information on deploying a VPN, refer to the various VPN design guides available under the Security
in WAN subsection of the Design Zone for Security, available at:
http://www.cisco.com/c/en/us/solutions/enterprise/design-zone-security/landing_wan_security.htm
l
Cisco Expressway
The mobile and remote access feature of the Cisco Expressway solution provides secure reverse proxy
firewall traversal connectivity, which enables remote users and their devices to access and consume
enterprise collaboration applications and services.
As shown in Figure 21-12, the Cisco Expressway solution encompasses two main components: the
Expressway-E node and the Expressway-C node. These two components work in combination with
Unified CM to enable secure mobile and remote access. The Expressway-E node provides the secure
edge interface to mobile and remote devices. This node normally resides in the DMZ area of the
enterprise network and creates a secure TLS connection with the Expressway-C node. The
Expressway-C node provides proxy registration to Unified CM for remote secure endpoint registration.
The Expressway-C node also provides media traversal capabilities.
21-29
Chapter 21
Mobile Collaboration
Figure 21-12
Secure Remote Collaboration with Cisco Expressway Mobile and Remote Access
Home
Office
802.11
Wi-Fi
Internet
Unified CM
M
Mobile
Data
a
E
Expressway-C Expressway-E
Voice and video
(RTP/sRTP)
Cisco Expressway
Unity
Connection
Coffee
Shop
Wi-Fi
H
t
Hotspot
348763
IM &
Presence
presence,
IM & prese
visual voicemail,
directory access
(XMPP, HTTPS)
S))
Once registered to Unified CM, the remote device is able to make and receive voice and video calls over
IP using SIP signaling and RTP media. The secure Cisco Expressway mobile and remote connection not
only enables device registration and voice and video calling, but it also enables additional collaboration
workflows including IM and presence, visual voicemail, and corporate directory access. The full
collaboration feature set is available from the enterprise without requiring a VPN tunnel. Voice and video
media as well as signaling and other collaboration traffic traverse the enterprise network at the
Expressway-C node. As shown in Figure 21-12, calls between two remote devices outside the enterprise
will be hairpinned at the Expressway-C node within the enterprise.
Unlike with VPN secure connections where all traffic from the secured endpoint traverses the VPN
tunnel back to the enterprise, Cisco Expressway mobile and remote access enables secure connectivity
to the enterprise for collaboration traffic only. Non-collaboration workflows and traffic do not traverse
the secure Cisco Expressway connection. Instead, all other traffic is sent directly to the local network or
the Internet and does not traverse the enterprise network.
The Cisco Expressway mobile and remote access functionality supports both Cisco hardware endpoints
and Cisco Jabber software-based client endpoints. Supported Cisco hardware endpoints include Cisco
TelePresence C, EX, MX, and SX Series video endpoints and Cisco DX, 7800, and 8800 Series desk
phones. Cisco Jabber desktop and mobile clients also support Cisco Expressway mobile and remote
21-30
Chapter 21
Mobile Collaboration
Remote Enterprise Mobility
access. In particular, Cisco Jabber mobile clients support Cisco Expressway mobile and remote access
connectivity while in motion, thus enabling secure real-time collaboration regardless of the mobile user's
location or network connectivity type.
Just as when relying on VPN for remote secure connectivity, Device Mobility configuration with
Expressway mobile and remote access is critical for ensuring Unified CM is able to track endpoint
locations for the purposes of monitoring call volume over low-speed links, negotiating appropriate
codecs, and routing calls using local gateway resources. When configuring Device Mobility in
environments with Expressway mobile and remote access, remember to:
Configure Device Mobility Info (DMI) with the IP subnet(s) used by the Expressway-C nodes.
Associate the DMI with the same device pool that is used for any devices co-located with the
Expressway-C node.
Cisco Expressway mobile and remote access functionality supports a maximum of 10,000 remote
endpoint registrations to Unified CM per Expressway-C and Expressway-E cluster pair. In addition,
Expressway cluster pairs support a maximum of 2,000 simultaneous video calls or 4,000 simultaneous
voice-only calls. For more information about Cisco Expressway capacity, including per-Expressway
node capacities, see the section on Cisco Expressway, page 25-35.
Deploy multiple Expressway clusters for increased scale or for designs spanning multiple geographic
locations. In the case of multi-site deployments, Expressway clusters should be distributed across
geographic regions to provide remote enterprise connectivity to users and their devices regardless of
location. In order to effectively distribute Expressway mobile and remote access connections so that
devices connect to the nearest Expressway service node or cluster, Geo DNS services are recommended.
With Geo DNS service, mobile devices are usually directed to the nearest Expressway service point
based on location as determined by the source IP address of the DNS query for Expressway DNS service
records or based on the shortest mean latency between the location of the device and available
Expressway service nodes.
For more information about the Cisco Expressway solution, refer to the data sheet and documentation
available at
http://www.cisco.com/c/en/us/products/unified-communications/expressway-series/index.html
21-31
Chapter 21
Mobile Collaboration
When using Device Mobility, remember to configure Device Mobility Info (DMI) with the IP
subnets distributed or owned by the VPN concentrators, or in the case of Expressway, with the
subnet(s) used by the Expressway-C nodes. Assign the DMI to the same device pool that is
configured for devices deployed in the same location as the VPN concentrators or Expressway-C
nodes.
Educate remote site users to select the nearest VPN concentrator for VPN connection.
Ensure appropriate VPN session capacity is available in order to provide connectivity to all remote
site locations and devices using VPN.
Ensure appropriate reverse proxy firewall traversal session capacity is available in order to provide
VPN-less secure connectivity to all remote devices. Ensure that sufficient Expressway-E and
Expressway-C nodes and session capacity are available. In all cases, sufficient Unified CM
registration capacity is required.
21-32
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
The fixed mobile convergence (FMC) mobility functionality delivered within the Cisco Unified Mobility
solution is provided through Cisco Unified CM and can be used in conjunction with Cisco mobile clients
and devices such as Cisco Jabber.
Cisco Unified Mobility provides the following mobility application functionality:
Mid-Call Features
Mid-call features allow a user to invoke hold, resume, transfer, conferencing, and directed call park
features from their mobile phone during in-progress mobility calls. These features are invoked from
the mobile phone keypad and take advantage of enterprise media resources such as music on hold
and conference bridges.
Cisco mobile clients and devices provide the ability to attach to both the mobile provider network and
802.11 wireless networks for voice and data connectivity. This enables users to leverage both enterprise
call control and in some cases mobile network call control from a single device. By leveraging the
enterprise telephony infrastructure for making and receiving calls whenever possible and, in the case of
dual-mode phones, falling back to the mobile voice network only when enterprise connectivity is
unavailable, mobile clients and devices can help reduce telephony costs. Dual-mode phones and the
clients that run on them also provide a handoff mechanism so that in-progress voice calls can be moved
easily between the WLAN and mobile voice interfaces as a user moves out of the enterprise.
In addition to enabling mobile devices to make voice or video calls over IP via 802.11 WLAN or mobile
data networks, Cisco mobile clients enable automated enterprise dialing using the Dial via Office
feature. Dial via Office calls are set up using SIP signaling over the IP network, while the media path is
over the mobile voice network and the PSTN. Cisco mobile clients and devices also provide other unified
communications services such as corporate directory access, presence and instant messaging (IM).
These devices and clients enable mobile users to remain productive whether inside or outside the
enterprise by providing access to collaboration applications while at the same time enabling users to
21-33
Chapter 21
Mobile Collaboration
make and receive enterprise calls from their mobile devices, whether outside the enterprise over public
or private WiFi hot spots or the mobile data network, or inside the enterprise and over the WLAN
network.
This section begins with a discussion of Unified Mobility features, functionality, and design and
deployment considerations. Given the various benefits of Unified Mobility and the fact that mobile
clients and devices can be integrated to take advantage of the features provided, this discussion paves
the way for examination of mobile client applications such as Cisco Jabber. This section also includes a
discussion of architecture, functionality, and design and deployment implications for the following
mobility applications and features:
21-34
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-13
IP Phone
Virtual Phone
IP
Line Level
Configuration
IP
Shared
Line
Remote Destination #1
408 555-6789
Remote Destination #2
XXX XXX-XXXX
Remote Destination #3
191959
As further shown in Figure 21-13, a mobility user can have one or more remote destinations configured
and associated with their remote destination profile. A remote destination represents a single PSTN
phone number where a user can be reached. A user can have up to 10 remote destinations defined. Call
routing timers can be configured for each remote destination to adjust the amount of time a call will be
extended to a particular remote phone, as well as the amount of time to wait before extending the call
and the amount of time that must pass before a call can be answered at the remote phone. Mobility users
can also configure filters for each remote destination to allow or deny calls from certain phone numbers
to be extended to that remote phone.
21-35
Chapter 21
Mobile Collaboration
Enterprise
408 555-7890
PSTN
Gateway
Cisco Unified
CM cluster
Remote
Destination:
408 555-7890
4
2
Phone A
Remote
IP
Destination 3
Profile
DN: 408 555-1234
DN: 408 555-1234
Shared Line
191960
PSTN
Typically a Single Number Reach user's configured remote destination is their mobile phone on a mobile
voice or cellular provider network; however, any destination reachable by means of the PSTN can be
configured as a user's remote destination. Furthermore, an SNR user can have up to 10 remote
destinations configured, so an incoming call could potentially ring as many as 10 PSTN phones as well
as the user's desk phone. Once the call is answered at the desk phone or at a remote destination phone,
any other call legs that have been extended to ring additional remote destinations or the desk phone (if
not answered at the desk phone) will be cleared. If the incoming call is answered at the remote
destination, the voice media path will be hairpinned within the enterprise PSTN gateway utilizing two
gateway ports. This utilization must be considered when deploying the SNR feature.
21-36
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Note
In order for Single Number Reach to work as in Figure 21-14, ensure that the user-level Enable Mobility
check box under the End User configuration page has been checked and that at least one of the user's
configured remote destinations has the Enable Single Number Reach check box checked.
408 555-7890
Enterprise
1
PSTN
Gateway
PSTN
2
IP
191961
Phone A
Desk phone pickup can be performed whenever an enterprise-anchored call is in progress at a configured
remote destination phone and that phone hangs up the call.
Note
An enterprise-anchored call refers to any call that has at least one call leg connected through an
enterprise PSTN gateway and that originated either from a remote destination to an enterprise DID or
from Single Number Reach, Mobile Voice Access, Enterprise Feature Access, or Intelligent Session
Control.
The option to pick up or resume the call at the desk phone is available for a certain amount of time. For
this reason, it is good practice for the Single Number Reach user to ensure that the calling phone hangs
up before the remote destination phone is hung up. This ensures that the call cannot be resumed at the
desk phone by someone else. By default, the call remains available for pickup at the desk phone for 10
seconds after the remote destination phone hangs up; however, this time is configurable and can be set
from 0 to 30000 milliseconds on a per-user basis by changing the Maximum Wait Time for Desk Pickup
parameter under the End User configuration page. Desk phone pickup can also be performed after
invoking the mid-call hold feature at the remote destination phone. However, in these cases, the
Maximum Wait Time for Desk Pickup parameter setting has no effect on the amount of time the call will
be available for pickup. A call placed on mid-call hold will remain on hold and be available for desk
phone pickup until manually resumed at either the remote or desktop phone.
21-37
Chapter 21
Mobile Collaboration
Another method for performing desk phone pickup is to use the mid-call session handoff feature. This
mid-call feature is invoked by manually keying *74, the default enterprise feature access code for session
handoff, which in turn generates a DTMF sequence back to Unified CM. When this feature is invoked,
Unified CM sends a new call to the user's enterprise desk phone. Once this new call is flashing or ringing
at the desk phone, the user then must answer the call to complete the session handoff.
The benefit of this desk phone pickup method over other methods (such as hanging up the call at the
mobile phone or using the mid-call hold feature) is that the conversation between the user and the far-end
phone is maintained throughout the handoff process. Once the *74 sequence has been keyed, the user
can continue the conversation because the handoff call is sent to the user's desk phone. When the user
answers the call at the desk phone, the call legs are shuffled so that the call leg to the far-end is connected
to the new call leg created at the desk phone, thus resulting in an uninterrupted or near-instantaneous
cut-through of the audio path. The original call leg at the mobile device is subsequently cleared.
Unlike the hang-up method for invoking desk phone pickup, where the end-users Maximum Wait Time
for Desk Pickup setting determines how long the call will be available for pickup at the desk phone, with
session handoff the Session Handoff Alerting Timer service parameter determines the amount of time
the call will ring or flash at the desk phone before the handoff call is cleared. The default handoff alerting
time is 10 seconds. Further, with session handoff, any call forward settings configured on the desk phone
do not get invoked. As a result, the handoff feature does not forward to voicemail or any other
call-forward destination. If a call is not answered by the end of Session Handoff Alerting Timer period,
then the call is cleared and the Remote In Use state is removed from the user's desk phone line. However,
in this scenario the original call is maintained at the mobile phone.
For additional information about session handoff and other mid-call features, see Mid-Call Features,
page 21-39.
21-38
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-16
408 555-7890
Enterprise
3
PSTN
PSTN
Gateway
MobileConnect On
IP
DN: 408 555-1234
Exit
Select
191962
Phone A
When a Single Number Reach user has multiple remote destinations configured, each remote destination
will ring when the Select softkey is pressed, and the user can answer the desired phone.
Note
In order for remote destination phone pickup to work as in Figure 21-16, ensure that at least one of the
user's configured remote destinations has the Mobile Phone check box checked. In addition, the Mobility
softkey must be configured for all mobility users by adding the softkey to each user's associated desk
phone softkey template. Failure to check the Mobile Phone check box and to make the Mobility softkey
available to mobility users will prevent the use of remote destination phone pickup functionality.
Note
Cisco TelePresence System C, EX, MX, SX, and TX Series video endpoints do not support remote
destination pickup as described above. These endpoints do not expose a mobility softkey or the "Send
call to Mobile Phone" option to the user. Therefore, these endpoints are unable to send in-progress calls
to the mobile device using remote destination pickup.
Mid-Call Features
As illustrated in Figure 21-17, once a user answers a Single Number Reach call at the remote destination
device (step 1: in this case, 408 555-7890), the user can invoke mid-call features such as hold, resume,
transfer, conference, directed call park, and session handoff by sending DTMF digits from the remote
destination phone to Unified CM via the enterprise PSTN gateway (step 2). When the mid-call feature
hold, transfer, conference, or directed call park is invoked, MoH is forwarded from Unified CM to the
held party (step 3: in this case, Phone A). In-progress calls can be transferred to another phone or
directed call park number, or additional phones can be conferenced using enterprise conference
resources (step 4).
21-39
Chapter 21
Mobile Collaboration
Figure 21-17
Enterprise
408 555-7890
Cisco Unified
CM cluster
PSTN
PSTN
Gateway
1
V
IP
Remote
Destination:
408 555-7890
Remote
Destination
Profile
DN: 408 555-1234
DN: 408 555-1234
Shared Line
Phone A
191963
IP
Mid-call features are invoked at the remote destination phone by a series of DTMF digits forwarded to
Unified CM. Once received by Unified CM, these digit sequences are matched to the configured
Enterprise Feature Access Codes for Hold, Exclusive Hold, Resume, Transfer, Conference, and Session
Handoff, and the appropriate function is performed.
Note
To enable the Directed Call Park mid-call feature, you must configure Cisco Unified CM with directed
call park numbers and call park retrieval prefixes.
Note
In order to perform the transfer, conference, and directed call park mid-call features, a second call leg is
generated by the remote destination phone to a system-configured Enterprise Feature Access DID that
answers the call, takes user input (including PIN number, mid-call feature access code, and target
number), and then creates the required call leg to complete the transfer, conference, or directed call park
operation.
With the mid-call session handoff feature, MoH is not forwarded to the far-end because the far-end is
never placed on hold. Instead, the original audio path is maintained until the mobile user answers the
handoff call at the desk phone. Once the call is answered, the call legs are shuffled at the enterprise
gateway and the audio path is maintained.
Mid-call features are invoked by manually keying the feature access codes and entering the appropriate
key sequences. Table 21-2 indicates the required key sequences for invoking mid-call features.
21-40
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Table 21-2
Mid-Call Feature
Enterprise Feature
Access Code (default) Manual Key Sequence
Hold
*81
Enter: *81
Exclusive Hold
*82
Enter: *82
Resume
*83
Enter: *83
Transfer
*84
Conference
*85
Session Handoff
*74
1. Enter: *74
2. Answer at the desk phone upon ring and/or flash.
Note
Media resource allocation for mid-call features such as hold and conference is determined by the Remote
Destination Profile configuration or, in the case of dual-mode phones and Unified Mobile
Communicator, the device configuration. The media resource group list (MRGL) of the device pool
configured for the Remote Destination Profile or the mobile client device is used to allocate a conference
bridge for the conferencing mid-call feature. The User Hold Audio Source and Network Hold MoH
Audio Source settings of the Remote Destination Profile or the mobile client device, in combination with
the media resource group list (MRGL) of the device pool, is used to determine the appropriate MoH
stream to be sent to a held device.
21-41
Chapter 21
Mobile Collaboration
Timer Control method With this method the system relies on a set of timers (one per remote
destination) in conjunction with system call-forward timers to ensure that, when and if a call is
forwarded to a voicemail system on ring-no-answer, the enterprise voicemail system receives the
call.
User Control method With this method the system relies on a DTMF confirmation tone from the
remote destination when the call is answered to determine if the call was received by the user or a
non-enterprise voicemail system.
System settings determine whether the timer control or user control method is used. The method used
can be set globally via the Voicemail Selection Policy service parameter or for individual remote
destinations via the Single Number Reach Voicemail Policy. By default the system and all remote
destinations use the timer control method
Timer Control Mobile Voicemail Avoidance
For this method, the system relies on a set of timers on the Remote Destination configuration page. The
purpose of these timers is to ensure that, when and if a call is forwarded to a voicemail system on
ring-no-answer, the call is forwarded to the enterprise voicemail system rather than any remote
destination voicemail system. These timers in conjunction with other system forward-no-answer timers
should be configured to avoid non-enterprise voicemail systems as follows:
Ensure the system forward-no-answer time is shorter at the desk phone than at the remote
destination phones.
To do so, ensure that the global Forward No Answer Timer field in Unified CM or the No Answer
Ring Duration field under the individual phone line is configured with a value that is less than the
amount of time a remote destination phone will ring before forwarding to the mobile voicemail
system. In addition, the Delay Before Ringing Timer parameter under the Remote Destination
configuration page can be used to delay the ringing of the remote destination phone in order to
further lengthen the amount of time that must pass before a remote destination phone will forward
to its own mobile voicemail box. However, when adjusting the Delay Before Ringing Timer
parameter, take care to ensure that the global Unified CM Forward No Answer Timer (or the
line-level No Answer Ringer Duration field) is set sufficiently high enough so that the mobility user
has time to answer the call on the remote destination phone. The Delay Before Ringing Timer
parameter can be set for each remote destination and is set to 4,000 milliseconds by default.
Ensure that the remote destination device stops ringing before the incoming call is forwarded to the
mobile voicemail system.
You can accomplish this with the Answer Too Soon and Answer Too Late timers for each remote
destination. First the Answer Too Soon Timer parameter under the Remote Destination
configuration page should be configured with a value that is more than the amount of time it takes
a call extended to a powered-off or out-of-range mobile phone to be forwarded to the mobile
voicemail system. By default this timer is set 1,500 milliseconds (or 1.5 seconds). If the call is
answered before the Answer Too Soon Timer expires, the system will disconnect the call leg to the
remote destination. This ensures that calls forwarded immediately to the mobile voicemail system
will not be connected, but those answered by the user after ring-in are connected.
21-42
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Next configure the Answer Too Late Timer parameter under the Remote Destination configuration
page with a value that is less than the amount of time that a remote destination phone will ring before
forwarding to its voicemail box. By default this timer is set to 19,00 milliseconds (or 19 seconds).
If the call is not answered before this timer expires, the system will disconnect the call leg to the
remote destination. This ensures that the remote destination phone stops ringing before the call is
forwarded to the mobile voicemail system.
Note
Incoming calls to a remote destination that are manually diverted by the mobility user can end up in the
mobile voicemail box if the manual diversion occurs after the Answer Too Soon timer has expired. To
prevent this from happening, mobility users should be configured for the user control method or advised
to ignore or silence the ringing of incoming calls they wish to divert to voicemail. This will ensure that
unanswered calls always end up in the enterprise voicemail system.
Note
In most deployment scenarios, the default Delay Before Ringing Timer, Answer Too Late Timer, and
Answer Too Soon Timer values are sufficient and do not need to be changed.
User Control Mobile Voicemail Avoidance
For this method, the system relies on DTMF confirmation tone from the remote destination when the call
is answered. If a DTMF tone is received by the system, then the system knows that the user answered
the call and pressed a key to generate the DTMF tone. On the other hand, if the DTMF tone is not
received by the system, the system assumes the call leg was answered by a non-enterprise voicemail
system and it disconnects the call leg.
When the user control method is enabled, on answer the end user will hear an audio prompt requesting
that they press a key pad button to generate a DTMF tone. By default the audio prompt is played to the
user one second after the call is answered. The user may not hear the audio prompt if they press the
keypad to generate a DTMF tone immediately upon answering. The audio prompt is played only on the
remote destination call leg and therefore the far-end party will not hear this prompt. Once the audio
prompt is played to the user, by default the system will wait 5 seconds to receive the DTMF tone. If the
tone is not received, the system disconnects the call leg but continues to ring the user's other configured
devices until the call is answered by the user or forwarded to the enterprise voicemail system.
Note
The user control mobile voicemail avoidance method is completely dependent on successful relay of the
DTMF tone from the remote destination on the mobile voice network or PSTN all the way to
Unified CM. The DTMF tone must be sent out-of-band to Unified CM. If DTMF relay is not properly
configured on the network and system, DTMF will not be received and all call legs to remote destinations
relying on the user control method will be disconnected. The system administrator should ensure proper
DTMF interoperation and relay across the enterprise telephony network prior to enabling the user control
method. If DTMF cannot be effectively relayed from the PSTN to Unified CM, then the timer control
mobile voicemail avoidance method should be used instead.
21-43
Chapter 21
Mobile Collaboration
Cisco Unified CM Administration or Cisco Unified CM Self Care Portal for end users
An administrator or user unchecks the Enable Single Number Reach box to disable, or checks the
Enable Single Number Reach box to enable, the feature. This is done per remote destination.
Note
The dialog box that appears when the Mobility softkey is pressed as described above uses the old feature
name, Mobile Connect, rather than the new feature name, Single Number Reach. The feature and
enable/disable functionality are the same.
21-44
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
The SNR phone user who wishes to either enable or disable the SNR feature or to pick up an
in-progress call on their remote destination phone pushes the Mobility softkey on their desk phone
(see step 1 in Figure 21-18).
2.
Unified CM returns the SNR status (On or Off) and offers the user the ability to select the Send Call
to Mobile Phone option when the phone is in the Connected state, or it offers the user the ability to
enable or disable the Mobile Connect status when the phone is in the On Hook state (see step 2 in
Figure 21-18).
3.
Single Number Reach users can use the Unified CM Self Care Portal to configure their own mobility
settings via the web-based configuration pages at
https://<Unified-CM_Server_IP_Address>/ucmuser/
where <Unified-CM_Server_IP_Address> is the IP address of the Unified CM publisher server (see
step 3 in Figure 21-18).
Figure 21-18
Cisco Unified
CM
M
Mobility User
Configuration
191964
IP
Unified CM servers
PSTN gateway
Each component must be redundant or resilient in order for Single Number Reach to continue
functioning fully during various failure scenarios.
21-45
Chapter 21
Mobile Collaboration
In order for SNR users to use the Unified CM Self Care Portal web interface to configure their mobility
settings (remote destinations and access lists), the Unified CM publisher server must be available. If the
publisher is down, users will not be able to change mobility settings. Likewise, administrators will be
unable to make mobility configuration changes to Unified CM; however, existing mobility
configurations and functionality will continue. Finally, changes to SNR status must be written by the
system on the Unified CM publisher server; if the Unified CM publisher is unavailable, then enabling or
disabling SNR will not be possible.
21-46
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Note
If the PSTN phone from which the Mobile Voice Access user is calling is configured as a Single Number
Reach remote destination for that user and the incoming caller ID can be matched against this remote
destination by Unified CM, the user does not have to enter their numeric user ID. Instead they will be
prompted to enter just the PIN number.
In the meantime, Unified CM has forwarded IVR prompts to the gateway, the gateway has played these
prompts to the user, and the gateway has collected user input including the numeric ID and PIN number
of the user. This information is forwarded to Unified CM for authentication and to generate the call to
9 1 972 555 3456 (step 3). After authenticating the user and receiving the number to be dialed,
Unified CM generates a call via the user's Remote Destination Profile (step 4). The outbound call to
972 555-3456 is routed via the PSTN gateway (step 5). Finally, the call rings at the PSTN destination
phone with number 972 555-3456 (step 6).
21-47
Chapter 21
Mobile Collaboration
Figure 21-19
Enterprise
1
408 555-7890
PSTN
Cisco Unified
CM cluster
M
6
972 555-3456
V
Mobile Voice
Access
DID: 408 555-2345 5
Remote
Destination
Profile
DN: 408 555-1234
DN: 408 555-1234
Shared Line
IP
252939
Note
In order for Mobile Voice Access to work as in Figure 21-19, ensure that the system-wide Enable Mobile
Voice Access service parameter is set to True and that the per-user Enable Mobile Voice Access check
box on the End User configuration page is also checked.
Note
The Mobile Voice Access feature relies on the Cisco Unified Mobile Voice Access Service, which must
be activated manually from the Unified CM Serviceability configuration page. This service can be
activated on the publisher node only.
21-48
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Note
When using Mobile Voice Access with hairpinning, users calling into the system will not be identified
automatically by their caller ID. Instead, users will have to key in their remote destination number
manually prior to entering their PIN. The reason the user is not automatically identified is that, for
hairpinning deployments, the PSTN gateway must first route the call to Unified CM to reach the
hairpinned Mobile Voice Access gateway. Because the call is routed to Unified CM first, the conversion
of the calling number from a mobile number to an enterprise directory number occurs prior to the call
being handled by the Mobile Voice Access gateway. This results in the Mobile Voice Access gateway
being unable to match the calling number with a configured remote destination, and therefore the system
prompts the user to enter their remote destination number. This is unique to hairpinning deployments;
with normal Mobile Voice Access flows, the PSTN gateway does not have to route the call to Unified CM
first in order to access Mobile Voice Access because the functionality is available on the local gateway.
In the meantime, the H.323 VoiceXML gateway collects and forwards the user input to Unified CM and
then plays the forwarded IVR prompts to the PSTN gateway and the Mobile Voice Access user.
Unified CM in turn receives user input, authenticates the user, and forwards appropriate IVR prompts to
the H.323 VoiceXML gateway based on user input (step 5). After receiving the number to be dialed,
Unified CM generates a call using the user's Remote Destination Profile (step 6). The outbound call to
972 555-3456 is routed through the PSTN gateway (step 7). Finally, the call rings at the PSTN
destination phone with number 972 555-3456 (step 8).
Mobile Voice Access Using Hairpinning
Mobile Voice
Access
DID: 408 555-2345
1
408 555-7890
H.323 VXML
PSTN
Gateway
PSTN
Gateway
Enterprise
Cisco Unified
CM cluster
M
M
PSTN
3
M
8
972 555-3456
Remote
Destination
Profile
DN: 408 555-1234
DN: 408 555-1234
Shared Line
IP
191966
Figure 21-20
21-49
Chapter 21
Mobile Collaboration
Note
When deploying Mobile Voice Access in hairpinning mode, Cisco recommends configuring the Mobile
Voice Access DID at the PSTN gateway and the Mobile Voice Access Directory Number within Cisco
Unified CM (under Media Resources > Mobile Voice Access) as different numbers. A translation
pattern within Unified CM can then be used to translate the called number of the Mobile Voice Access
DID to the configured Mobile Voice Access directory number. Because the Mobile Voice Access
directory number configured within Unified CM is visible to the administrator only, translation between
the DID and directory number will be invisible to the end user and there will be no change in end-user
dialing behavior. This is recommended in order to prevent mobility call routing issues in multi-cluster
environments. This recommendation does not apply to Mobile Voice Access in non-hairpinning mode.
Note
Mobile Voice Access in hairpinning mode is supported only with H.323 VXML gateways.
Note
Unlike with Mobile Voice Access, Enterprise Feature Access requires that all two-stage dialed calls must
originate from a phone that has been configured as a remote destination in order to match the caller ID
and PIN against the end-user account. There is no provision within Enterprise Feature Access in which
the mobility user can enter their remote destination number or ID to identify themselves to the system.
Identity can be established only via the combination of incoming caller ID and entered PIN.
Next the outgoing call is originated via the user's remote destination profile (step 3), and the call to
PSTN number 972 555-3456 is routed via the enterprise PSTN gateway (step 4). Finally, the call rings
the PSTN phone (step 5: in this case, 972 555-3456). As with Mobile Voice Access, the voice media path
of each Enterprise Feature Access two-stage dialed call is hairpinned within the enterprise PSTN
gateway utilizing two gateway ports.
21-50
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-21
Enterprise
1
408 555-7890
Cisco Unified
CM cluster
PSTN
Gateway
PSTN
Remote
Destination:
408 555-7890
4
M
3
Remote
Destination
Profile
DN: 408 555-1234
DN: 408 555-1234
Shared Line
972 555-3456
Note
191967
IP
In order for Enterprise Feature Access two-stage dialing to work as in Figure 21-21, ensure that the
system-wide Enable Enterprise Feature Access service parameter is set to True.
21-51
Chapter 21
Mobile Collaboration
Note
When the Enable Mobile Voice Access service parameter is set to False, resulting in an inability to make
two-stage dialed calls, Mobile Voice Access still provides users with the ability to enable and disable
Single Number Reach remotely. As long as the Mobile Voice Access Directory Number has been
configured on the system, the user's account has been enabled for Mobile Voice Access, and the Cisco
Unified Mobile Voice Access service is running on the publisher, an authorized calling user can still
enable or disable Single Number Reach.
21-52
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
To control the nature of this matching, consider the following two approaches.
Using Complete Caller ID Matching
With this approach, remote destination numbers are configured exactly as the caller ID would be
presented from the PSTN. For example, if the caller ID from the PSTN for a remote destination phone
is presented to the system as 4085557890, then this number should be configured on the Remote
Destination configuration page.
In order to route Single Number Reach calls appropriately to this remote destination, it is necessary to
configure the dial plan to use either +E.164 dialing methods or a digit prefix mechanism to prefix
necessary PSTN access codes and other required digits. For example, if you are not using a global
+E.164 dial plan and assuming a 9 or other PSTN steering digits or country codes are required to reach
the PSTN when dialing calls from the enterprise, then digit prefixing must be configured to add the
appropriate PSTN steering digit and country code to the beginning of the configured remote destination
number. Digit prefixing should be facilitated by using translation patterns, route patterns, or route list
constructs within the Unified CM system. When using this complete match approach and a digit
prefixing method, the Matching Caller ID with Remote Destination parameter should be left at the
default setting of Complete Match.
Application Dial Rules may also be used to provide digit prefixing in these scenarios. However, it is
worth noting that Application Dial Rules are applied based on called digit-string length and cannot be
partitioned, meaning that they are applied globally across the system. This severely limits the use of
Application Dial Rules, especially in scenarios where multiple dialing domains (for example, different
countries) need to be supported on a single Unified CM cluster.
Note
Not only are Application Dial Rules applied to Single Number Reach, Mobile Voice Access, and
Enterprise Feature Access calls, but they are also applied to calls made with Cisco WebDialer, Cisco
Unified CM Assistant, and Cisco Jabber applications. For this reason, exercise care when configuring
these rules to ensure that dialing behavior across all applications is as expected.
The recommended dial plan approach is always to globalize the caller ID to +E.164 on ingress from the
PSTN and always to configure remote destinations as +E.164. This will guarantee that the caller ID from
the PSTN (after normalization) will always provide a unique match when compared against all
configured remote destinations. Combined with a dial plan supporting +E.164 dialing, this eliminates
the need for digit prefixing and ensures unique identification of remote destination users and numbers
even when supporting multiple international numbering plans. Because the recommended dial plan
approach is to globalize the caller ID on ingress and localize on egress according to trunk requirements
and/or user expectations, using the unmodified caller ID as presented from the PSTN is not compatible
with this approach.
Using Partial Caller ID Matching
With this approach, remote destinations are configured as they would be dialed from the system to the
PSTN. For example, if the number for the remote destination is 14085557890 and PSTN access from the
system requires a 9, then this number should be configured on the Remote Destination configuration
page as 914085557890. This approach precludes the need for configuration of a digit prefixing
mechanism on the system, but it requires setting the Matching Caller ID with Remote Destination service
parameter to Partial Match and setting the Number of Digits for Caller ID Partial Match to the
appropriate number of consecutive digits that should be matched against the remote destination caller
ID. For example, if the caller ID for a remote destination is 14085557890 and the remote destination is
configured as 914085557890, then the Number of Digits for Caller ID Partial Match would ideally be
set to 10 or 11. In this example, this parameter could be set to a lower number of digits; however, always
ensure that enough consecutive digits are matched so that all configured remote destinations in the
21-53
Chapter 21
Mobile Collaboration
system are matched uniquely. If there is no exact match or if more than one configured remote
destination number is matched when using partial caller ID matching, the system treats this as if there
is no matching remote destination number, thus requiring the user to enter their remote destination
number/ID manually in the case of Mobile Voice Access before providing their PIN. With Enterprise
Feature Access, there is no mechanism for the user to enter their remote destination number; therefore,
when using this functionality, ensure that only unique matches occur.
Note
If the PSTN service provider sends variable-length caller IDs, using partial caller ID matching is not
recommended because ensuring a unique caller ID match for each inbound call might not be possible.
In these scenarios, using complete caller ID matching and/or a +E.164 dial plan is the preferred method.
Unified CM forwards IVR prompts and instructions to the H.323 or SIP VXML gateway via HTTP
(see step 1 in Figure 21-22). This provides the VXML gateway with the ability to play these prompts
for the inbound Mobile Voice Access callers.
2.
The H.323 or SIP VXML gateway uses HTTP to forward Mobile Voice Access user input back to
Unified CM (see step 2 in Figure 21-22).
3.
The PSTN gateway forwards DTMF digits in response to user or Smart Phone key sequences from
the remote destination phone for Enterprise Feature Access two-stage dialing and mid-call features
(see step 3 in Figure 21-22).
Figure 21-22
PSTN
Gateway
Cisco Unified
CM
Note
252940
HTTP
While Figure 21-22 depicts the H.323 or SIP VoiceXML gateway as a separate box from the PSTN
gateway, this is not an architectural requirement. Both VoiceXML functionality and PSTN gateway
functionality can be handled by the same box, provided there are no requirements for the PSTN gateway
to run a protocol other than H.323 or SIP. An H.323 or SIP gateway is required for Mobile Voice Access
VoiceXML functionality.
21-54
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
High Availability for Mobile Voice Access and Enterprise Feature Access
The Mobile Voice Access and Enterprise Feature Access features rely on the same components and
redundancy mechanisms as the Single Number Reach feature (see High Availability for Single Number
Reach, page 21-45). Unified CM Groups are necessary for PSTN gateway registration redundancy.
Likewise, PSTN physical gateway and gateway connectivity redundancy should be provided. Redundant
access between the PSTN and the enterprise is required for remote destination phones to access Mobile
Voice Access and Enterprise Feature Access features in the event of a gateway failure. However, while
physical redundancy can and should be provided for the H.323 or SIP VoiceXML gateway, there is no
redundancy mechanism for the Cisco Unified Mobile Voice Access service on Unified CM. This service
can be enabled and run on the publisher node only. Therefore, if the publisher node fails, Mobile Voice
Access functionality will be unavailable. Enterprise Feature Access and two-stage dialing functionality
have no such dependency on the publisher and can therefore provide equivalent functionality to mobility
users (without the IVR prompts).
21-55
Chapter 21
Mobile Collaboration
When configuring the Remote Destination Profile Rerouting CSS, Cisco recommends that the route
patterns within this CSS point to a gateway that is in the same call admission control location as the
gateway used to route the inbound call to the user's desk phone. This ensures that a call admission control
denial due to insufficient bandwidth between two locations will not occur when routing calls out to the
remote destination. Further, because subsequent call admission control checks after the initial Single
Number Reach call is routed will not result in a denial if there is insufficient WAN bandwidth, routing
the inbound and outbound call legs out a gateway or gateways in the same call admission control location
ensures that subsequent desk phone or remote destination pickup operations during this call will not
require call admission control, which could result in WAN bandwidth oversubscription.
When using route patterns pointing to route lists that use Standard Local Route Group, the local route
group configured on the callers device pool will be used. In this case the egress gateway for the call leg
to the remote destination will be local to the original calling device. For calls coming in from the PSTN,
this will help to fulfill the above requirement to use egress gateways in the same call admission control
location as the original caller (in this case the incoming gateway).
Likewise, it is equally important to ensure that call admission control denials are minimized when
placing two-stage dialed calls. Call admission control denials for two-stage dialed calls can be
minimized or avoided by using local route group constructs so that the egress gateway used to route the
outbound call leg is chosen by the ingress gateway of the inbound call leg. With this method, the ingress
and egress gateways used will be in the same call admission control location. Alternatively, the route
patterns within the Remote Destination Profile device-level CCS should point to an egress gateway that
is in the same call admission control location as the ingress gateway that handled the inbound call leg to
the Mobile Voice Access or Enterprise Feature Access system access number. However, be aware that a
subsequent desk phone pickup can result in WAN bandwidth oversubscription if the desk phone is in a
different call admission control location than the gateway through which the Mobile Voice Access or
Enterprise Feature Access system access numbers are reached.
Note
Automatic inbound caller ID matching for configured remote destination numbers is affected by whether
the Matching Caller ID with Remote Destination service parameter is set to Partial or Complete Match.
See Remote Destination Configuration and Caller ID Matching, page 21-52, for more information about
this setting.
In addition to automatic enterprise call anchoring, inbound and outbound call routing must also be
considered when a configured remote destination phone is calling into the enterprise. Inbound call
routing for calls from configured remote destinations occurs in one of two ways, depending on the
setting of the service parameter Inbound Calling Search Space for Remote Destination. By default, this
service parameter is set to Trunk or Gateway Inbound Calling Search Space. With the service
parameter set to the default value, inbound calls from configured remote destinations will be routed
using the Inbound Calling Search Space (CSS) of the PSTN gateway or trunk on which the call is coming
in. If, on the other hand, the parameter Inbound Calling Search Space for Remote Destination is set to
21-56
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
the value Remote Destination Profile + Line Calling Search Space, inbound calls coming from remote
destinations will bypass the Inbound CSS of the PSTN gateway or trunk and will instead be routed using
the associated Remote Destination Profile CSS (in combination with the line-level CSS).
Given the nature of inbound call routing from remote destination phones, it is important to make sure
that calling search spaces are configured appropriately in order to provide access for these inbound calls
to any partitions required for reaching internal enterprise phones, thus ensuring proper call routing from
remote destination phones.
Note
Incoming calls that do not come from a configured remote destination phone are not affected by the
Inbound Calling Search Space for Remote Destination service parameter because they will always use
the trunk or gateway inbound CSS.
Outbound call routing for Mobile Voice Access or Enterprise Feature Access calls always uses a
concatenation of the Remote Destination Profile line CSS and device-level CSS, therefore it is important
to make sure that these calling search spaces are configured appropriately in order to provide access to
any route patterns necessary for off-net or PSTN access, thus ensuring proper outbound call routing from
remote destination phones.
21-57
Chapter 21
Mobile Collaboration
digits 91 in addition to the number they are calling, then rerouting and resulting enterprise call anchoring
will not occur. This is because the user dialed 91 408 555 1234 to reach the remote destination on the
PSTN but the remote destination was configured as 408 555 1234, so there is no match.
For this feature to function properly, matching must occur between the configured remote destination
and the number that must be dialed to reach this remote destination on the PSTN. To ensure that this
matching happens, set the service parameter Matching Caller ID with Remote Destination to Partial
Match. By setting this parameter to Partial Match and then specifying the number of digits to partially
match using the Number of Digits for Caller ID Partial Match service parameter, it is still possible to
match the configured remote destination number with the dialed number even if it contains PSTN
steering digits.
Using the previous example and assuming that system has been set to use partial match on ten digits, the
dialed number 9 1 408 555 1234 can be matched to the configured remote destination 408 555 1234.
This is because, with partial matching, the system attempts to match the same number of digits as
specified by the Number of Digits for Caller ID Partial Match, which in this case is ten digits. The system
attempts to match the two numbers by matching digits from right to left. The last ten digits of the dialed
number 9 1 408 555 1234 are 408 555 1234, and these ten digits match the ten digits of the configured
remote destination (408 555 1234). In this example, the resulting call is anchored in the enterprise and
the called mobile user is able to invoke mid-call features and perform desk phone pickup or session
handoff.
At first glance it might appear that an easier way to handle this feature would be to configure remote
destination or mobility identity numbers that include any required PSTN steering digits. However, when
configuring these numbers with required PSTN steering digits, if you do not also configure partial caller
ID matching, the system will not be able to perform automatic caller ID matching and enterprise
anchoring for inbound calls from configured remote destinations or mobility identities. In the previous
example, if the remote destination number had been configured as 9 1 408 555 1234 and complete caller
ID matching had been used, an inbound call from the remote destination would present caller ID of
408 555 1234 and a match would not occur, meaning the inbound call from the remote destination would
not be anchored as expected.
Based on this potential for mismatch between dialed numbers for outbound calls and configured remote
destination numbers for inbound calls, Cisco recommends enabling partial (rather than complete) caller
ID matching when using the Intelligent Session Control feature for all deployments that require one or
more steering digits to reach the PSTN. This ensures that calls made directly to the remote destination
number using PSTN steering digits are still matched and anchored. On the other hand, if steering digits
are not required to reach the PSTN and users are able to dial the full E.164 number to route calls to the
PSTN, then Cisco recommends the complete caller ID matching setting because the remote destination
is configured to match the caller ID and is the same number as dialed by internal users to reach the
remote destination or mobility identity on the PSTN.
When enabling the Intelligent Session Control feature, it is also important to understand the behavior of
the enterprise and remote destination lines during the reroute feature operation. On call reroute, remote
destination line settings Do Not Disturb (DND), Access Lists and Time of Day call filtering, and the
Delay Before Ringing Timer are ignored. All reroute calls are routed unfiltered and immediately.
Enterprise desk phone line settings are also ignored or bypassed by default. However, Call Forward All
settings on the enterprise desk phone line can be honored during reroute feature operation by setting the
Ignore Call Forward All on Enterprise DN service parameter to False. If this parameter is set to False,
on reroute operation, calls will not be routed to the remote destination if the enterprise desk phone line
has a call-forward-all destination set. Instead, the call will be routed to the call-forward-all destination.
By default, this service parameter is set to True, and call-forward-all settings on enterprise desk phone
lines are ignored.
21-58
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Intelligent Session Control functionality may be further enhanced by using the Ring All Shared Lines
feature. This feature is enabled by setting the Ring All Shared Lines service parameter to True. By
default, this service parameter is set to True and the feature is enabled. However, the Ring All Shared
Lines feature is dependent on the Intelligent Session Control feature, which must also be enabled in
order use the Ring All Shared Line functionality. When the Ring All Shared Lines and Intelligent
Session Control features are both enabled, not only will the system route internally originated calls to
the dialed remote destination by way of the PSTN, but all of the user's other shared-line devices will also
receive the call. This includes the user's enterprise desk phone as well as other configured remote
destinations. The called user will then be able to answer the incoming call on any of their devices and
the call will be anchored in the enterprise.
Note
If Ring All Shared Lines is enabled, mobile client devices will not receive calls at the cellular voice
interface of the device when the device is registered to Unified CM.
Caller ID Transformations
Any calls made into the cluster by configured remote destination numbers will automatically have their
caller ID or calling number changed from the calling remote destination phone number to the enterprise
directory number of the associated desk phone. For example, if a remote destination phone with number
408 555-7890 has been configured and associated to a user's enterprise desk phone with number
555-1234, then any call from the user's remote destination phone destined for any directory number in
the cluster will automatically have the caller ID changed from the remote destination number of 408
555-7890 to the enterprise directory number of 555-1234. This ensures that the active call caller ID
display and call history log caller ID reflect a mobility user's enterprise desk phone number rather than
their mobile phone number, and it ensures that any return calls are made to the user's enterprise number,
thus anchoring those calls within the enterprise.
Likewise, calls from a remote destination phone to external PSTN destinations and anchored in the
enterprise via Mobile Voice Access or Enterprise Feature Access two-stage dialing, or those calls forked
to the PSTN as a result of Single Number Reach, will also have caller ID changed from the calling remote
destination phone number to the associated enterprise directory number.
Finally, in order to deliver the calling party number as an enterprise DID number rather than an enterprise
directory number to external PSTN phones, calling party transformation patterns can be used. By using
calling party transformation patterns to transform caller IDs from enterprise directory numbers to
enterprise DIDs, return calls from external destinations will be anchored within the enterprise because
they will be dialed using the full enterprise DID number. For more information about these
transformations and dial plan implications, see Special Considerations for Cisco Unified Mobility,
page 14-85.
21-59
Chapter 21
Mobile Collaboration
The Cisco Unified Mobility solution is verified with only Cisco equipment. This solution may also work
with other third-party PSTN gateways and Session Border Controllers (SBCs), but each Cisco Mobility
feature is not guaranteed to work as expected. If you are using this solution with third-party PSTN
gateways or SBCs, Cisco technical support may not be able to resolve problems that you encounter.
The following guidelines and restrictions apply with regard to deployment and operation of Single
Number Reach within the Unified CM telephony environment:
Single Number Reach is supported only with PRI TDM PSTN connections. T1 or E1-CAS, FXO,
FXS, and BRI PSTN connections are not supported. This PRI requirement is based on the fact that
Cisco Unified CM must receive expeditious answer and disconnect indication from the PSTN in
order to ensure full feature support. Answer indication is needed in order for Cisco Unified CM to
stop ringing the desk phone and other remote destinations when a Single Number Reach call is
answered at a particular remote destination. In addition, answer indication is required in order to
support the single enterprise voicemail box feature. Finally, disconnect indication is required for
desk phone pickup. A PRI PSTN connection will always provide answer or disconnect indication.
Single Number Reach is also supported over SIP trunk VoIP PSTN connections. Use of Cisco IOS
Unified Border Element is recommended as the demarcation point between the Unified CM SIP
trunk and the service provider trunk. A VoIP-based PSTN connection is still able to provide
expeditious answer and disconnect indication to Unified CM due to the end-to-end signaling path
provided by VoIP-based PSTN connections.
Single Number Reach can support up to two simultaneous calls per user. Any additional calls that
come in are automatically transferred to the users voicemail.
Single Number Reach does not work with Multilevel Precedence and Preemption (MLPP). If a call
is preempted with MLPP, Single Number Reach features are disabled for that call.
Single Number Reach services do not extend to video calls. A video call received at the desktop
phone cannot be picked up on the cellular phone.
Remote destinations must be Time Division Multiplex (TDM) devices or off-system IP phones on
other clusters or systems. You cannot configure IP phones within the same Unified CM cluster as
remote destinations.
For additional guidelines and restrictions, refer to the information on Cisco Unified Mobility in the latest
version of the Feature Configuration Guide for Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
Note
A mobility-enabled user is defined as a user that has a remote destination profile and at least one remote
destination or a mobile client device and a mobility identity configured.
21-60
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Note
A mobility identity is configured just like a remote destination within the system, and it has the same
capacity implications as a remote destination. Unlike a remote destination, however, the mobility
identity is associated directly to a phone device rather than a remote destination profile. The mobility
identity applies only to dual-mode mobile client devices running Cisco Jabber.
Scalability and performance of Cisco Unified Mobility ultimately depends on the number of mobility
users, the number of remote destinations or mobility identities each user has, and the busy hour call
attempt (BHCA) rates of those users. Multiple remote destinations per user and/or high BHCA per user
can result in lower capacity for Cisco Unified Mobility. For more information on Cisco Unified Mobility
sizing, including Unified CM server node capacities and hardware specific per-node and per-cluster
capacities, see the chapter on Collaboration Solution Sizing Guidance, page 25-1.
Ensure that the PSTN gateway protocol is capable of out-of-band DTMF relay or allocate media
termination points (MTPs) in order to covert in-band DTMF to out-of-band DTMF. When using
Cisco IOS gateways for PSTN connectivity, out-of-band DTMF relay will be supported. However,
third-party gateways might not support a common out-of-band DTMF method, and as a result an
MTP might be required. In order to use Enterprise Feature Access Two-Stage Dialing and mid-call
features, DTMF digits must be received out-of-band by Cisco Unified CM.
Note
When relying on MTP for converting in-band DTMF to out-of-band DTMF, be sure to
provide sufficient MTP capacity. If heavy or frequent use of Enterprise Feature Access
Two-Stage Dialing or mid-call features is anticipated, Cisco recommends a hardware-based
MTP or Cisco IOS software-based MTP.
Prior to deploying Unified Mobility, it is important to work with the PSTN provider to ensure the
following:
Caller ID is provided by the service provider for all inbound calls to the enterprise. This is a
expectation that mobility-enabled users will receive the caller ID of the original caller at their
remote destination rather than a general enterprise system number or other non-meaningful
caller ID.
Note
Some providers restrict outbound caller ID on a trunk to only those DIDs handled by
that trunk. For this reason, a second PRI trunk that does not restrict caller ID might have
to be acquired from the provider. To obtain an unrestricted PRI trunk, some providers
might require a signed agreement from the customer indicating they will not send or
make calls to emergency numbers over this trunk.
21-61
Chapter 21
Mobile Collaboration
Note
Because mobility call flows typically involve multiple PSTN call legs, planning and allocation of
PSTN gateway resources is extremely important for Unified Mobility. In cases where there are large
numbers of mobility-enabled users, PSTN gateway resources will have to be increased. The
following methods are recommend to minimize or reduce PSTN utilization:
Limit the number of remote destinations per mobility-enabled user to one (1). This will reduce
the number of DS0s that are needed to extend the inbound call to the user's remote destination.
One DS0 is consumed for each configured remote destination when a call comes into the user's
enterprise directory number, even if the call is not answered at one of the remote destinations.
Note that a DS0 per remote destination may be used for as long as 10 seconds, even if the call
is not answered at the remote destination.
Use access lists to block or restrict the extension of calls to a particular remote destination based
on incoming caller ID. Because access lists can be invoked based on the time of day, this
eliminates the need for repeated updates of access lists by the end-user or the administrator.
Educate end-users to disable Single Number Reach when not needed, to further eliminate DS0
utilization when a call comes in for that user's enterprise number. If Single Number Reach is
disabled, incoming calls will still ring the desk phone and will still forward to enterprise
voicemail if the call goes unanswered.
Due to the potential for call admission control denials resulting from insufficient WAN bandwidth
between locations and the possibility that a desk phone pickup or remote destination pickup might
result in WAN bandwidth over-subscription, Cisco recommends configuring Remote Destination
Profile CSS and Rerouting CSS so that route patterns within these CSSs point to gateways that are
located within the same call admission control location as the gateway on which the inbound call
leg comes in. For more information, see Remote Destination Profile Configuration, page 21-55.
If you enable the Intelligent Session Control feature in deployments where PSTN steering digits
must be dialed to access the PSTN, Cisco recommends setting the Matching Caller ID with Remote
Destination service parameter to Partial Match and configuring the appropriate number of digits
(Number of Digits for Caller ID Partial Match service parameter) to achieve a partial match of
configured remote destinations or mobility identities. This will ensure proper functioning of the
Intelligent Session Control feature and the mobility automatic caller ID matching and anchoring
features.
21-62
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Cisco Jabber A mobile client available for Android and Apple iOS mobile devices, including
iPhone and iPad, providing the ability to make voice and/or video calls over IP on the enterprise
WLAN network or over the mobile data network as well as the ability to access the corporate
directory and enterprise voicemail services and XMPP-based enterprise IM and presence.
Cisco Spark A mobile client available for Android and Apple iOS devices, including iPhone and
iPad, providing 1-to-1 and 1-to-many cloud-based collaboration rooms enabling voice and/or video
calls over IP, secure persistent messaging, and file sharing.
Cisco WebEx Meetings A mobile client available for Android, BlackBerry, Windows Mobile, and
Apple iOS devices including iPhone and iPad, enabling users to attend and participate in Cisco
WebEx meetings while mobile.
Cisco AnyConnect Mobile A mobile client available for Android and Apple iOS devices,
enabling secure remote VPN connectivity to the enterprise for access to on-premises collaboration
applications and services even when the user is outside of the enterprise.
In addition, this section discusses high availability and capacity planning considerations for Cisco
mobile clients and devices.
21-63
Chapter 21
Mobile Collaboration
Note
The use of the term dual-mode phone in this section refers specifically to devices with 802.11 radios in
addition to the cellular radio for carrier voice and data network connectivity. Dual-mode devices that
provide Digital Enhanced Cordless Telecommunications (DECT) or other wireless radios and/or
multiple cellular radios are outside the scope of this section.
Figure 21-23 depicts the basic Cisco mobile clients and devices solution architecture for connecting and
enabling mobile client devices for Cisco Collaboration deployments. For voice and video services,
mobile client devices associate to the enterprise WLAN or connect over the Internet (from a public or
private WLAN hot spot or the mobile data network), and the Cisco mobile client registers to Cisco
Unified CM as an enterprise phone using the Session Initiation Protocol (SIP). Once registered, the
client device relies on the underlying enterprise Cisco IP telephony network for making and receiving
calls. When the mobile device is connected to the enterprise network and the client is registered to
Unified CM, the device is reachable through the user's enterprise number. Any inbound calls to the user's
enterprise number will ring the mobile client device. If the user has a Cisco IP desk phone, then the
mobile client registration enables a shared line instance for the user's enterprise number so that an
incoming call rings both the user's desk phone and the mobile device. When unregistered, the mobile
client device will not receive incoming enterprise calls unless the mobile device has an active cellular
voice radio, the user has been enabled for Cisco Unified Mobility, and Single Number Reach has been
turned on for the user's mobile phone number. In these scenarios the mobile voice network and PSTN
are used for making and receiving voice-only calls.
Unified Mobility features such as Single Number Reach are not compatible with tablets and other mobile
client devices that do not have cellular voice radios because these non-dual-mode devices do not have a
native PSTN reachable number. Non-dual-mode devices are able to make and receive enterprise calls
only when connected to the enterprise and registered to the enterprise call control system.
As shown in Figure 21-23, when attached to the enterprise, Cisco mobile clients and devices can also
communicate directly with other back-end application servers such as the corporate directory, Cisco
Unity Connection enterprise voicemail system, and the Cisco IM and Presence Service for access to
additional enterprise collaboration services such as messaging and presence. Cisco mobile clients and
devices also integrate with cloud-based collaboration services such as Cisco WebEx, which delivers IM
and presence and web conferencing services.
21-64
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
348816
Figure 21-23
Mobile
Voice
PSTN
PSTN
P
Ga
Gateway
Enterprise
WLAN
(802.11)
DMZ
MZ
Cisco Unified CM
Cisco ASA
Public/Private
WLAN
(802.11)
Cisco IM &
Presence
Mobile
Data
Internet
Cisco
W bE
WebEx
Cisco
Expressway-E
Cisco
Expressway-C
Spark
Cloud-based Collaboration
Collaboratio
S
i
Services
Note
Cisco Unity
Connection
Cisco Expressway
Corporate
Directory
On-Premises Enterprise
and Collaboration
Cisco Internal
Applications
The voice and video quality of calls will vary depending on the Wi-Fi or mobile data network connection.
Cisco Technical Assistance Center (TAC) is not able to troubleshoot connectivity or voice and video
quality issues over 3G/4G mobile data networks or non-corporate Wi-Fi networks.
Dual-mode mobile client devices must be capable of dual transfer mode (DTM) in order to be connected
simultaneously to both the mobile voice and data network and the WLAN network. This allows the
device to be reachable and able to make and receive calls on both the cellular radio and WLAN interface
of the device. In some cases proper mobile client operation might not be possible if mobile voice and
data networks do not support dual-connected devices.
21-65
Chapter 21
Mobile Collaboration
devices rely on the underlying WLAN infrastructure for carrying both critical signaling and other traffic
for setting up calls and accessing various applications as well as the real-time voice and video media
traffic, deploying a WLAN network optimized for both data and real-time media traffic is necessary. A
poorly deployed WLAN network will be subjected to large amounts of interference and diminished
capacity, leading not only to poor voice and video quality but in some cases dropped or missed calls.
This will in turn render the WLAN deployment unusable for making and receiving calls. Therefore,
when deploying dual-mode phones and other mobile devices, it is imperative to conduct a WLAN radio
frequency (RF) site survey before, during, and after the deployment to determine appropriate cell
boundaries, configuration and feature settings, capacity, and redundancy to ensure a successful
deployment of voice and video over WLAN. Each mobile device type and/or client should be tested on
the WLAN deployment to ensure proper integration and operation prior to a production deployment.
Using a WLAN that has been deployed and configured to provide optimized real-time traffic over
WLAN services (such as the Cisco Unified Wireless Network), including quality of service, will ensure
a successful mobile client device deployment.
Cisco recommends relying on the 5 GHz WLAN band (802.11a/n/ac) whenever possible for connecting
mobile clients and devices capable of generating voice and video traffic. 5 GHz WLANs provide better
throughput and less interference for voice and video calls.
For more information on voice and video over WLAN deployments and wireless device roaming, see
Wireless Device Roaming, page 21-5.
Note
While dual-mode phones and other mobile client devices are capable of connecting back to the enterprise
through the Internet for call control and other Unified Communications services, Cisco cannot guarantee
voice and video quality or troubleshoot connectivity or voice and video quality issues for these types of
connections. These types of connections include remote connections to the enterprise through public or
private WLAN access points (APs) or hot spots or through the mobile data network. Cisco recommends
an enterprise class voice and video-optimized WLAN network for connecting dual-mode phones and
other mobile client devices. Most public and private WLAN APs and hot spots are tuned for data
applications and devices. In these cases, the AP radios are turned to maximum power, and
dynamic-power control results in devices enabling maximum power on network attachment, which
allows for larger client capacities. While this may be ideal for data applications that are capable of
retransmitting dropped or lost packets, for real-time traffic applications this can result in poor voice and
video quality due to the potential for large numbers of dropped packets. Likewise, mobile provider data
networks are susceptible to congestion and/or dropped connections, which can also result in poor call
quality and dropped calls.
WebEx Meetings, which provides web-enabled voice and video conferencing with content sharing.
WebEx Messenger, which provides XMPP IM and presence as well as point-to-point audio and
video calling.
Spark, which provides 1-to-1 and 1-to-many collaboration rooms with voice and video calling,
messaging, and file sharing.
21-66
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Layer 3 Marking
Traffic Type
DSCP1
PHB2
DSCP 46
PHB EF
DSCP 34
PHB AF41
Call Signaling
DSCP 24
PHB CS3
Cisco mobile client Layer 2 802.11 WLAN packet marking (User Priority, or UP) presents challenges
given the various mobile platform and firmware restrictions. Because Cisco mobile clients run on a
variety of mobile devices, Layer 2 wireless QoS marking is inconsistent and therefore Layer 2 wireless
QoS marking cannot be relied on to provide appropriate treatment to traffic on the WLAN.
Despite appropriate mobile client application Layer 3 or even Layer 2 packet marking, mobile devices
present many of the same challenges as desktop PCs in terms of generating many different types of
traffic, including both data and real-time traffic. Given this, mobile devices generally fall into the
untrusted category of collaboration endpoints. For deployments where mobile client devices are not
considered trusted endpoints, packet marking or re-marking based on traffic type and port numbers is
required to ensure that network priority queuing and dedicated bandwidth is applied to appropriate
traffic. In addition to re-marking the mobile device traffic, Cisco recommends using network-based
policing and rate limiting to ensure that the mobile client devices do not consume too much network
bandwidth.
Alternatively, given appropriate Cisco mobile client Layer 3 marking and assuming mobile client
devices are trusted, Cisco mobile client traffic will be queued appropriately as it traverses the enterprise
network by using priority voice queuing and dedicated video media and call signaling bandwidth queues.
Because Cisco mobile clients and devices are capable of making and receiving calls using the enterprise
telephony infrastructure and call control services, it is important to understand the nature and behavior
of call routing as it pertains to mobile client devices.
Inbound Call Routing
When mobile clients and devices register to Unified CM as an enterprise device with enterprise number,
the mobile device rings when incoming calls to the system are destined for the user's enterprise number.
This occurs for incoming calls originated on the PSTN or from other Unified CM clusters or enterprise
IP telephony systems as well as for incoming calls originated within the Unified CM cluster by other
21-67
Chapter 21
Mobile Collaboration
users. If the mobile client device user has other devices or clients that are also associated to the enterprise
number, these devices will also ring as shared lines; and once the call is answered at one of the devices
or clients, ringing of all other devices and clients ceases.
In scenarios where a user has been enabled for Cisco Unified Mobility, and when Single Number Reach
is enabled for the user's dual-mode mobile phone number, the incoming call may also be extended to the
mobility identity corresponding to the user's mobile phone number. However, this depends on whether
the mobile device is connected to the enterprise WLAN network or attached to the enterprise network
through a secure connection and registered to Unified CM. In situations in which the device is connected
to the enterprise network directly or through a secure remote connection, an incoming call to the user's
enterprise number will not be extended by Single Number Reach to the mobility identity of the mobile
device even if Single Number Reach is enabled on for this mobile number. The reason an incoming call
to the enterprise number is not extended to the mobility identity of a dual-mode mobile device when it
is registered to Unified CM is that the system is aware the device is connected to the enterprise network
and available. Thus, in order to reduce utilization of enterprise PSTN resources, Unified CM does not
extend the call to the dual-mode mobile phone's mobile voice network interface through the PSTN.
Instead, only the WLAN or mobile data network interface corresponding to the enterprise number
receives the call.
Note
In cases where dial via office is enabled (see Dial Via Office, page 21-73), even if the client is registered,
Unified CM will extend inbound calls to the user's mobile number using Single Number Reach rather
than via VoIP to the enterprise number.
For situations in which the mobile device is not connected to the enterprise network directly or through
a secure remote connection or is not registered to Unified CM, incoming calls to the enterprise number
will be extended to the dual-mode mobile phone number per the configured mobility identity, assuming
that the user has been enabled for Unified Mobility and that Single Number Reach for the mobility
identity is turned on. For more information on integration of mobile clients and devices with Unified
Mobility, see Interactions Between Cisco Jabber and Cisco Unified Mobility, page 21-92.
The same behavior and logic described above also applies with the Ring All Shared Lines feature. If this
feature is enabled, calls are extended to the mobility identity or cellular number only when the
dual-mode mobile client device is not registered to Unified CM. For more information on the Ring All
Share Line feature, see Intelligent Session Control and Ring All Shared Lines, page 21-57.
In all cases, incoming calls made directly to the dual-mode device's mobile network phone number will
always be routed directly to the mobile voice interface of the dual-mode device on the mobile network,
unless the provider network or device settings are such that calls are not extended to the device by the
mobile network. This is considered appropriate behavior because these calls were not made to the user's
enterprise number. These would be considered personal calls, and as such should not be routed through
the enterprise.
Note
Mobile client devices that do not have cellular voice radios, such as tablet devices, are not dual-mode
devices and as such cannot be reached on a mobile voice network interface. These devices can be reached
only at the enterprise number by voice-over-IP.
Outbound Call Routing
For outbound calls from the dual-mode mobile device, the interface used depends on the location and
connectivity of the device at that particular time. If the dual-mode device is not connected to the
enterprise and not registered to Unified CM, then calls are routed by the cellular voice radio interface to
the mobile voice network as usual. However, when connected to the enterprise and registered to
Unified CM, the mobile device should make all calls through the enterprise telephony infrastructure. If
21-68
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
no enterprise connectivity is available or the mobile client is unregistered, then outbound calling is not
possible from the enterprise number, and instead calls would have to use the mobile number of the
mobile client device for making calls over the mobile voice network. Alternatively, users may use the
two-stage dialing features provided with Cisco Unified Mobility (see Mobile Voice Access and
Enterprise Feature Access, page 21-46).
Dial Plan
The enterprise dial plan determines the dialing behavior of the mobile client device when it is connected
to the enterprise and registered to Unified CM. For example, if the enterprise dial plan is configured to
allow abbreviated dialing to reach internal extensions, then a mobile device registered to Unified CM
can use this abbreviated dialing. While it is certainly a convenience for dual-mode mobile phone users
to be able to dial within the enterprise using enterprise dialing habits and abbreviated dialing as well as
site-based and/or PSTN steering digits for outbound calls, it is also a somewhat unnatural dialing scheme
because mobile phone users typically dial numbers for outgoing calls on their mobile phone by using
full E.164 dial strings since this is what is expected by the mobile voice network for outbound calling.
The enterprise dialing experience for an end-user is ultimately up to the enterprise policies and
administrator of the enterprise telephony deployment. However, for dual-mode mobile devices, Cisco
recommends normalizing required dialing strings for dual-mode client devices so that user dialing habits
are maintained whether the device is connected to the enterprise network and registered to Unified CM
or not. Because dialing on the mobile voice network is typically done using full +E.164 (with a
preceding '+') and mobile phone contacts are typically stored with full +E.164 numbers, Cisco
recommends configuring the enterprise dial plan to accommodate full +E.164 with preceding '+' for
dual-mode mobile devices. When the dial plan is configured within Unified CM to handle this type of
outbound dialing for dual-mode phones, it is possible for users to store a single set of contacts on the
phone in the +E.164 format and, when dialed from these contacts or manually using the full +E.164
number, calls will always be routed to the appropriate destination, whether the device is connected to the
enterprise network directly or over secure remote connection and registered to Unified CM or connected
only to the mobile voice network. Configuring the enterprise dial plan in this manner provides the best
possible end-user dialing experience so that users' mobile device dialing habits are maintained and they
do not have to be aware of whether the device has enterprise connectivity and is registered to
Unified CM.
To achieve normalized dialing from dual-mode phones, whether connected to the enterprise or just the
mobile voice network, configure the dial plan within Unified CM with the following considerations in
mind:
Ensure that the enterprise dial plan is capable of handling dial strings from dual-mode phones
typically used on the mobile voice network. For example, the dial plan should be configured to
handle the following strings, which might be dialed from a mobile phone to reach a particular phone
through the mobile voice network: +1 408 555 1234, 408 555 1234. Supporting the latter 10-digit
dialing method (for example, 408 555 1234) might potentially overlap with other dialing habits such
as abbreviated intra-site dialing. In that case the administrator has to decide which of the colliding
dialing habits (10-digit dialing or abbreviated intra-site) should be available for dual-mode phones
registered to the enterprise network. The set of dialing habits supported on dual-mode phones often
differs from the set of dialing habits supported on regular endpoints.
For calls destined for other enterprise numbers, systems configured for abbreviated dialing should
be capable of modifying dial strings and rerouting to enterprise extensions as appropriate. For
example, assuming the enterprise dial plan is based on five-digit internal dialing, the system should
be configured to handle call routing to an enterprise extension so that a call to made to +1 408 555
1234 or 408 555 1234 is modified and rerouted to 51234 if the call is made while the dual-mode
device is registered to Unified CM.
21-69
Chapter 21
Mobile Collaboration
Ensure that all inbound calls to the enterprise destined for dual-mode devices have the calling
number and/or caller ID prefixed with appropriate digits so that missed, placed, and received call
history lists are in full +E.164 formats. This will allow dual-mode device users to dial from call
history lists without the need for editing the dial string. Instead, users will be able to select a number
from the call history list to redial, whether connected to the enterprise or not. For example, if an
incoming call from inside the enterprise originates from 51234 to a dual-mode user's enterprise
number and the call goes unanswered, Unified CM should be configured to manipulate the calling
number so that the resulting entry within the history list of the dual-mode device shows either
408 555 1234 or +1 408 555 1234. This number can be dialed whether the dual-mode device is
connected to the enterprise or just to the mobile voice network without the need for further
manipulation.
The one exception to normalized dialing for dual-mode devices is for scenarios in which some enterprise
extensions or phones are reachable only internally (that is, they have no externally reachable
corresponding DID number). In these situations, non-externally reachable numbers can be dialed
(manually or from contacts) using abbreviated formats. Because these numbers will never be available
externally and can be dialed only from inside the enterprise, some sort of enterprise-only indication
should be made when storing these numbers in the contact list. Further, incoming calls from these
internal-only numbers should not have the calling number modified for call history lists because these
numbers may be called only inside the enterprise. Instead, calls from these extensions should be listed
in all call history lists without modification so that the abbreviated dial strings can be successfully dialed
only while the device is connected to the enterprise and registered to Unified CM.
Mobile client devices that do not have cellular voice radios, such as tablets, are dependent exclusively
on enterprise connectivity and enterprise voice and video telephony or cloud-based collaboration
services.
Emergency Services and Dialing Considerations
Mobile client devices do present a slight challenge when it comes to making calls to emergency service
numbers such as 911, 999, and 112. Because the mobile client devices may be located inside or outside
the enterprise, providing location indication of a device and its user in the event of an emergency must
be considered. Dual-mode mobile devices with cellular voice radios receive location services from their
provider networks, and these location services are always available when the device is connected and
typically able to pinpoint locations far more precisely than enterprise wireless networks; therefore, Cisco
recommends that dual-mode device users rely on the mobile voice network for making emergency calls
and determining device and user location. To ensure that Cisco dual-mode client devices rely exclusively
on the mobile provider voice network for emergency and location services, these clients force all calls
made to numbers configured in the Emergency Numbers field on the mobile client device configuration
page to route over the mobile voice network. Further, dual-mode phone users should be advised to make
all emergency calls over the mobile voice network rather than the enterprise network.
While making emergency calls over WLANs or mobile data networks is not recommended, mobile
devices that do not have cellular voice radios are capable of making calls only through these data
interfaces. Mobile devices that do not have cellular voice radios should not be relied upon for making
emergency calls.
Enterprise Caller ID
When mobile client devices are connected to the enterprise and registered to Unified CM (either through
the mobile data network or a WLAN), all calls made with the enterprise line over the WLAN or mobile
data network will be routed with the user's enterprise number as caller ID. This ensures that returned
calls made from call history lists at the far-end are always routed through the enterprise because the
return call is to the user's enterprise number. If a dual-mode mobile device user has been enabled for
21-70
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Cisco Unified Mobility, and Single Number Reach is turned on for the mobile phone number, return calls
to the enterprise number would also be extended to the dual-mode device through the PSTN whenever
it is not connected to the enterprise.
Mid-Call Features
When mobile client devices are connected to the enterprise and registered to Unified CM as enterprise
endpoints, they are able to invoke call processing supplementary services such as hold, resume, transfer,
and conference, using SIP call signaling methods as supported by Unified CM. Just as with any IP phone
or client registered to Unified CM, these devices are able to leverage enterprise media resources such as
music on hold (MoH), conference bridges, media termination points, and transcoders.
External Call Routing
When dual-mode mobile client devices are not connected to the enterprise and/or not registered to
Unified CM, they may make and receive calls only over the mobile voice network. For this reason,
Unified CM has no visibility into any calls being made or received at the dual-mode mobile device while
it is unregistered. The mobile number is the caller ID being sent to the network when calls are made from
dual-mode phones not connected to the enterprise. This will likely result in unanswered calls being made
directly back to the dual-mode device's mobile number instead of being routed back through the
enterprise.
If the dual-mode mobile client device is integrated with Cisco Unified Mobility, enterprise two-stage
dialing services may be leveraged for making calls through the enterprise network even when the
dual-mode device is outside the enterprise and not registered to Unified CM. Unified Mobility two-stage
dialing is done using either Mobile Voice Access or Enterprise Feature Access and requires the user to
dial an enterprise system access DID number and enter credentials prior to dialing the number they are
calling. For more information on Unified Mobility two-stage dialing features, see Mobile Voice Access
and Enterprise Feature Access, page 21-46.
Likewise, if the dual-mode phone is integrated with Unified Mobility, a user can also receive incoming
calls to their enterprise number at the mobile number through Single Number Reach; can invoke mid-call
features using DTMF key sequences including hold, resume, transfer, and conference; and can perform
desk phone pickup to move an active call from the mobile phone to the enterprise desk phone.
Remote Secure Enterprise Connectivity
Mobile client devices can utilize the IP telephony infrastructure for enterprise voice and video over IP
calling and other collaboration services, even when not inside the enterprise, provided they have a secure
connection back to the enterprise in order to register the client with Unified CM and to access other
collaboration applications and services. Remote secure connectivity for these devices requires the use of
the Cisco AnyConnect mobile client VPN solution or the VPN-less Cisco Expressway mobile and
remote access feature in order to secure the client connection over the Internet.
Voice and video quality and user experience for remotely attached mobile client devices will vary
depending on the nature of the Internet-based network connection. Cisco cannot guarantee acceptable
voice and video quality nor successful connectivity for these types of client connections. Care should be
taken when relying on these types of connections for business-critical communications. In the case of
dual-mode devices with unreliable or low-bandwidth Internet connections, users with dual-mode devices
should be advised to make calls over the mobile voice network if connectivity is available rather than
relying on the remote enterprise telephony infrastructure.
Additional Services and Features
In addition to call processing or call control services, Cisco mobile clients and devices are capable of
providing the additional features and services described in this section.
21-71
Chapter 21
Mobile Collaboration
One very important aspect of dual-mode device deployments is call preservation as a user moves in and
out of the enterprise or as the device connects to and disconnects from the enterprise network and
network connectivity changes from the cellular voice radio to the WLAN radio, and vice versa. Because
dual-mode phone users are often mobile, it is important to maintain any active call as a dual-mode user
moves in or out of the enterprise. For this reason, dual-mode client devices and the underlying enterprise
telephony network should be capable of some form of call handoff.
There are two types of call handoff that should be accommodated by both the dual-mode client and the
underlying IP telephony infrastructure:
Hand-out
Call hand-out refers to the movement of an active call from the WLAN or mobile data network
interface of the dual-mode phone to the cellular voice interface of the dual-mode phone. This
requires the call to be handed out from the enterprise IP network to the mobile voice network
through the enterprise PSTN gateway.
Hand-in
Call hand-in refers to the movement of an active call from the cellular voice interface of the
dual-mode phone to the WLAN or mobile data network interface of the dual-mode phone. This
requires the call to be handed in from the mobile voice network to the enterprise IP network through
the enterprise PSTN gateway.
The handoff behavior of a dual-mode phone depends on the nature of the dual-mode client and its
particular capabilities. Dual-mode client handoff may be invoked manually by the user or automatically
based on network conditions. In manual handoff scenarios, the dual-mode users are responsible for
engaging and completing the handoff operation based on their location and needs. With automatic
handoff, the mobile client monitors the WLAN signal and makes handoff decision based on
strengthening or weakening of the WLAN signal at the client. Hand-out is engaged in the case of a
weakening WLAN signal, while hand-in is engaged in the case of a strengthening WLAN signal.
Automatic handoff depends on the mobile device to provide capabilities for monitoring WLAN signal
strength.
Handoff operations are critical for maximizing utilization of the enterprise IP telephony infrastructure
for phone calls. These operations are also necessary for providing voice continuity and good user
experience so that users do not have to hang up the original call and make another call to replace it.
XMPP-Based IM and Presence
Some mobile clients are capable of providing enterprise instant messaging (IM) and presence services
based on the Extensible Messaging and Presence Protocol (XMPP), through integration to an
on-premises or off-premises application server or service. In either case, the IM and presence
capabilities of these mobile clients enable the following:
While IM and presence are not required functionality for mobile clients, they do enable users to make
their availability status visible to contacts and to view the availability status of contacts, thus improving
productivity. Further, users can send enterprise-based IM messages rather than incurring costs for mobile
Short Message Service (SMS) messages.
21-72
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Mobile clients and devices are capable of accessing the enterprise directory for contact lookups.
Enterprise directory access is enabled using either:
Note
Lightweight Directory Access Protocol (LDAP) for communication between the clients and a
compatible LDAP directory
REST-based (HTTPS) communications between the clients and the User Data Services (UDS) API,
which provides a set of operations that enable authenticated access to user contact information
stored within the end-user database of the Unified CM cluster
Except in the case of very small deployments, LDAP-based directory access is recommended over
UDS-based access due to the fact that UDS-based directory access greatly reduces Unified CM node
endpoint capacity.
While corporate directory access is not a required feature for mobile devices and clients, it does provide
productivity improvements for mobile users when they are able to access corporate directory information
from their mobile device.
Enterprise Voicemail Services
Many mobile clients and devices are also capable of accessing enterprise voicemail services. Cisco
mobile clients are capable of receiving enterprise message waiting indication whenever an unread
voicemail is in the user's enterprise voicemail box and the mobile device is attached to the enterprise
network. Further, mobile clients can be used to retrieve enterprise voicemail messages. Typically
enterprise voicemail messages are retrieved when the user dials the voicemail system number and
navigates to their voicemail box after providing required credentials. However, Cisco Jabber mobile
clients provide the ability to retrieve voicemail messages from the voicemail box by downloading and
displaying a list of all messages in the voicemail box and then by selecting individual messages to be
downloaded to the mobile device for listening. This is sometimes referred to as visual voicemail. Both
the mobile client and the enterprise voicemail system must be capable of providing and receiving
message waiting indication (MWI), voicemail message information, and downloads of the messages
over the network. Cisco Unity Connection supports visual voicemail through REST (HTTPS) and
provides MWI, voicemail lists, and message downloads.
Dial Via Office
Dial via office (DVO) functionality provides automated enterprise dialing capabilities that enable
dual-mode mobile devices to initiate calls through the enterprise telephony infrastructure. Deploying
DVO calling provides the following benefits to the enterprise:
Cost savings for calls to international and (possibly) long distance destinations as compared to
direct-dialed cellular calls. Note that, in cases of mobile data traversal, mobile data costs must also
be considered.
Ability to dial internal enterprise numbers. Because DVO calls are made using the enterprise line,
non-DID or internal-only enterprise extensions are reachable.
Mobile phone number masking. For DVO calls, the system sends the user's enterprise number as
caller ID, and not the mobile phone number.
Centralized enterprise call detail records (CDRs) and call logs. Because DVO calls are made through
the enterprise telephony infrastructure, administrators have complete visibility to these calls even
though they traverse the PSTN and mobile voice network.
Enterprise call anchoring. DVO calls are anchored in the enterprise, thus enabling users to leverage
Cisco Unified Mobility DTMF-based mid-call features and desk phone pickup.
21-73
Chapter 21
Mobile Collaboration
Dual-mode mobile devices running the Cisco Jabber client are able to make DVO calls using the
Unified CM telephony infrastructure and enterprise PSTN gateway to make calls using the enterprise
line. However, unlike voice over IP (VoIP) calling where voice media traverses the IP network, this
functionality is facilitated by SIP signaling between the client and Unified CM over an IP connection
(WLAN or mobile data) and voice media between the mobile device and the mobile voice network and
PSTN, as shown in Figure 21-24.
Figure 21-24
PSTN
Gateway
PSTN
Mobile Voice
Network
Call Media
(Voice)
Cisco Unified CM
Enterprise
WLAN
(802.11)
Call Signaling
(SIP)
Cisco ASA
Public/Private
WLAN
(802.11)
Mobile Data
Network
Internet
Note
348765
(Data Channel)
For DVO calls, all voice or media from the user's mobile phone will always travel through the mobile
voice network, PSTN, and enterprise PSTN gateway. Media never traverses the IP data connection to the
enterprise. The mobile data network connection is used only for call signaling traffic and other
application interactions.
For details on dial via office as implemented for Cisco Jabber clients, refer to Cisco Jabber Dial Via
Office for Dual-Mode Devices, page 21-83.
21-74
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Cisco mobile clients provides a streamlined configuration method for simplifying first-time end-user
client configuration at the mobile client device. This configuration method relies on RFC 2782 standard
Domain Name Service records (DNS SRV) within the corporate DNS server to automatically discover
collaboration services on the network. DNS SRV records direct the mobile client to appropriate
application servers for call control and IM and presence services. This configuration and provisioning
method alleviates the need for the user to manually configure the XMPP IM and presence server and
voice and video call control server or TFTP server host name or IP address. Instead the user simply enters
their user ID and domain name, and the client application automatically discovers the available
collaboration services and connects to these back-end servers, with the application prompting the user
for credentials as appropriate. If no services are discovered or if service discovery operation fails, then
the mobile client application reverts to manual configuration mode, requiring users to enter collaboration
application server host names or IP addresses and credentials. Multiple DNS SRV records with priority
and weighting indication ensure high availability of back-end collaboration application services as well
as mobile client distribution across multiple servers providing these services.
Note
Mobile client user simplified configuration does not simplify administrative tasks related to client and
service configuration and provisioning on the back-end application servers. All administrative tasks to
add user accounts, mobile client devices, and services configuration are still required in addition to
creating the DNS SRV record or records in the corporate DNS server.
Basic Access This use case enables basic Internet-only access for guest devices. This use case
provides the ability to enable employee-owned personal device network connectivity without
providing access to corporate resources.
Limited Access This use case enables full access to corporate network resources, but it applies
exclusively to corporate-owned devices.
Enhanced Access This use case enables granular access to corporate network resources for both
corporate-owned devices and employee-owned personal devices based on corporate policies.
Cisco collaboration mobile clients, whether running on corporate or personal devices, usually require
access to numerous back-end on-premises enterprise application components for full functionality. For
this reason the Limited or Enhanced Access use case scenarios generally apply to applications such as
Cisco Jabber for Android or iPhone. The chief difference between these two access models is that with
Limited Access, the corporate-owned devices are given full access to corporate network resources. In
the case of Enhance Access, not only is the scope expanded to include employee-owned devices, but
21-75
Chapter 21
Mobile Collaboration
access to corporate network resources can also be provided in a granular way so that devices and the
applications that run on them are able to access only specific resources based on corporate security
policies.
In the case of cloud-based collaboration services, Cisco mobile clients and devices connect directly to
the cloud through the Internet without the need for enterprise network attachment. In these scenarios,
user and mobile devices can be deployed using the Basic Access model because these use cases require
only Internet access.
For more information about the Cisco BYOD infrastructure and BYOD access use cases, refer to the
Cisco Bring Your Own Device (BYOD) Smart Solution Design Guide, available at
http://www.cisco.com/c/en/us/solutions/enterprise/data-center-designs-cloud-computing/own_devi
ce.html#~overview
When deploying Cisco mobile clients and devices within the Cisco BYOD infrastructure, consider the
following high-level design and deployment guidelines:
The network administrator should strongly consider allowing voice and video-capable clients to
attach to the enterprise network in the background (after initial provisioning), without user
intervention, to ensure maximum use of the enterprises telephony infrastructure. Specifically, use of
certificate-based identity and authentication helps facilitate an excellent user experience by
minimizing network connection and authentication delay.
In scenarios where Cisco mobile clients and devices are able to connect remotely to the enterprise
network through a secure VPN or VPN-less connection:
The network administrator should weigh the corporate security policy against the need for
seamless secure connectivity without user intervention in order to maximize utilization of the
enterprise telephony infrastructure. The use of certificate-based authentication and enforcement
of a device pin-lock policy provides seamless attachment without user intervention and
functionality similar to two-factor authentication because the end user must possess the device
and know the pin- lock to access the network. If two-factor authentication is mandated, then user
intervention will be required in order for the device to attach remotely to the enterprise.
It is important for the infrastructure firewall configuration to allow all required client
application network traffic to access the enterprise network. Failure to provide an appropriate
access solution or to open access to appropriate ports and protocols at the corporate firewall
could result in an inability of the Cisco mobile clients or devices to register to on-premises
Cisco call control for voice and video telephony services and/or the loss of other client features
such as enterprise directory access or enterprise visual voicemail.
When enterprise collaboration applications such as Cisco Jabber are installed on employee-owned
mobile devices, if the enterprise security policy requires the device to be wiped or reset to factory
default settings under certain conditions, device owners should be made aware of the policy and
encouraged to backup personal data from their device regularly.
When deploying Cisco collaboration mobile clients and devices, it is important for the underlying
network infrastructure from end-to-end to support the necessary QoS classes of service, including
priority queuing for voice media and dedicated video and signaling bandwidth, to ensure the quality
of client application voice and video calls and appropriate behavior of all features.
21-76
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Android
Various models of Android phones and tablets. (Consult the release notes referenced below for
specific device and firmware support information.) These devices must be running a minimum
firmware version of 4.1(2). Although not officially supported, Cisco Jabber for Android runs on
many Android devices running version 4.1(2) or later, with various degrees of limitations depending
on the device. The WLAN interfaces of most Android devices support 802.11a, 802.11b, 802.11g,
802.11n, and 802.11ac network connectivity.
Apple iOS
Various Apple iOS devices including iPhone and iPad. (Consult the release notes referenced below
for specific device and firmware support information.) These devices must be running a minimum
iOS version of 7.1. The WLAN interfaces of most Apple iOS devices support 802.11a, 802.11b,
802.11g, and 802.11n network connectivity. Some newer Apple devices support 802.11ac.
For details on the latest specific device and firmware version support, refer to the product release notes
for:
Android
http://www.cisco.com/c/en/us/support/unified-communications/jabber-android/products-release-no
tes-list.html
21-77
Chapter 21
Mobile Collaboration
The Cisco Jabber for Android, iPad, and iPhone clients not only provide voice and video over IP phone
services but also provide XMPP-based enterprise instant messaging (IM) and presence, corporate
contact and directory services when configured to access the enterprise contact source, and enterprise
voicemail message waiting indication (MWI) and visual voicemail when integrated to Cisco Unity
Connection.
The Cisco Jabber clients running on smartphones (Android and iPhone) are capable of performing only
manual hand-out as described in the section on Cisco Jabber Dual-Mode Handoff, page 21-80.
For more information about the Cisco Jabber Android and Apple iOS clients, additional feature details,
and supported hardware and software versions, refer to the Cisco Jabber documentation for:
Android
http://www.cisco.com/c/en/us/support/unified-communications/jabber-android/tsd-products-suppo
rt-series-home.html
As indicated previously, Cisco mobile clients such as Jabber are able to discover available collaboration
services by relying on DNS lookups and DNS SRV service record resolution. When service discovery is
properly configured, the user needs to enter only their user name and domain, and the client will
automatically discover and connect to available collaboration services.
As shown in Figure 21-25, during initial client configuration or in the case of network connection
changes, Jabber discovers collaboration services by querying DNS for the following SRV records:
_cisco_uds._tcp.<domain>
SRV record or records of this type are added to the enterprise DNS server when Jabber is deployed
in phone-only mode enabling voice and video over IP calling or in full UC mode enabling both voice
and video calling well as IM and presence. If the query for this record is resolved by DNS, Cisco
Jabber connects to Unified CM, determines the authenticator, and locates available services.
_cuplogin._tcp.<domain>
SRV record or records of this type are added to the enterprise DNS server when Jabber is deployed
in IM-only mode enabling XMPP-based IM and presence. If the query for this record is resolved by
DNS, Cisco Jabber connects to Unified CM IM and Presence and authenticates.
In the case of hybrid deployments with Cisco WebEx Messenger, during initial configuration and on
network connection changes, the client also issues an HTTP query to a central authentication service
(CAS) URL for Cisco WebEx Messenger service to determine if the domain is a valid WebEx domain.
When the client receives positive confirmation to the HTTP query that a valid WebEx domain has been
entered, the client then connects to and authenticates with the WebEx Messenger service and retrieves
client configuration and information on available UC services as configured in the Cisco WebEx Org
Admin.
21-78
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-25
Unified CM
IM and Presence
(imp.example.com)
Post discovery:
Connect
Authenticate
Download
configuration
Unified CM
(ucm.example.com)
Cisco
DNS
WebEx
Corporate
DNS
348776
While the UDS service runs on all nodes in the Unified CM cluster, when configuring DNS SRV records
for Unified CM UDS service, administrators should configure records for resolution to Unified CM
subscriber nodes only. This ensures that client interaction with the UDS service avoids the publisher
node and instead distributes the load across call processing nodes within the cluster.
In deployments where service discovery is not configured or reliance on DNS is not possible, the Jabber
client will revert to manual configuration, requiring the user to enter authenticator and service node IP
addresses. Manually configured IP addresses are cached by the Jabber client for use on subsequent
connections.
Once service discovery or manual configuration is complete, Jabber must authenticate and download a
service profile and/or the jabber-config.xml file (if available), which directs the client to additional
back-end application services such as voicemail and directory and enables appropriate configuration.
Cisco Jabber Corporate Directory Access
Cisco Jabber mobile clients rely on various methods for accessing enterprise contact information. In
addition to local device contacts and contacts previously added to the Jabber buddy list, Jabber mobile
clients are also able to access corporate directory services using the following methods:
21-79
Chapter 21
Mobile Collaboration
The jabber-config.xml file is used to configure the directory integration method for Jabber clients as well
as to configure certain directory related settings for Jabber clients.
Except for very small deployments, we recommend using the BDI method of directory access for Jabber
clients due to performance limitations of Unified CM UDS. When directory integration is configured to
use the UDS method, Unified CM node Jabber endpoint capacity is reduced by 50%. For example, when
deploying a 5,000 user OVA Unified CM node and using the UDS method of directory access for Jabber
clients, the Jabber endpoint capacity of the Unified CM node reduces from 5,000 to 2,500 Jabber devices
(assuming no other endpoints are configured on the node).
Cisco Jabber Dual-Mode Handoff
To properly deploy Cisco dual-mode clients such as Cisco Jabber, it is important to understand the nature
of handoff operations within the client. The handoff method used by the Cisco Jabber dual-mode client
depends on the Transfer to Mobile Network setting on the Cisco Dual Mode for iPhone or Cisco Dual
Mode for Android device configuration page.
There are two methods of handoff, depending on the Transfer to Mobile Network setting:
Note
Handoff capabilities apply only to dual-mode smartphones. This functionality is not supported on
devices without cellular voice radios, such as the Samsung Galaxy Note Pro.
Mobility Softkey Method of Hand-Out
The operation depicted in Figure 21-26 is of an active call on an iPhone or Android dual-mode device
within the enterprise being moved manually from the WLAN interface to the mobile voice network or
cellular interface of the device through the enterprise PSTN gateway. As shown, there is an existing call
between the mobile client device associated to the enterprise WLAN and registered to Unified CM, and
a phone on the PSTN network (step 1). Because this is a manual process, the user must select the Use
Mobile Network button from the in-call menu within the Cisco Jabber client, which signals to
Unified CM the intention to hand-out the call (step 2). Next Unified CM generates a call to the
configured mobility identity number corresponding to this mobile device through the enterprise PSTN
gateway (step 3). This call to the mobility identity is made to the mobile voice network or cellular
interface of the iPhone or Android device. The user can now move out of the enterprise and away from
WLAN network coverage (step 4). In the meantime, the inbound call from Unified CM is received at the
mobile voice network interface, and the user must answer the call manually to complete the hand-out.
Once the inbound call on the cellular interface is answered, the RTP stream that was traversing the
WLAN is redirected to the PSTN gateway, and the call continues uninterrupted between the mobile
client device and the original PSTN phone, with the call anchored in the enterprise gateway (step 5).
21-80
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-26
Enterprise
PSTN
Gateway
Mobile
Voice
Network
PSTN
Cisco Unified
CM Cluster
3
1
5
Handoff Number: 1234
Handoff Number DID: +1 408 555 1234
AccessPoint
Mobile
Voice
WLAN
Same
Phone
253854
Figure 21-27 illustrates the same hand-out operation as in Figure 21-26, where an active call on an
iPhone dual-mode phone within the enterprise is moved manually from the WLAN interface to the
mobile voice network or cellular interface of the device through the enterprise PSTN gateway. However,
in this case the Handoff Number method of hand-out is used.
As shown in Figure 21-27, there is an existing call between the dual-mode device associated to the
enterprise WLAN and registered to Unified CM, and a phone on the PSTN network (step 1). Because
this is a manual process, the user must select the Use Mobile Network button from the in-call menu
within the Cisco Jabber dual-mode client, which signals to Unified CM the intention to hand-out the call
(step 2). Next the Cisco Jabber client automatically generates a call through the cellular interface over
the mobile voice network to the configured Handoff Number within the Unified CM system (step 3). The
user can now move out of the enterprise and away from WLAN network coverage (step 4). In the
meantime, the inbound call from the Cisco Jabber client is received by Unified CM. Assuming the
inbound calling number matches the user's configured mobility identity, the RTP stream that was
traversing the WLAN is redirected to the PSTN gateway, and the call continues uninterrupted between
the Cisco Jabber mobile client and the original PSTN phone, with the call anchored in the enterprise
gateway (step 5).
21-81
Chapter 21
Mobile Collaboration
Figure 21-27
Enterprise
PSTN
Gateway
Mobile
Voice
Network
PSTN
Cisco Unified
CM Cluster
3
1
5
Handoff Number: 1234
Handoff Number DID: +1 408 555 1234
AccessPoint
Mobile
Voice
WLAN
Same
Phone
254273
Note
The Handoff Number method of hand-out requires Unified CM to receive an inbound calling number
from the PSTN network that matches the mobility identity number configured under the Cisco Dual
Mode device attempting the hand-out. If the caller ID is not sent by the dual-mode device, if the PSTN
provider does not send the inbound caller ID to the enterprise, or if the inbound caller ID does not match
the user's configured mobility identity, the hand-out operation will fail.
Note
Cisco Jabber dual-mode clients do not support hand-in. In scenarios where an in-progress call is active
between the dual-mode mobile voice network or cellular interface and an enterprise phone (or a PSTN
phone with the call anchored in the enterprise gateway), the only way to move the call to the WLAN
interface of the dual-mode device is to hang up the call and redial once the dual-mode client has
connected to the enterprise network and registered to Unified CM.
WLAN Design Considerations for Cisco Jabber Mobile Clients
Consider the following WLAN guidelines when deploying Cisco Jabber mobile clients:
Whenever possible, ensure that Cisco Jabber mobile clients roam on the WLAN only at Layer 2 so
that the same IP address can be used on the WLAN interface of the device. In Layer 3 roaming
scenarios where subnet boundaries are crossed due to device IP address changes, calls will be
dropped.
Deploy Cisco Jabber mobile clients on WLAN networks where the same SSID is used across all
APs. Roaming between APs is much slower if SSIDs are different.
21-82
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Ensure all APs in the WLAN broadcast their SSID(s). If the SSID is not broadcast by the AP, the
user may be prompted by the device to join other Wi-Fi networks or the device may automatically
join other Wi-Fi networks. When this occurs the call is interrupted.
Whenever possible, deploy Cisco Jabber mobile clients on the 5 GHz WLAN band (802.11a/n/ac).
5 GHz WLANs provide better throughput and less interference for voice and video calls.
The Unified CM administrator can enable or disable dial via office (DVO) calling for each dual-mode
device by using the Product Specific Configuration Layout section of the Cisco Dual Mode for iPhone
or Android device configuration page. Once DVO is enabled, the user can turn on DVO using the Calling
Options setting within the Cisco Jabber application. It is important to note that the DVO calling options
dictate not only the outbound calling method used by the Jabber client but also the inbound calling
method. Table 21-4 shows the various calling options and the corresponding outbound and inbound
calling method based on the type of network connectivity.
Table 21-4
Inbound and Outbound Calling Method with Cisco Jabber Dial Via Office Calling Options
Device
IP Connection
Outbound Call
Inbound Call
Voice over IP
Mobile Data
Single
Number
Reach
No IP
802.11 WLAN
(Corporate/enterprise)
Outbound Call
802.11 WLAN
(Non-corporate/enterprise)
Voice over IP
Inbound Call
Outbound Call
Inbound Call
Single
Number
Reach
Voice over IP
Voice over IP
Note
The Dial via Office calling feature applies only to dual-mode smartphones. This functionality is not
supported on tablets such as the Apple iPad because there is no cellular voice radio on those devices.
Note
The Dial via Office calling feature is not supported over secure connections using Cisco Expressway. If
Dial via Office calling is required, Cisco AnyConnect VPN should be used for enterprise secure remote
attachment.
21-83
Chapter 21
Mobile Collaboration
When dial via office is enabled for Cisco Jabber clients, as with Single Number Reach, the mobile
voicemail avoidance or single enterprise voicemail box feature of Cisco Unified Mobility is engaged. In
the case of dial via office, this voicemail avoidance feature ensures that, given a failure in the network
path or some other communication error during a DVO call setup, the called user does not end up in the
calling user's voicemail box. Typically the User Control method of voicemail avoidance provides the best
overall user experience because, if a DVO call leg inadvertently ends up being answered by a voicemail
system, the call leg will be disconnected when a DTMF tone is not received by Unified CM, and the DVO
call will be cleared. When Cisco Jabber users are enabled for the User Control method of mobile
voicemail avoidance, they should be reminded that they must press a button on the mobile device key
pad when receiving a mobility call at the client device. Failure to do so will result in call setup failure.
Note
Because the User Control method of mobile voicemail avoidance is completely dependent on successful
relay of the DTMF tone from the mobile device over the PSTN connection and PSTN gateway and
out-of-band to Unified CM, failure to propagate inbound DTMF from the PSTN to Unified CM results
in a disconnect of all enterprise calls made (dial via office reverse) or received (single number reach) by
the mobile device. If DTMF cannot be effectively relayed from the PSTN to Unified CM, then the Timer
Control mobile voicemail avoidance method should be used instead.
For more information about the single enterprise voicemail box voicemail avoidance feature, see Mobile
Voicemail Avoidance with Single Enterprise Voicemail Box, page 21-42.
Dial Via Office Calling Option Use Cases
When deploying dial via office, consider the following Cisco Jabber client calling option user profiles:
Voice over IP
The typical user profile for the Voice over IP calling option is a user that is mobile within the office
(home or enterprise) but for whom enterprise calling is not typically required outside the enterprise.
Additionally, with this user profile, mobile voice and data costs are usually an important
consideration for both corporate-paid and employee-paid mobile voice and data service.
Cisco Jabber clients support dial via office reverse (DVO-R). With this method of DVO, the call setup is
facilitated by an inbound call from the Unified CM system to the user's configured mobility identity or
mobile phone number.
Figure 21-28 illustrates a DVO-R call flow. In this example, a Cisco Jabber user wishes to dial a PSTN
phone at +1 408 555-7890. The user dials the number or selects the number from the contact list from
within the Cisco Jabber client, which generates a SIP call setup request over the IP connection to the
enterprise and Unified CM (step 1). Based on the call setup request, Unified CM generates a reverse call
back to the user's configured mobility identity (mobile phone number) using the enterprise PSTN
gateway (step 2). Once the incoming call from Unified CM is answered at the mobile device, a call is
21-84
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
extended to the number the user called or selected (step 3; in this case, +1 408-555-7890). Once the call
is answered at the far end, the media path is connected and the call is anchored through the enterprise
PSTN gateway (step 4). Because the call is now anchored in the enterprise gateway, the user has the
ability at any point during this call to use the Unified Mobility desk phone pickup feature as well as to
invoke Unified Mobility DTMF-based mid-call features.
Figure 21-28
+1 (408) 555-7890
3
4
PSTN
Gateway
PSTN
Cisco Unified CM
Mobile Voice
Network
Cisco ASA
Public/Private
WLAN
(Data Channel)
(802.11)
Internet
Note
348766
Mobile Data
Network
The call flow shown in Figure 21-28 assumes that Cisco Jabber is registered to Unified CM, that DVO
is enabled for the user, and that the client calling option setting is either Mobile Voice Network or
Autoselect. If the client setting is Autoselect, the dual-mode device running Cisco Jabber must be
IP-connected via the mobile data network. If connected over 802.11 WLAN, then the client would use
voice over IP rather than DVO.
By default the DVO-R callback call leg will be extended to the user's mobile device, as shown in
Figure 21-28; however; a user may specify an alternate callback number in the DVO Callback Number
field within the Cisco Jabber client. By default the DVO Callback Number field is populated with the
user's configured mobility identity. If the user configures a different number in this field, the DVO-R
callback call leg will be extended to that number. For example, rather than receiving the callback on the
mobile phone, the user may wish to direct the callback to their home phone.
21-85
Chapter 21
Mobile Collaboration
Note
When invoking DVO-R with an alternate callback number, if the callback call leg from Unified CM is
directed to a user-specified alternate number, the call is not anchored in the enterprise. In such cases,
users cannot perform desk phone pickup or invoke DTMF-based mid-call features on DVO-R calls using
an alternate callback number. In addition, voicemail avoidance does not engage for DVO-R alternate
number calls.
Mobile Profiles and Dial Via Office Reverse
Cisco Unified CM mobility profiles may be assigned to the mobility identity for mobile client devices.
While not required, the mobility profile specifies the caller ID sent by the system during setup of the
DVO-R callback call leg to the mobility identity or alternate callback number. The number configured
in the Callback Caller ID field of the Dial-via-Office Reverse Callback Configuration section of the
mobility profile configuration page is the number sent as caller ID. If no mobility profile is assigned to
the mobility identity or if the Callback Caller ID field is left blank, the system will send the configured
default Enterprise Feature Access Number.
Note
The Mobile Client Calling Option field of the mobility profile has no impact on DVO operation;
regardless of the setting, the Cisco Jabber client makes DVO-R calls when enabled for DVO calling. Dial
via Office Forward (DVO-F) is not a currently available calling option.
Cisco Jabber and Expressway Mobile and Remote Access
The mobile and remote access feature of the Cisco Expressway solution provides secure firewall
traversal for Cisco Jabber, enabling remote Jabber users to access enterprise collaboration applications
and services from their mobile devices when outside the enterprise.
All collaboration traffic traversing the Expressway mobile and remote access connection is encrypted,
including call media and signaling. The encrypted connection is between the Jabber endpoint and the
Expressway-C node inside the enterprise. Traffic between Expressway-C and endpoints or applications
inside the enterprise is unencrypted by default. Media and signaling traffic inside the enterprise is
encrypted only when the Unified CM cluster is configured as mixed mode with device authentication,
SRTP media, and TLS SIP signaling encryption facilitated by security configuration relying on the
Unified CM Cisco Certificate Trust List (CTL) Provider and Certificate Authority Proxy Function
(CAPF) services.
Jabber determines its location relevant to the enterprise (inside or outside) based on DNS query
resolution and a split DNS resolution design whereby the service records for Unified CM
(_cisco-uds._tcp) and Unified CM IM and Presence (_cuplogin._tcp) are configured only in the
corporate DNS and the service record for Expressway (_collab-edge._tls) is configured only on the
public DNS. This split design ensures that corporate DNS resolution points Jabber directly to
collaboration services when inside the enterprise and public DNS resolution points Jabber to connect
through Expressway. DNS queries are sent by Jabber whenever the network connection of the mobile
device changes.
As shown in Figure 21-29, Jabber queries DNS for three SRV service records: _cisco-uds._tcp,
_cuplog._tcp, and _collab-edge._tls. When inside the enterprise, the Jabber client receives resolution
from corporate DNS either pointing to Unified CM or Unified CM IM and Presence. In this case, Jabber
will connect directly to the resolved collaboration application service node(s). When outside the
enterprise, Jabber does not receive resolution for Unified CM or Unified CM IM and Presence from
public DNS, but instead receives resolution for Expressway directing the client to connect to the
enterprise through Expressway.
21-86
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-29
Cisco Jabber: Split DNS Resolution Inside and Outside the Enterprise
Enterprise Network
DMZ
Unified CM
IM and Presence
(imp.example.com)
Public Network
Expressway-E
Unified CM
Expressway-C
(expressway.example.com)
(ucm.example.com)
Internet
Corporate
DNS
Firewall
Firewall
DNS
Public
DNS
DNS
348778
Split DNS
DNS query/response:
_cisco-uds._tcp. ? = ucm.example.com
_cuplogin._tcp. ?
= imp.example.com
_collab-edge._tls. ? = [No resolution]
DNS query/response:
_cisco-uds._tcp. ? = [No resolution]
_cuplogin._tcp. ?
= [No resolution]
_collab-edge._tls. ? = expressway.example.com
Note
In cases where Cisco AnyConnect VPN is used for remote enterprise connectivity, Jabber will receive
DNS query resolution from corporate DNS through the VPN tunnel and will connect directly to
collaboration service nodes.
When deploying Expressway mobile and remote access for Cisco Jabber mobile clients, consider the
following unsupported features and functions:
Dual-mode hand-out
Moving an active call from the WLAN interface of the Jabber device to the cellular voice interface
is not supported over Expressway connections.
CAPF enrollment for endpoint authentication and media and signaling encryption
If secure media and signaling is required on the enterprise network, the Jabber device must complete
CAPF enrollment while on-premises and prior to connecting over Expressway.
21-87
Chapter 21
Mobile Collaboration
Session persistency
All calls and other collaboration application connections over Expressway mobile and remote access
are cleared whenever the network path changes or is lost.
If any of the above features and functions is required for the deployment, consider using AnyConnect
VPN instead of Expressway for remote secure enterprise access.
Cisco Jabber and Expressway Mobile and Remote Access with Cisco AnyConnect VPN Split-Tunnel
In some cases VPN and Expressway might need to be deployed in parallel, enabling Jabber users to
connect via either VPN or Expressway. In these situations, there are two methods of use. Jabber users
can rely on the Expressway mobile and remote access feature for collaboration workloads and rely on
VPN for all device traffic when connectivity back to the enterprise requires workloads outside of
collaboration. In these scenarios, when the Cisco AnyConnect VPN client establishes a connection back
to the enterprise, either due to VPN on-demand triggering or manual launch by the user, active
connections are dropped and the user must wait for the Jabber client to reconnect to provisioned
collaboration services over VPN before resuming use. This results in a poor user experience.
Alternatively, AnyConnect VPN and Expressway may be used simultaneously with split-tunneling to
force collaboration flows through the Expressway mobile and remote access connection and all other
traffic through the VPN tunnel. This alternative method often provides a better user experience because
it prevents the Jabber client from disconnecting from Expressway and reconnecting over VPN when the
VPN tunnel is established.
As shown in Figure 21-30, the split-tunneling afforded by this method of deployment relies on two basic
principles
DNS filtering at the Cisco Adaptive Security Appliance (ASA) VPN head-end
Traffic filtering at the ASA is used to filter DNS queries from the Jabber client for
_cisco-uds._tcp.<domain> and _cuplogin._tcp.<domain>. Because these DNS queries are filtered,
the Jabber client is unable to resolve Unified CM or IM and Presence service record requests for
direct connection to collaboration services. Therefore, the only DNS resolution will be for
_collab-edge._tcp.<domain>, which always results in Expressway connection and traversal.
21-88
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-30
Cisco Jabber: Expressway Mobile and Remote Access and Cisco AnyConnect VPN
Enterprise Network
DMZ
Unified CM
IM and Presence
(imp.example.com)
Public Network
Expressway-E
Unified CM
Expressway-C
(expressway.example.com)
(ucm.example.com)
Internet
64.53.42.256
Firewall
Corporate
DNS
Collaboration
traffic
Firewall
ASA
Split-tunnel
All other traffic
DNS
IP address exclude list:
64.53.42.256
[expressway.example.com]
DNS filter list:
348779
_cisco-uds._tcp.example.com
_cuplogin._tcp.example.com
In the case of AnyConnect VPN split-tunneling with Expressway mobile and remote access, the same
Expressway DNS SRV record (_collab-edge._tls) configured in the public DNS is added to the corporate
DNS. This prevents the need to provide access and forward DNS queries to the public DNS through the
VPN tunnel.
Although configuring an identical _collab-edge._tls SRV record in the corporate DNS would seem to
violate the foundational split DNS design expected with Jabber and Expressway mobile and remote
access deployments, in fact, Jabber's order of connection preference ensures appropriate behavior.
Jabber's order of connection preference is Unified CM (_cisco-uds._tcp) first, then IM and Presence
(_cuplogin._tcp), and finally Expressway (_collab-edge._tls). Therefore, even when the
_collab-edge._tls query is resolved by the corporate DNS, the client will still connect directly to
collaboration services because the corporate DNS will also resolve queries for _cisco-uds._tcp or
_cuplogin._tcp services.
For more information about Jabber and Expressway mobile and remote access with AnyConnect VPN,
refer to the information on mobile and remote access collaboration with Cisco Expressway Series, found
in the Cisco Unified Access (UA) and Bring Your Own Device (BYOD) CVD available at
http://www.cisco.com/c/en/us/td/docs/solutions/Enterprise/Borderless_Networks/Unified_Access/
BYOD_Design_Guide.html
21-89
Chapter 21
Mobile Collaboration
Cisco Jabber mobile clients are able to leverage single sign-on (SSO) using the Security Assertion
Markup Language (SAML) version 2. Jabber and Cisco collaboration infrastructure including
Unified CM, Unified CM IM and Presence, and Unity Connection leverage web-based SSO SAML v2
in order to identify and authenticate user connections, thus enabling the use of a single set of Jabber user
credentials for access to all collaboration services.
As depicted in Figure 21-31, Cisco Jabber SSO depends on pre-established trust relationships between
collaboration applications such as Unified CM, called service providers, and the identity provider (IdP).
Unified CM and Unity Connection service providers rely on LDAP sync and integration with the
corporate LDAP directory to identify users. Likewise, the IdP relies on the LDAP corporate directory
for authentication of users. Supported IdPs for Cisco Jabber and collaboration services include Ping
Federate, Microsoft Active Directory Federation Services (ADFS), and Open Access Manager
(OpenAM).
Figure 21-31 shows a basic Jabber SSO flow. The SSO flow begins with the Jabber client requesting
access to a collaboration service provider for example, access to Unified CM for call control services.
Rather than logging in directly to the collaboration service provider for access, the service provider
redirects the Jabber client to the IdP with a SAML authentication request. The IdP requests
authentication credentials from the Jabber user and authenticates the user against the corporate LDAP
directory. Assuming that the user is authenticated successfully, the IdP returns a signed assertion which
Jabber forwards to the collaboration service provider using HTTP POST. The collaboration service
provider then validates the signed assertion and provides authorization to the Jabber client. For example,
Jabber successfully registers to Unified CM.
21-90
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
Figure 21-31
LDAP
Directory
LDAP
sync
Unity
Connection
User
authentication
Identity
Provider
(IdP)
Trust
Unified CM
relationship
SP redirect
with SAML
authentication
request
SAML
authentication
and credential
exchange
348777
Access
request and
authorization
IdP signed
response and
HTTP post
In addition to forwarding a signed assertion to the Jabber client, the IdP stores a security context for the
authenticated Jabber client. Should the client request access to other collaboration service providers, the
IdP is able to provide subsequent signed assertions without requiring another exchange of credentials.
In this way, SSO enables the Jabber user or client to access multiple collaboration services by entering
their credentials once.
It is worth noting that the collaboration service provider never communicates directly with the IdP when
authenticating the user.
For more information about SSO, refer to the Identity Management Architecture Overview, page 16-32,
and the SAML SSO Deployment Guide for Cisco Unified Communications Applications available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
In addition to SSO user identification and authentication to on-premises collaboration applications and
services, SAML SSO can also be enabled for user authentication over Expressway mobile and remote
access connections. In these scenarios, an HTTPS reverse proxy is deployed in the DMZ of the enterprise
to broker authentication for inbound remote access connections. The HTTPS reverse proxy
communicates with the internal enterprise IdP and brokers the SAML request and authentication
21-91
Chapter 21
Mobile Collaboration
exchange between the remote client and the enterprise IdP. While the HTTPS reverse proxy in the DMZ
can be any generic HTTPS reverse proxy, some IdP vendors offer an option to install an IdP instance in
the DMZ to serve an IdP proxy role for brokering or proxying SSO SAML requests.
Interactions Between Cisco Jabber and Cisco Unified Mobility
The Cisco Jabber mobile clients can be integrated with Cisco Unified Mobility to leverage Cisco Single
Number Reach, mid-call DTMF features, two-stage dialing, and single enterprise voicemail box mobile
voicemail avoidance.
Integration with Unified Mobility requires the iPhone or Android dual-mode mobile phone number to be
configured within Unified CM as a mobility identity associated with the Cisco Dual Mode for iPhone or
Cisco Dual Mode for Android device. Once the mobile number is configured as a mobility identity
within the system, Single Number Reach can be leveraged so that incoming calls to the user's enterprise
number will be extended to the iPhone or Android dual-mode device through the mobile voice network
as long as the iPhone or Android dual-mode device is not connected to the enterprise and not registered
to Unified CM. In situations where the dual-mode device is connected to the enterprise, registered to
Unified CM, and the client calling options are set so that inbound voice-over-IP calling is enabled
("Voice over IP" or "Autoselect" when the device is connected to a WLAN), an inbound call to the
enterprise number will not be extended to the mobile voice network interface of the device. When the
iPhone or Android dual-mode device is connected to the enterprise, only the WLAN or mobile data
interface of the device will receive the inbound call. This prevents unnecessary consumption of
enterprise PSTN gateway resources.
When handling enterprise calls through the cellular voice network, the iPhone or Android dual-mode
device can invoke mid-call features by means of DTMF and perform desk phone pickup for any
enterprise anchored call. The dual-mode device can also leverage Mobile Voice Access and Enterprise
Feature Access two-stage dialing features when making outbound calls to route these calls through the
enterprise and anchor them in the enterprise PSTN gateway.
In addition to configuring a mobility identity for the iPhone or Android dual-mode device, you can
configure additional mobile phone numbers or off-system phone numbers as remote destinations and
associate them to the Cisco Dual Mode for iPhone or Cisco Dual Mode for Android device within
Unified CM. When associating the mobility identity and additional remote destinations to the dual-mode
device, you do not have to configure a remote destination profile.
When mobile users are provisioned with multiple Cisco mobile clients across multiple mobile devices
(for example, a user running Cisco Jabber for Android on their Android smartphone and Cisco Jabber
for iPhone and iPad on their Apple iPad), associate the mobility identity with the dual-mode device (for
example, Cisco Dual Mode for Android) rather than with the tablet device (Cisco Jabber for Tablet).
Because the dual-mode device leverages functionality unique to the mobility identity, including
dual-mode handoff and dial via office, the mobility identity should be associated to this device.
Associate all other remote destinations to the same device as the mobility identity. Associating different
remote destinations on different mobile client devices for the same user makes configurations more
complex and troubleshooting issues more difficult.
For more information about the Cisco Unified Mobility feature set as well as design and deployment
considerations, see Cisco Unified Mobility, page 21-34.
Interactions Between Cisco Jabber and Cisco Intelligent Proximity for Mobile Voice
The Intelligent Proximity for Mobile Voice feature is designed to enable hands-free audio for the cellular
or mobile line of a dual-mode devices. For this reason, usually only calls on the cellular line of the Jabber
client device are enabled for hands-free audio play out on an Intelligent Proximity-capable IP endpoint.
In the case of voice or video over IP calls on Cisco Jabber, Intelligent Proximity for Mobile Voice is not
invoked. The one exception to this is with the Cisco IP Phone 8851 and 8861 endpoints. Because these
IP phones are audio-only, with Intelligent Proximity for Mobile Voice, audio for a Jabber IP-based call
21-92
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
is streamed through the 8851 or 8861 phone while the video portion of this call remains on the Jabber
client device. In the case of other hardware endpoints capable of Intelligent Proximity for Mobile Voice,
audio for Jabber IP-based calls is not played by the IP endpoint.
Cisco Spark
The Cisco Spark mobile client is available for Android and Apple iOS mobile devices, including iPad
and iPhone. Once the client application is downloaded from the appropriate application store (Apple
Application Store or Google Play) and installed on the Apple iOS or Android device, users must enter
their email address and activate their account with the resulting provisioning email. Once a user activates
their account, the client connects to the Cisco Collaboration Cloud and the user can begin creating secure
collaboration rooms with one or more people to communicate using encrypted instant messaging (IM).
The user should access Cisco Spark at http://web.ciscospark.com/ using a web browser at least once in
order to set a password for their account. Alternatively, the user can use the desktop Spark client
available for download from http://download.ciscospark.com/. Failure to do this will require the user to
activate their account via email each time they connect with the mobile client.
Cisco Spark for Android, iPad, and iPhone clients not only provide secure persistent IM collaboration
rooms, but they also provide encrypted voice and video calling over IP and file sharing capabilities.
For proper Cisco Spark client operation, the mobile device must be able to reach the Internet by
connecting to a wireless network (enterprise or public/private 802.11 WLAN or mobile provider data
network).
For more information about the Cisco Spark Android and Apple iOS clients, additional feature details,
and supported hardware and software versions, refer to the Cisco Spark documentation at
http://support.ciscospark.com/
Cisco Cloud Collaboration Services: SAML SSO for Cisco Spark and Cisco WebEx
Just as with on-premises enterprise and collaboration edge deployments described earlier, enterprise
SSO can be used to facilitate secure logins to cloud collaboration services such as Cisco Spark and Cisco
WebEx. With these types of deployments the enterprise IdP in combination with an HTTPS reverse
proxy deployed in the enterprise DMZ leverage enterprise credentials to identify and authenticate user
access to Cisco Spark and Cisco WebEx.
21-93
Chapter 21
Mobile Collaboration
When employing VPN network connectivity for connections over the mobile data network or public or
private Wi-Fi hot spots, it is important to deploy a high-bandwidth secure VPN infrastructure that
adheres to the enterprise's security requirements and policies. Careful planning is needed to ensure that
the VPN infrastructure provides high bandwidth, reliable connections, and appropriate session or
connection capacity based on the number of users and devices using this connectivity.
For more information on secure remote VPN connectivity using Cisco AnyConnect, refer to the Cisco
AnyConnect Secure Mobile Client documentation available at
http://www.cisco.com/c/en/us/support/security/anyconnect-secure-mobility-client/tsd-products-su
pport-series-home.html
21-94
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
21-95
Chapter 21
Mobile Collaboration
The above considerations are certainly not all unique to mobile clients and devices. They apply to all
situations in which devices and users are added to Unified CM, resulting in additional load to the overall
system.
For more information on general system sizing, capacity planning, and deployment considerations, see
the chapter on Collaboration Solution Sizing Guidance, page 25-1.
Dual-mode mobile devices must be capable of dual transfer mode (DTM) in order to be connected
simultaneously to both the mobile voice and data network and the WLAN network so that the device
is reachable and able to make and receive calls on both the cellular radio and WLAN interface of
the device. In some cases, proper dual-mode client operation might not be possible if mobile voice
and data networks do not support dual-connected devices.
WLAN APs should be deployed with a minimum cell overlap of 20%. This overlap ensures that a
mobile device can successfully roam from one AP to the next as the device moves around within a
location, while still maintaining voice and data network connectivity.
WLAN APs should be deployed with cell power level boundaries (or channel cell radius) of
-67 dBm in order to minimize packet loss. Furthermore, the same-channel cell boundary separation
should be approximately 19 dBm. A same-channel cell separation of 19 dBm is critical for ensuring
that APs or clients do not cause co-channel interference to other devices associated to the same
channel, which would likely result in poor voice and video quality.
Whenever possible rely on the 5 GHz WLAN band (802.11a/n/ac) for connecting mobile clients and
devices capable of generating voice and video traffic. 5 GHz WLANs provide better throughput and
less interference for voice and video calls.
The enterprise wired and wireless LAN should be deployed and configured to support the necessary
end-to-end QoS classes of service, including priority queuing for voice media and dedicated video
and signaling bandwidth, to ensure the quality of client application voice and video calls and the
appropriate behavior of all features. While most clients mark traffic appropriately at Layer 3 based
on Cisco QoS recommendations, appropriate Layer 2 WLAN UP marking is dependent on the client
device and vendor implementation. For this reason, Layer 2 marking is not consistent across
platforms and as such cannot be relied upon.
Because mobile devices are similar to desktop computers and can generate a large variety of data
and real-time traffic, these devices are typically considered untrusted. For this reason, the network
should be configured to re-mark all traffic from these client devices based on port number and/or
protocol. Likewise, rate limiting and policing on ingress to the network is recommended.
Cisco recommends using only an enterprise-class voice and video optimized WLAN network for
connecting mobile devices and clients. While most mobile client devices are capable of attaching to
public or private WLAN access points or hot spots for connecting back to the enterprise through the
Internet for call control and other collaboration services, Cisco cannot guarantee voice and video
quality for these types of connections.
When deploying Cisco collaboration mobile clients and devices on a Cisco Bring Your Own Device
(BYOD) infrastructure, administrators should consider a network attachment method that does not
require user intervention and which maximizes utilization of the IP telephony infrastructure.
Further, for remote connectivity scenarios, all relevant ports must be opened in the corporate firewall
in order for Cisco mobile clients and devices to be able to access collaboration services.
21-96
Chapter 21
Mobile Collaboration
Mobility Beyond the Enterprise
If corporate policy dictates that the BYOD infrastructure must remotely wipe or factory-reset lost
or stolen mobile devices, employees using personal mobile devices should be aware of the policy
and should regularly back up personal data.
The Unified Mobility Single Number Reach feature will not extend incoming calls to the dual-mode
device's configured mobility identity if the dual-mode device is inside the enterprise and registered
to Unified CM. This is by design in order to reduce utilization of enterprise PSTN resources.
Because the dual-mode device registers to Unified CM, the system knows whether the device is
reachable inside the enterprise; and if it is, there is no reason to extend the call to the PSTN in order
to ring the dual-mode device's cellular voice radio. Only when the dual-mode device is unregistered
will Single Number Reach extend incoming calls to the user's enterprise number out to the mobility
identity number on the PSTN.
When you deploy mobile devices, Cisco recommends normalizing required dialing strings so that
users are able to maintain their dialing habits, whether the mobile device is connected to the
enterprise or not. Because dialing on the mobile network is typically done using full E.164 (with or
without a preceding '+') and mobile phone contacts are typically stored with full E.164 numbers,
Cisco recommends configuring the enterprise dial plan to accommodate full E.164 or full E.164 with
preceding '+' for mobile client devices. By configuring the enterprise dial plan in this manner, you
can provide the best possible end-user dialing experience so that users do not have to be aware of
whether the device is registered to Unified CM.
Cisco recommends that dual-mode phone users rely exclusively on the mobile voice network for
making emergency calls and determining device and user location. This is because mobile provider
networks typically provide much more reliable location indication than WLAN networks. To ensure
that dual-mode phones rely exclusively on the mobile voice network for emergency and location
services, configure the Emergency Numbers field of the dual-mode devices within Unified CM with
emergency numbers such 911, 999, and 112 in order to force these calls over the mobile voice
network. Dual-mode phone users should be advised to make all emergency calls over the mobile
voice network rather than the enterprise network. Although making emergency calls over corporate
WLANs or mobile data networks is not recommended, mobile devices that do not have cellular voice
radios are capable of making calls only through these data interfaces. Mobile devices that do not
have cellular voice radios should not be relied upon for making emergency calls.
When deploying Cisco Jabber on mobile devices, configure the WLAN network to accommodate
the following deployment guidelines:
Minimize roaming of Cisco Jabber mobile client devices at Layer 3 on the WLAN. Layer 3
roaming, where a device IP address changes, will result in longer roam times and dropped voice
packets and could even result in dropped calls.
Configure the same SSID across all APs utilized by the Cisco Jabber mobile client devices
prompts to join other APs within the WLAN infrastructure, which could result in interrupted
calls.
Provide sufficient wireless voice and video call capacity on the enterprise wireless network for
Cisco mobile clients and devices by deploying the appropriate number of wireless APs to handle the
desired call capacity based on mobility-enabled user BHCA rates. Each 802.11g/n (2.4 GHz) or
802.11a/n/ac (5 GHz) channel cell can support a maximum of 27 simultaneous voice-only calls with
24 Mbps or higher data rates. Each 802.11g/n (2.4 GHz) or 802.11a/n/ac (5 GHz) channel cell can
support a maximum of 8 simultaneous video calls assuming 720p video resolution at up to 1 Mbps
bit rate. For 2.4 GHz WLAN deployments, Bluetooth must be disabled to achieve this capacity.
Actual call capacity could be lower depending on the RF environment, wireless endpoint type, and
WLAN infrastructure.
21-97
Chapter 21
Mobile Collaboration
When deploying Dial via Office Reverse (DVO-R), use of the User Control method of voicemail
avoidance ensures that called users do not end up in the calling user's voicemail box. This method
of voicemail avoidance requires the calling user to press a number on the mobile device key pad in
order to connect the DVO-R call. Failure to press a key on the mobile device results in the DVO call
being cleared.
DVO-R calls using the alternate callback number are not anchored in the enterprise and therefore
desk phone pickup and DTMF-based mid-call features may not be used on these calls. In addition,
voicemail avoidance is not engaged for calls to alternate callback numbers.
The following features and capabilities are not supported over Expressway mobile and remote
access connections: DVO, WLAN to cellular dual-mode handoff, LDAP directory access, per-user
or per-device access restrictions, and session persistency during network path changes. If any of
these features are required, consider implementing a Cisco AnyConnect VPN solution for Jabber
mobile clients.
Given the performance limitations of Unified CM UDS, Jabber mobile clients should use the basic
directory integration (BDI) method of directory access whenever possible. When Jabber clients rely
on UDS for directory access, Unified CM node registration capacity is reduced by 50%. Jabber
clients connected over Expressway mobile and remote access always use UDS for directory access
even when BDI has been configured.
When mobile users are provisioned with multiple Cisco mobile clients across multiple mobile
devices, the mobility identity and any additional remote destinations should always be associated to
the Cisco Jabber dual-mode device type.
After initially downloading, installing, and activating the Spark account via the mobile device, the
user should access Cisco Spark using a web browser or desktop client in order to create a password
for their account. Once this is done, the user will be able to access Cisco Spark using any client
(mobile, desktop, or web browser). Failure to set a password results in the user having to re-activate
their account through email after sign-out each time.
21-98
CH A P T E R
22
This chapter describes the Cisco Unified Contact Center solutions available with the Cisco Unified
Communications System. It includes information on Cisco products such as Cisco Unified Contact
Center Express, Cisco Unified Contact Center Enterprise, and Cisco Unified Customer Voice Portal. It
also covers the design considerations for deploying these Cisco Unified Contact Center products with
Cisco Unified Communications Manager and other Unified Communications components.
This chapter covers the following topics:
This chapter starts with a high-level overview of the main Cisco Unified Contact Center Portfolio. Then
it covers the various Unified Communications deployment models for contact centers. Finally, it
discusses design considerations on topics such as bandwidth, latency, Cisco Unified Communications
Manager integration, and sizing.
The intent of this chapter is not to provide details on each contact center product and their various
components but rather to discuss the design considerations for their integration with the Cisco Unified
Communications System. Detailed design guidance for each Unified Contact Center product is covered
in specific design guides for the Cisco Unified Contact Center Express, Cisco Unified Contact Center
Enterprise, and Cisco Unified Customer Voice Portal products. Links to the product-specific design
guides are listed at
http://www.cisco.com/go/ucsrnd
22-1
Chapter 22
Described in:
Revision Date
Contact Sharing
Context Service
In addition, Unified CM call queuing provides statistics on the number of calls currently waiting in
queue and the longest call waiting time, along with other statistics, through the serviceability counters
based on the hunt pilot number. This allows the supervisor to monitor the queue status using the Real
Time Monitoring Tool (RTMT). For details on the serviceability counters, refer to the latest version of
the Feature Configuration Guide for Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
22-2
Chapter 22
For each hunt pilot, callers can be routed to an alternate configurable destination such as voicemail or
another hunt pilot if any of the following situations occurs:
Note
The number of calls in the queue reaches the maximum that is set by the Maximum Number of
Callers Allowed in Queue parameter.
The wait time of a caller in queue exceeds the threshold that is configured by the Maximum Wait
Time in Queue parameter.
For calls routed to the queue-enabled hunt pilot number through a SIP trunk, the SIP Rel1XX Options
should be set to Send PRACK if 1XX contains SDP in the SIP profile associated with the SIP trunk.
For additional information on the Unified CM call queuing option, refer to the latest version of the
Feature Configuration Guide for Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
http://www.cisco.com/en/US/products/ps12586/index.html
http://www.cisco.com/en/US/products/ps12586/tsd_products_support_series_home.html
http://docwiki.cisco.com/wiki/Packaged_CCE
Call Router
The Call Router makes all the decisions on how to route a call or customer contact.
Logger
The Logger maintains the system database that stores contact center configurations and temporarily
stores historical reporting data for distribution to the data servers. The combination of Call Router
and Logger is called the Central Controller.
22-3
Chapter 22
Peripheral Gateway
The Peripheral Gateway (PG) interfaces to various "peripheral" devices, such as Cisco Unified CM,
Cisco Unified IP Interactive Voice Response (Unified IP IVR), Cisco Unified CVP, or multichannel
products such as Cisco Unified Web Interaction Manager (Unified WIM) and Cisco Unified E-Mail
Interaction Manager (Unified EIM). A Peripheral Gateway that interfaces with Unified CM is also
referred to as an Agent PG.
The Cisco Unified CCE solution is based on the integration with Cisco Unified Communications
Manager (Unified CM), which controls the agent phones. For deployments without Unified CM but with
traditional ACD, use Cisco Unified Intelligent Contact Management Enterprise (Unified ICME) instead
of Unified CCE.
The queuing and self-service functions are provided by Cisco Unified IP Interactive Voice Response
(Unified IP IVR) or Cisco Unified Customer Voice Portal (Unified CVP) and are controlled by the
Unified CCE Call Router.
Most of the Unified CCE components are required to be redundant, and these redundant instances are
referred to as side A and side B instances. For example, Call Router A and Call Router B are redundant
instances of the Call Router component running on two different virtual machines.
Agents can use a large variety of endpoints, including some video endpoints and some Cisco
TelePresence endpoints such as the Cisco DX70 and DX80. For a list of supported endpoints, refer to
the Compatibility Matrix for Unified CCE, available at
http://docwiki.cisco.com/wiki/Compatibility_Matrix_for_Unified_CCE
22-4
Chapter 22
Unified CVP can be deployed standalone or integrated with Unified CCE to offer voice and video
self-service and queuing functions. The Unified CVP solution now supports the G.711 a-law codec
end-to-end.
The Basic Video Service in Unified CVP is available when Unified CVP is deployed along with Cisco
Contact Center Enterprise (Unified CCE) in a comprehensive deployment model. This service allows a
video caller to interact with an audio-only IVR and subsequently connect with a video agent. It supports
Cisco TelePresence endpoints such as the Cisco DX70 and DX80 as customer and agent endpoints. The
video agents can also conference in a second audio-only agent by dialing a direct extension from their
phone.
Video in Queue (VIQ) Basic Video is an optional feature in Unified CVP, and it can be enabled to play
video to callers while they wait for a video-enabled agent or expert. Cisco MediaSense or Cisco
TelePresence Content Server enables the video streaming. The caller can subsequently connect to a video
agent.
For more information on Unified CVP system design and detailed call flows, refer to the latest version
of the Cisco Unified Customer Voice Portal Design Guide, available at
http://www.cisco.com/c/en/us/support/customer-collaboration/unified-customer-voice-portal/produ
cts-implementation-design-guides-list.html
22-5
Chapter 22
Cisco SocialMiner
Cisco SocialMiner is a social media customer-care solution that can help you proactively respond to
customers and prospects by communicating through public social media networks such as Twitter,
Facebook, or other public forums or blogging sites. By providing social media monitoring, queuing, and
workflow to organize customer posts on social media networks and deliver them to your social media
customer care team, your company can respond to customers in real time using the same social network
the customers are using. For more information, refer to the documentation available at
http://www.cisco.com/en/US/products/ps11349/index.html
22-6
Chapter 22
In addition, Cisco Unified Contact Center Management Portal (Unified CCMP) can be deployed to
simplify the operations and procedures for performing basic administrative functions such as managing
agents and equipment. Unified CCMP is a browser-based management application designed for use by
contact center system administrators, business users, and supervisors. It is a dense multi-tenant
provisioning platform that overlays the Cisco Unified CCE, Unified CM, and Unified CVP equipment.
Reporting
Cisco Unified Intelligence Center is the main reporting tool for the Cisco Contact Center solutions. It is
supported by Unified CCE, Unified CCX, and Unified CVP. This platform is a web-based application
offering many Web 2.0 features, high scalability, performance, and advanced features such as the ability
to integrate data from other Cisco Unified Communications products or third-party data sources.
Cisco Unified Intelligence Center gets source data from a database, such as an Unified CCE
Administration & Data Server database or the Unified CVP Reporting Informix database. Reports are
then generated and provided to a reporting client.
Multichannel Support
The Cisco Unified Enterprise solution supports web interaction and email interaction for multichannel
support. Cisco Unified Web Interaction Manager (Unified WIM) technology helps ensure that
communication can be established from nearly any web browser. Cisco Unified E-Mail Interaction
Manager (Unified EIM) provides inbound email routing, automated or agent assisted email responses,
real-time and historical reporting, and role-based hierarchical rights management for agents,
supervisors, administrators, and knowledge base administrators.
For more design information on these products, refer to the Cisco Unified Web and E-Mail Interaction
Manager Solution Reference Network Design Guide, available at
http://www.cisco.com/en/US/products/ps7236/products_implementation_design_guides_list.html
For more details on call recording and monitoring, see the chapter on Call Recording and Monitoring,
page 23-1.
22-7
Chapter 22
Contact Sharing
Contact Sharing enables large contact centers to grow larger. Centralized self-service (IVR ICM
deployment model) uses a Contact Sharing routing node to distribute calls to two Unified CCE instances
for horizontal scaling. Live Data is a prerequisite for Contact Sharing and must be installed and
configured prior to use of Contact Sharing. Contact Sharing also requires the IVR Cisco Intelligent
Contact Management (ICM) deployment model to be enabled in the deployment. For details about
Contact Sharing, refer to the latest version of the Cisco Unified Contact Center Enterprise Features
Guide, available at
http://www.cisco.com/c/en/us/support/customer-collaboration/unified-contact-center-enterprise/pr
oducts-feature-guides-list.html
Context Service
Context Service is a cloud-based storage service that provides a repository for customer journey data. It
enables Cisco Contact Center customers to deliver a seamless omnichannel experience through
integration with other Cisco Customer Collaboration products as well as APIs for third-party integration,
as depicted in Figure 22-1.
Figure 22-1
Social
Media
Chat
Customer
Journey
Phone
SMS
IVR
Retail
Mobile
Web
Cisco Customer
Collaboration Platforms
Partner Applications
348811
IoE
22-8
Chapter 22
Context Service allows any application to write and read customer journey activity. Cisco Contact Center
customers, referred to in this section as the business, have access to Context Service from within their
Cisco Contact Center platforms. The Cisco Contact Center platforms are enabled and optionally
configured to post context data about contact center interactions.
Context Service stores this data in an element called a Piece of Data (POD). A POD can store any
metadata about the consumer interaction, except for the media (such as audio recording). Businesses
choose which fields (metadata) to store in the POD and the level of data privacy for each field. PODs
can be organized by customer and also grouped together as part of a collection of interactions called a
Request (see Figure 22-2). Context Service also provides tagging capability to group PODs for
correlation, trending, and analytics.
Figure 22-2
Chat
Phone
Email
IVR
Agent
Social
Media
SMS
Phone
Customer
Request
Pod
Tag Pod
POD
Tag
T
a Pod
Pod
POD
Mobile
Web
IoE
Partner Applications
348812
Retail
Context Service is hosted on Cisco Intercloud, which is an ecosystem of Cisco and partner data centers
that is managed and operated by the Cisco data center team across the globe. Context Service follows a
data privacy model very similar to Cisco Spark, in which each business controls access to its data. The
data is encrypted/decrypted on-premises at the client and stored as an encrypted blob in Cisco data
centers. Businesses can choose to host the encryption keys (Keystore) on their premises. This is
analogous to valuables stored in a safety deposit box (locker) at a bank; even though the valuables are
in the bank, the customer has the key to the deposit box and controls access to it. This is a newer approach
to data privacy, and it puts the customer in control without the overhead of hosting a private cloud.
Context Service provides a level of data privacy classification so that businesses can store their
customers Personally Identifiable Information (PII) separately from other encrypted data, and thereby
allowing businesses to provide third-party analytics vendors with controlled access to their encrypted
data without giving access to their customers PII data.
22-9
Chapter 22
Context Service is managed by Cisco Collaboration Management (Atlas), which is the management
portal for all Cisco cloud collaboration offerings, including Cisco Spark. Cisco partners and businesses
use Collaboration Management to connect on-premises clients, manage the POD data model (fields),
monitor POD usage, and so forth.
Context Service provides an open API and Java/JS SDK to make it easy for technology partners to
integrate their applications with Context Service.
For more details on Cisco Context Service, refer to the Cisco Unified Contact Center design guides
available at http://www.cisco.com/go/ucsrnd.
22-10
Chapter 22
Figure 22-3
IP
PSTN
Contact Center
IP
IP
253824
WAN
Because the agents or the voice gateways in this type of deployment are located in remote sites, it is
important to consider the bandwidth requirements between the sites. It is also important to carefully
configure call admission control, Quality of Service (QoS), codecs, and so forth. For more information
on the general design considerations for Unified Communications solutions, refer to the chapter on
Collaboration Deployment Models, page 10-1.
Contact center deployments in a Unified Communications system typically have the following additional
bandwidth requirements:
The traffic volume handled by the agents is higher than that of typical users, and therefore voice and
signaling traffic is also higher for agents.
Agents and supervisors use desktops with screen popup, reports and statistics, and so forth. This
causes data traffic between the agent or supervisor desktops and the contact center servers. In
addition, bandwidth calculations must account for reporting information if, for example, an agent or
supervisor is remote and pulls data from a server in a central location. For more information and
guidance, refer to the design guides for the individual Cisco Contact Center products, available at
http://www.cisco.com/go/ucsrnd.
Depending on type of IVR solution, there could be traffic between the voice gateway and the IVR
system. For example, if the voice gateways are distributed and calls arrive at a voice gateway located
in a remote site with Unified IP IVR, there would be voice traffic across the WAN between the voice
gateway and Unified IP IVR. With Unified CVP, the call could be queued at the remote site, with
the VXML Gateway providing call treatment and queuing and therefore avoiding voice traffic across
the WAN for IVR and reducing overall WAN bandwidth requirements.
Remote agents (for example, agents working from home) are also supported with Cisco Unified Contact
Center. There are mainly two solutions. The first one requires the agent to use an IP phone that is
connected to the central site by a broadband internet connection. In this solution, the phone is CTI
22-11
Chapter 22
controlled by the Cisco Unified Contact Center application. The second solution is based on Cisco
Unified Mobile Agent, which enables an agent to participate in a call center with any PSTN phone such
as cell phone.
V
CC
PSTN
IP
Unified CCX
CC
Unified CCX
V
WAN
V
CC
IP
253826
IP
Unified CCX
Requirements for Unified CCE differ from Unified CCX. A single Unified CCE system can span across
multiple Unified CM clusters distributed across multiple geographic locations. A Unified CCE Agent
PGs must be installed in each Unified CM cluster location and could be physically remote from the
Unified CCE Central Controller (Call Router + Logger). Figure 22-5 illustrates this type of deployment
and highlights the placement of the Agent PG.
22-12
Chapter 22
Figure 22-5
Agent PG
IP
PSTN
CC
Unified CCE
Agent PG
V
WAN
IP
253827
IP
Agent PG
If you require multiple contact center deployments, you could connect those deployments through
Unified ICM by using the parent/child deployment model to form a single virtual contact center. The
parent/child model provides several benefits, such as enterprise queuing and enterprise reporting across
all the contact center deployments. It also provides complete site redundancy and higher scalability. For
more details on the parent/child model, refer to the following documents:
Cisco Contact Center Gateway Deployment Guide for Cisco Unified ICME/CCE, available at
http://www.cisco.com/c/en/us/support/customer-collaboration/unified-contact-center-enterprise/pr
oducts-installation-guides-list.html
Similarly to the multisite model with centralized call processing, multisite deployments with distributed
call processing require careful configuration of QoS, call admission control, codecs, and so forth.
22-13
Chapter 22
For example, with Unified CCE, the side A components could be remote from the Unified CCE side B
components and separated from them by an IP WAN connection. (For more details on Unified CCE high
availability, see High Availability for Contact Centers, page 22-15.) The following design considerations
apply to this type of deployment:
The IP WAN between the two sites must be highly available, with no single point of failure. For
example, the IP WAN links, routers, and switches must be redundant. WAN link redundancy could
be achieved with multiple WAN links or with a SONET ring, which is highly resilient and has
built-in redundancy.
The Agent Peripheral Gateway (PG) and the CTI Manager to which it is connected must be located
in the same data center. Because of the large amount of redirect and transfer traffic and additional
CTI traffic, the Intra-Cluster Communication Signaling (ICCS) bandwidth requirements between
the Unified CM nodes are higher when deploying Unified CCE.
If the primary Unified CCE and Unified CM nodes are located in one site and the secondary
Unified CCE and Unified CM nodes are in another site, the maximum latency between the two sites
is dictated by the Unified CM latency requirement of 80 ms round trip time (RTT). However, if the
Unified CCE nodes are in different locations than the Unified CM nodes, it is possible to have a
higher latency between the redundant Unified CCE nodes.
Figure 22-6 illustrates a deployment of Unified CCE using clustering over the WAN. For more details,
refer to the Cisco Unified Contact Center Enterprise Design Guide, available at
http://www.cisco.com/c/en/us/support/customer-collaboration/unified-contact-center-enterprise/pr
oducts-implementation-design-guides-list.html
Figure 22-6
PSTN
CC
CC
V
V
V
WAN
Unified CCE
Side B
253828
Unified CCE
Side A
IP
IP
With Unified CCX and Unified IP IVR solutions, the primary Unified CCX or Unified IP IVR node
could also be remote from the backup node. The requirements for Unified CCX deployments are
different than the ones for Unified CCE deployments. For example, redundant WAN links are not
required with Unified CCX. Also, the maximum latency between the primary and backup Unified CCX
nodes is 80 ms RTT. Figure 22-7 illustrates this type of deployment. For more details, refer to the Cisco
Unified Contact Center Express Design Guide, available at
http://www.cisco.com/c/en/us/support/customer-collaboration/unified-contact-center-express/prod
ucts-implementation-design-guides-list.html
22-14
Chapter 22
Figure 22-7
PSTN
CC
CC
Unified CCX
Unified CCX
253829
WAN
IP
IP
22-15
Chapter 22
In addition to the redundancy of the Unified Contact Center components themselves, their integration
with Unified CM can also be redundant. For example, each Unified CCX or Unified IP IVR node can
connect to a primary CTI Manager and also to a backup CTI Manager in case the primary CTI Manager
fails. With Unified CCE, a PG side A would connect to a primary CTI Manager, while the redundant PG
side B connects to the secondary CTI Manager, thus providing high availability if one CTI Manager
fails.
For more details, refer to the Cisco Unified Contact Center design guides available at
http://www.cisco.com/go/ucsrnd.
Bandwidth Provisioning
Contact center solutions require sufficient WAN bandwidth to accommodate the following main types
of traffic:
Voice traffic between the ingress gateway and the IVR system. With Unified IP IVR, if the
Unified IP IVR cluster is in a central location and PSTN gateways are in remote locations, there will
be voice traffic over the WAN. With Unified CVP, it is possible to queue the call at the edge and
therefore keep the voice traffic local to the remote site to avoid voice traffic across a WAN link.
Video queuing is also supported with the Unified CVP Video in Queue (ViQ) feature, so also
consider the video traffic between the caller and the video media server.
Voice traffic between the ingress gateway and the agent, or voice traffic between the caller and agent
for internal calls. There could also be video traffic between the caller and the agent if the contact
center deployment supports video.
Voice and video signaling traffic. This is typically for the signaling traffic between the ingress
gateway or caller endpoint and Unified CM, and between the agent phone and Unified CM.
VXML Gateway traffic if Unified CVP is deployed. The traffic includes media file retrieval from
the media server and VXML documents exchanged with the VXML server.
Data traffic between the Finesse agent or supervisor desktop and the application server(s) hosting
Finesse gadgets.
Reporting traffic between the reporting user and the Unified Contact Center Reporting server.
Traffic between Unified Contact Center servers if they are remote from each other. For example, this
type of traffic occurs with clustering over the IP WAN or with multisite and distributed call
processing with PGs remote from the Unified CCE Central Controller.
Voice traffic due to recording and silent monitoring. Depending on the solution, one or two RTP
streams could be sent in order to silently monitor or record the conversation with an agent.
Bandwidth calculations and guidelines are provided in the Cisco Unified Contact Center design guides
available at http://www.cisco.com/go/ucsrnd.
22-16
Chapter 22
Latency
Agents and supervisors can be located remotely from the call processing components and the contact
center. Technically, the delay between the Finesse Server and the Finesse desktop could be very high
because of high time-out values. Long latency will affect the user experience and might cause confusion
or become unacceptable from the user perspective. For example, the phone could start ringing but the
desktop might not be updated until later.
Latency requirements between the contact center and the call processing components, and between the
contact center components themselves, depend on the contact center solutions. For example, the
Unified CCX redundant nodes can be located remotely from each other, with a maximum latency of
80 ms RTT. With Unified CCE, the maximum latency between the Unified CCE components and
Unified CM, or between the Unified CCE components themselves, is greater than 80 ms RTT.
For more details, refer to the Cisco Unified Contact Center design guides available at
http://www.cisco.com/go/ucsrnd.
QoS
Similar to deployments with other Unified Communications components, contact center deployments
require the configuration of Quality of Service (QoS) to prioritize time-sensitive or critical traffic. QoS
marking for voice and voice signaling in a contact center environment is the same as with other Unified
Communications deployments. Traffic specific to the contact center must be marked with specific QoS
markings. For example, some of the traffic for the Unified CCE private network must be marked as
AF31, while other traffic must be marked as AF11. The QoS marking recommendations and QoS design
guidance are documented for each Unified Contact Center solution in their respective Cisco Unified
Contact Center design guides available at http://www.cisco.com/go/ucsrnd.
22-17
Chapter 22
For administration and upgrade purposes, Cisco recommends separate Unified CM clusters for
contact center and non-contact center deployments. If separate clusters are not possible, then Cisco
recommends separate Unified CM subscriber nodes for contact center and non-contact center
applications.
With contact center deployments, Cisco recommends that you do not use a 2:1 redundancy scheme
for Unified CM. Use 1:1 redundancy to provide higher resiliency and faster upgrades.
The integration between Unified CM and Unified CCX, Unified IP IVR, or Unified CCE is done
through JTAPI. The Unified CCX cluster connects to a primary CTI Manager and also has a backup
connection to a secondary CTI Manager. With Unified CCE, the Agent PG connects to only one CTI
Manager. The redundant Agent PG connects to the backup CTI Manager only. If the primary CTI
Manager fails, the primary Agent PG will also fail and trigger the failover.
A single PG can control and monitor agent phones on all Unified CM subscriber pairs in a
centralized deployment, as illustrated in Figure 22-8.
It is possible to integrate multiple Unified CCX deployments with a single Unified CM cluster.
For more details on Unified CM integration, refer to the Cisco Unified Contact Center design guides
available at http://www.cisco.com/go/ucsrnd.
Deployment with One Agent PG and Four Unified CM Subscriber Pairs
PG A
PG B
Primary
Backup
Primary
Backup
Primary
Backup
Primary
Backup
253830
Figure 22-8
Because Unified CVP allows queuing at the edge, deploying Unified CVP instead of Unified IP IVR
could lower the bandwidth requirements for multisite deployments.
Most of the Cisco Unified Contact Center products and components can be installed in a virtualized
environment based on VMware.
Media termination point (MTP) resources might be required in some scenarios. For example, with
Mobile Agents and inbound calls through SIP trunks, MTPs are required for the associated CTI ports
when RFC 2833 is negotiated. MTPs are also required in some scenarios with Unified CVP. With
Unified CCX Extend and Connect, MTPs are required for the associated CTI Remote Device when
RFC 2833 is negotiated.
22-18
Chapter 22
Transcoders might be required. For example, if phones in a WAN- connected location support only
the G.729 codec but Unified CVP is configured for G.711 support, then Unified CM will engage
transcoders. However, an inbound call that arrives from a gateway or Cisco Unified Border Element
can start with G.711 at Unified CVP then later renegotiate to G.729 with the agents without the need
for transcoders.
Some third-party contact center products are also supported with Unified CM. The integration with
Unified CM could be based on JTAPI and could use CTI ports for call treatment and queuing and
CTI route points. To size Unified CM correctly, it is important to have a good understanding of the
call flows and their impact on Unified CM. It is also important to understand how the redundancy is
implemented and whether or not it impacts Unified CM or CTI scalability.
For more detailed design considerations, refer to the Cisco Unified Contact Center design guides
available at http://www.cisco.com/go/ucsrnd.
The maximum number of Unified CCE agents in a single Unified CM cluster depends on the IVR
solution. With Unified IP IVR, CTI route points and CTI ports are used during the call treatment
queuing, which consume Unified CM resources. With Unified CVP, the call treatment and queuing
are typically handled by the VXML Gateway, Unified CVP VXML server, and Unified CVP call
server, with no impact on Unified CM. Therefore, a single Unified CM cluster can support more
agents with Unified CVP than with Unified IP IVR.
The Unified CCE Mobile Agent feature relies on CTI ports and therefore needs additional resources
from Unified CM subscribers. Therefore, Unified CM scalability is reduced when Mobile Agents
are deployed.
With Unified CCE deployments, SIP dialing is supported. With the SIP dialer, each outbound call
is placed directly from the SIP dialer port to the egress voice gateway. The call reaches Unified CM
only when the call is transferred to an agent. Therefore, Unified CM capacity is much higher when
the SIP dialer is used.
When sizing Unified CM, it is also important to account for any additional CTI applications. For
example, some PC clients can control a phone remotely through CTI. Some call recording
applications can also integrate directly with Unified CM through the CTI Manager and can monitor
agent phones, which could require additional resources from Unified CM. For more details, refer to
22-19
Chapter 22
Computer Telephony Integration (CTI), page 9-27, and to the Cisco Unified Contact Center design
guides available at http://www.cisco.com/go/ucsrnd.
Some silent monitoring and recording solutions (such as the silent monitoring and recording feature
based on Unified CM) consume resources from Unified CM, whereas other solutions such as SPAN
or desktop silent monitoring and recording do not.
When Unified CCX uses Cisco Finesse for agent desktop, Cisco MediaSense is utilized for silent
monitoring and recording, and this will consume resources on Unified CM or the voice gateway. For
details, refer to the latest version of the Solution Reference Network Design for Cisco MediaSense,
available at
http://www.cisco.com/en/US/products/ps11389/products_implementation_design_guides_list.html
Again, due to the complexity associated with sizing, all deployments must be sized with the Cisco
Unified Communications Sizing Tool, available to Cisco employees and partners only (with proper
login authentication) at http://tools.cisco.com/cucst
For more details, refer to the Cisco Unified Contact Center design guides available at
http://www.cisco.com/go/ucsrnd.
Increased productivity
Subject-matter experts can use a single platform to reach customers, regardless of device or location.
The Cisco Remote Expert Solution employs industry-leading, high-quality collaboration products and
services supported by a Cisco Validated Design reference architecture and partner ecosystem.
22-20
Chapter 22
For more details on Remote Expert, refer to the Cisco Remote Expert Solution Design Guide, available at
http://www.cisco.com/c/en/us/solutions/enterprise/design-zone/remote_expert.html
22-21
Chapter 22
22-22
CH A P T E R
23
Call monitoring and recording solutions provide a way to monitor and record audio and video calls that
traverse various components in a Unified Communications and Collaboration solution, such as Cisco IP
Phones, Cisco Unified Border Element devices, or Cisco switches. These recordings can then be used by
call centers and other enterprise functions for various purposes such as compliance, transcription, speech
analysis, podcasting, and blogging. This chapter provides an overview of various call recording solutions
available for Cisco Unified Communications and Collaboration solutions for both audio and video calls.
The chapter also outlines basic design considerations for call recording solutions embedded within a
Cisco Unified Communications and Collaboration solution.
Described in:
Revision Date
23-1
Chapter 23
SPAN-Based Solutions
Recording solutions based on a Switched Port Analyzer (SPAN) use the packet sniffing technology for
recording calls. SPAN is a method of monitoring network traffic. When SPAN is enabled on a switch
port or VLAN, the switch sends a copy of all network packets traversing that port or VLAN to another
port where a recording or monitoring server (such as Cisco Unified Workforce Optimization Quality
Management or a third-party recording server, for example) analyzes those packets. It detects and
decodes the VoIP RTP packets embedded in the network traffic and stores them as audio on a storage
device. SPAN can be enabled on the ports connected to a Cisco Voice Gateway or Cisco IP Phones, as
required. For example, for recording internal calls between IP phones, SPAN should be enabled on
switch ports connected to the IP phones.
Figure 23-1 illustrates a SPAN-based recording solution deployment for recording internal calls. The
ports marked as source ports connected to IP phones are mirrored to the destination port connected to
the recording server.
Figure 23-1
Voice
V Gateway
Cisco Unified CM
WAN
Destination Port
Source Ports
IP Phone
IP Phone
Recording
Server
348767
Switch
IP Phone
Several Cisco partners provide SPAN-based recording servers and applications for Cisco Unified
Communications and Collaboration solutions. For technical details, refer to the specific partner product
information in the Cisco Developer Network Marketplace Solutions Catalog, available at
https://marketplace.cisco.com/catalog/search?utf8=%E2%9C%93&x=48&y=6&search%5Btechno
logy_category_ids%5D=1900
23-2
Chapter 23
In addition, network traffic flow needs to be considered for appropriate bandwidth provisioning when
port mirroring is enabled.
SPAN-Based Recording and Virtualization
This section reviews some common SPAN-based deployments with virtualization enabled and lists some
of the limitations. VMware provides support for the SPAN feature on VMware vSphere Distributed
Switch (VDS) starting with vSphere 5.0.
In a virtualized setup, some of the Unified Communications applications, contact center applications,
and the port analyzer application may be deployed on virtual machines on the same host or on different
hosts. There are some limitations to SPAN-based recording solutions in a virtualized setup. For example,
the following features are not supported for deployments of Cisco Unified Contact Center Enterprise
(Unified CCE) with virtualization:
Note
SPAN-based silent monitoring and recording on Cisco Unified Computing System (UCS) B-Series
chassis
SPAN-based silent monitoring and recording is not supported on the UCS B-Series chassis.
23-3
Chapter 23
Figure 23-2
Voice
Gateway
SIP
Agent
Phone
Agent + Customer
Voice
Supervisor
Phone
PSTN
348822
Caller
23-4
Chapter 23
recording request by pressing the Start Recording button on the endpoint or by sending the recording
request from the JTAPI or TAPI application. To start the recording, Unified CM sends the request to the
forking device to fork the media of the conversation to the recording server, where the media is recorded.
Cisco
Unified
IP Phone
Voice
Gateway
Cisco
Jabber
Softphone
SIP
End User
Voice
PSTN
SIP
Caller
Voice
Mobile
Remote
Destination
348821
Recording
Server
Caller
For a list of Cisco Unified IP Phones that support call monitoring and recording with Unified CM, refer
to the Unified CM Silent Monitoring/Recording Supported Device Matrix, available at
https://developer.cisco.com/site/uc-manager-sip/faq/supported/
23-5
Chapter 23
Note
Invoking media forking from a voice gateway produces two RTP streams, and if silent monitoring is
required, the application is responsible for mixing the streams.
Figure 23-4 illustrates the basic setup for Cisco Unified CM network-based recording using gateways.
Cisco Unified CM and the voice gateway are connected through a recording-enabled SIP trunk.
Unified CM registers with the UC Gateway Services API running on the gateway through its HTTP
interface. This enables Unified CM to receive call event notifications for all calls passing through the
gateway and to decide when to start or stop the recording. Depending on the recording option configured,
when a gateway call is connected with an end user on the phone, Unified CM might notify the gateway
immediately to fork the media or wait for the user indication to start the recording before notifying the
gateway. Unified CM notifies the gateway to stop forking the media upon user indication to stop the
recording, or the gateway automatically stops the recording upon call termination. The requests to start
or stop the recording are sent over the HTTP interface using the Extended Media Forking (XMF) API.
23-6
Chapter 23
Figure 23-4
Voice
Gateway
Cisco
Unified
IP Phone
HTTP
Cisco
Jabber
Softphone
SIP
PSTN
Caller
Voice
End User
Voice
SIP
Mobile
Remote
Destination
348769
Recording
Server
Caller
With Unified CM network-based recording with a gateway, the end user phone and the media forking
device (voice gateway) are decoupled. They can register to the same Unified CM cluster (as shown in
Figure 23-4) or to separate Unified CM clusters. Therefore, this solution could be deployed in a
multi-cluster environment such as Cisco Unified CM Session Management Edition (SME). Figure 23-5
illustrates an example of deploying Unified CM network-based recording with SME, where the voice
gateway registers to the SME cluster and the end user phone registers to the leaf cluster. The SME cluster
and leaf cluster are connected by a SIP intercluster trunk (ICT) with the gateway recording option
enabled on both sides. Thus, the recording invocation requests and responses can be sent between SME
and leaf clusters. Also, customers have the option to deploy the recording server centrally in the SME
cluster with the voice gateway or to distribute the recording servers in all the leaf clusters.
23-7
Chapter 23
Figure 23-5
Media Between
Caller and End User
SME Cluster
SIP
Intercluster
Trunk
HTTP
Voice
Gateway
Leaf Cluster
End User
Phone
PSTN
Caller
Voice
End User
Voice
Caller
SIP
SIP
Central
Recording
Server
Branch
Recording
Server
348770
SIP
This solution is supported on a variety of platforms; for example, Cisco Unified Border Element
running on a Cisco Integrated Services Router (ISR) G2. For detail requirements, refer to the latest
version of the Feature Configuration Guide for Cisco Unified Communications Manager, available
at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-mana
ger-callmanager/products-maintenance-guides-list.html
Only SIP is supported between the voice gateway and Cisco Unified CM, but SIP proxy servers are
not supported.
For inter-cluster recording, only a SIP trunk is supported to interconnect the clusters.
23-8
Chapter 23
Cisco MediaSense
Cisco MediaSense is a SIP-based network application that provides voice and video media recording
capabilities for Cisco Collaboration devices. It is fully integrated into the Unified CM architecture and
can record calls that traverse appropriately configured Unified CM IP phones, gateways, or Cisco
Unified Border Element devices by invoking media forking capabilities on the IP phones and Cisco
Unified Border Element devices. In addition, an IP phone user or SIP endpoint device may call the Cisco
MediaSense system directly in order to leave a recording that consists of only media generated by that
user. Such recordings may include video as well as audio, thus offering a simple and easy method for
recording video blogs and podcasts. While the recording is in progress, it can also be streamed live using
the built-in media player or a third-party media player such as VLC or Apple QuickTime. Cisco
MediaSense uses an HTTP interface to access and play back recordings or perform live streaming.
Note
Most but not all Cisco Unified IP Phones support media forking. Those that do not support media forking
cannot be used for phone-based recording. For a list of IP phones that support phone-based media
recording with Cisco MediaSense, refer to the latest version of the Solution Reference Network Design
for Cisco MediaSense, available at
http://www.cisco.com/en/US/products/ps11389/products_implementation_design_guides_list.html.
Cisco MediaSense also provides an administration and reporting interface to configure the cluster and
manage recordings. It does not currently support secure media relay using sRTP or other means.
Secondary Provides high availability for the database as well as both database and media
operations.
Expansion Provides additional capacity for media operations but no data operations. It is not
supported in Cisco UCS-E deployments.
When deploying multiple Cisco MediaSense clusters, Cisco recommends partitioning the IP phones
carefully among the various clusters so that each IP phone gets recorded by only one cluster.
Note
SIP proxy servers are not supported between Cisco MediaSense and Unified CM or Cisco Unified
Border Element.
Cisco MediaSense integrates with the following Cisco Collaboration technologies to capture the media:
Cisco Unified CM network-based recording (with gateway or built-in bridge) for audio call
recording
Cisco Unified Border Element media forking for video and audio call recording
23-9
Chapter 23
Integration of Cisco MediaSense and Unified CM Network-Based Recording Using Built-in Bridge
End User Phones
Media Between
Caller and End User
Cisco
Unified
IP Phone
Voice
Gateway
Cisco
Jabber
Softphone
SIP
End User
Voice
SIP
PSTN
Caller
Voice
Mobile
Remote
Destination
348823
Cisco MediaSense
Caller
Cisco MediaSense can be utilized to perform compliance or on-demand recording through Unified CM
configuration. Under the control of Unified CM, call recording media can be sourced from either a Cisco
Unified IP Phone, TDM gateway, or Cisco Unified Border Element that connects with Unified CM over
a SIP trunk. Depending on the call flow and call participants, Unified CM dynamically selects the device
to source the media to be captured. Unified CM network-based recording enables Unified CM to route
recording calls, regardless of device, location, or geography.
Figure 23-7 illustrates the basic deployment of Unified CM network-based recording using a gateway
with Cisco MediaSense. For more information about Unified CM network-based recording, refer to the
section on Cisco Unified CM Network-Based Recording with a Gateway, page 23-6.
23-10
Chapter 23
Figure 23-7
Media Between
Caller and End User
Cisco Unified CM
End-User Phone
End-U
HTTP
Cisco Unified
Border Element
SIP
End-User
Voice
PSTN
SIP
Caller
Voice
348784
Cisco MediaSense
Caller
Integration of Cisco MediaSense with Cisco Unified Border Element Media Forking
With this integration, Cisco MediaSense can be utilized to perform compliance recording. When a call
going through Cisco Unified Border Element (CUBE) is connected, CUBE can be configured to fork the
media to Cisco MediaSense and thus provides the ability to capture the end-to-end conversation from a
caller's perspective, no matter how the call traverses through the enterprise. Figure 23-8 illustrates a
basic Cisco MediaSense deployment using Cisco Unified Border Element media forking for call
recording. In this configuration, Cisco MediaSense is directly integrated with Cisco Unified Border
Element for media forking, and the media forking control messages are sent between the two
components without involving Unified CM. The Cisco Unified Border Element device does media
forking by means of a recorder profile configuration attached to one or more dial peers. Cisco
recommends attaching the recording profile to the outbound dial peer.
23-11
Chapter 23
Figure 23-8
Integration of Cisco MediaSense with Cisco Unified Border Element Media Forking for
Call Recording
Voice Between
Caller and End User
Cisco Unified CM
End-User Phone
End-U
SIP
Cisco Unified
Border Element
S
SIP
End-User Voice
Caller Voice
PSTN
Caller
348785
Cisco MediaSense
If the caller and end user are connected via the video-enabled endpoints in which the call traverses
through Cisco Unified Border Element (CUBE), CUBE can be configured to fork both the audio and
video of the caller and end user to Cisco MediaSense for recording. Figure 23-9 illustrates the Cisco
MediaSense deployment using Cisco Unified Border Element media forking for recording in a
video-enabled contact center. In this deployment, once the caller and agent are connected on the video
call, CUBE forks the caller and agent's audio and video to Cisco MediaSense for capturing.
23-12
Chapter 23
Figure 23-9
Integration of Cisco MediaSense with Cisco Unified Border Element Media Forking for
Audio and Video Recording
Contact Center
Application
CTI
Cisco
Unified CM
SIP
Voice and
Video between
Caller and Agent
Caller
Cisco Unified
Border Element
Agent
Agent Voice
and Video
Cisco MediaSense
Note
348786
Caller Voice
and Video
Recording using Cisco Unified Border Element media forking is supported only for SIP-to-SIP call
flows. Cisco Unified Border Element software with media forking runs only on Cisco Integrated
Services Router (ISR) platforms. Media forking is not supported on Cisco Aggregation Services Routers
(ASR).
Any requirements around call recording need to be considered when doing capacity planning for Cisco
Unified Border Element devices because they require additional DSP resources and memory resources.
For the memory requirements, Cisco recommends provisioning the Cisco Unified Border Element
devices with the maximum amount of memory when enabling call recording. Also, media forking
increases bandwidth usage on the link between the Cisco Unified Border Element and the Cisco
MediaSense server. The percentage of calls getting recorded needs to be factored in when calculating
bandwidth requirement.
For details on configuring Cisco Unified Border Element devices to enable network-based recording,
refer to the section on Network-Based Recording Using Cisco UBE in the Cisco Unified Border Element
Protocol-Independent Features and Setup Configuration Guide, available at
http://www.cisco.com/en/US/docs/ios-xml/ios/voice/cube_proto/configuration/15-mt/cube-proto-1
5-mt-book.html
23-13
Chapter 23
Agent Desktop
Agent desktop monitoring and recording solutions are specific to contact center deployments that enable
supervisors to do silent monitoring and initiate call recording when needed. Several agent desktop
monitoring and recording solutions are available, such as:
Cisco Unified Contact Center Enterprise Solution Reference Network Design, available at
http://www.cisco.com/en/US/products/sw/custcosw/ps1844/products_implementation_design_gui
des_list.html
Cisco Unified Contact Center Express Solution Reference Network Design, available at
http://www.cisco.com/en/US/products/sw/custcosw/ps1846/products_implementation_design_gui
des_list.html
23-14
Chapter 23
servers in a cluster must be located at the same physical site, within a network round-trip time (RTT) not
exceeding 10 ms to the Network Attached Storage (NAS) and Structured Query Language (SQL)
servers. Figure 23-10 illustrates TelePresence Content Server clustering.
Note
In Cisco TelePresence Content Server versions prior to 6.2, the clustering option supports H.323
protocol only. For support of SIP registration and SIP calling, ensure that you are running Cisco
TelePresence Content Server version 6.2 or later. For additional clustering requirement details, refer to
the latest version of the Cisco TelePresence Content Server Administration and User Guide, available at
http://www.cisco.com/en/US/products/ps11347/prod_maintenance_guides_list.html.
Figure 23-10
Centralized
Database
Load Balancer
Node 2
Node 1
Network Attached Storage
Cisco TelePresence
Content Server Cluster
348984
Cisco Unified
Communications
Manager
Cisco TelePresence Content Server registered to Cisco Video Communication Server (VCS)
This deployment relies on H.323 registration to VCS rather than a SIP trunk to Unified CM.
23-15
Chapter 23
Figure 23-11 illustrates a sample deployment where the Cisco TelePresence Content Server is trunked
to Cisco Unified CM. The Cisco TelePresence System endpoint and Cisco EX90 are also registered to
Unified CM. A VBrick Rev is used to transcode the video recordings and publish to the VBrick Rev
Portal. The TelePresence Content Server records the call between the two endpoints by joining the
TelePresence bridge. It then sends the recorded video to the VBrick Rev transcoder by means of Secure
File Transfer Protocol (SFTP). The VBrick Rev transcodes the video and publishes it to the VBrick
Portal application.
Deployment of Cisco TelePresence Content Server with Cisco Unified CM
Cisco TelePresence
System Endpoint
Cisco TelePresence
Content Server
Cisco EX90
Cisco Unified
Communications
Manager
VBrick Rev
348985
Figure 23-11
For details on the Cisco Telepresence Content Server, refer to the latest version of the Cisco
TelePresence Content Server Administration and User Guide, available at
http://www.cisco.com/en/US/products/ps11347/prod_maintenance_guides_list.html
23-16
Chapter 23
With Unified CM call recording, each recorded call adds two calls to the call processing component
BHCA capacity. Forking media from an IP phone or voice gateway consumes resources from
Unified CM or the voice gateway, respectively.
Bandwidth requirements increase when media forking is enabled on IP phones or Cisco Unified
Border Element devices to send forked media to the recording server. In case of agent desktop
monitoring and recording, the bandwidth utilization can be bursty, depending on how many calls are
being monitored or recorded at a given time.
DSP resource utilization is impacted on Cisco Unified Border Element when doing media forking
to Cisco MediaSense.
Memory utilization on Cisco Unified Border Element increases for each call that is recorded.
In cases where CTI applications interact with Cisco Unified CM to invoke recording and
monitoring, you should consider the Unified CM cluster deployment model and load-balance the
CTI applications across the cluster.
Due to the complexity associated with sizing, all deployments must be sized with the Cisco Unified
Communications Sizing Tool, available to Cisco employees and partners only (with proper login
authentication) at
http://tools.cisco.com/cucst
23-17
Chapter 23
23-18
PART
Network Management
CH A P T E R
24
Once the network, call routing, call control infrastructure, and applications and services have been put
in place for your Cisco Unified Communications and Collaboration System, network and application
management components can be added or layered on top of that infrastructure. There are numerous
applications and services that can be deployed in an existing Cisco Unified Communications and
Collaboration infrastructure to monitor and manage the operations of the system. These applications and
services can be classified into four basic areas:
User and Device Provisioning Services Provide the ability to centrally provision and configure
users and devices for Unified Communications and Collaboration applications and services.
Voice Quality Monitoring and Alerting Provide the ability to monitor on an ongoing basis various
call flows occurring within the system to determine whether voice and video quality are acceptable
and to alert administrators when the quality is not acceptable.
Operations and Fault Monitoring Provides the ability to centrally monitor all application and
service operations and to issue alerts to administrators regarding network and application failures.
Network and Application Probing Provides the ability to probe and collect network and
application traffic information at various locations throughout the deployment and to allow
administrators to access and retrieve this information from a central location.
This part of the SRND covers the applications and services mentioned above. It provides an introduction
to the various network management applications and services, followed by discussions surrounding
architecture, high availability, capacity planning, and design considerations. The discussions focus on
design-related aspects of the applications and services rather than product-specific support and
configuration information, which is covered in related product documentation.
This part of the SRND also contains detailed information on how to size a Cisco Unified
Communications and Collaboration deployment as well as some recommended methods for migrating
from third-party and legacy communications systems to a Cisco Unified Communications and
Collaboration System.
24-1
Chapter 24
Architecture
Architecture
As with other network and application technology systems, operations and serviceability applications
and services must be layered on top of the underlying network, system, and application infrastructures
in order to be able to monitor and control these infrastructures. Unified Communications and
Collaboration operations and serviceability services such as user and device provisioning, voice and
video quality monitoring and altering, operations and fault monitoring, and network and application
probing, all rely on the underlying network infrastructure for network connectivity for various operations
and serviceability applications and probes. While there is no direct reliance on the Unified
Communications and Collaboration call routing, call control infrastructure, or Unified Communications
and Collaboration clients and services, these infrastructures and applications are what the various
operational and management services actually manage and configure. For example, user and device
provisioning services as well as various monitoring and alerting services leverage the network
infrastructure for connectivity to various Unified Communications and Collaboration applications and
service nodes in order to configure and monitor various components and operations. These same services
also communicate directly with, and in some cases change configurations on or receive alerts from,
components such as call processing agents, PSTN and IP gateways, media resources, endpoints, and
various Unified Communications and Collaboration applications for messaging, rich media
conferencing, and collaboration clients. In addition to relying on these infrastructure layers and basic
Unified Communications and Collaboration services and applications, services pertaining to operations
and serviceability are also often dependent upon each other for full functionality.
24-2
Chapter 24
High Availability
As with network, call routing, and call control infrastructures and critical Unified Communications and
Collaboration applications and services, operations and serviceability services should be made highly
available to ensure that required provisioning, monitoring, and altering will continue even if failures
occur in the network or applications. It is important to understand the various types of failures that can
occur as well as the design considerations around those failures. In some cases, the failure of a single
operations and management application or server can impact multiple services because the Unified
Communications and Collaboration operations and serviceability components are dependent on other
components or services. For example, while the various application service components of a network
management deployment might be functioning properly, the loss of network connectivity to, or a failure
of, a network probe would effectively eliminate the ability to monitor network health or voice and video
quality unless redundant network probes had been deployed along with alternate paths of connectivity.
For operations and serviceability functions such as user and device provisioning, high availability
considerations include temporary loss of functionality due to network connectivity or application server
failures resulting in the inability of administrators to provision users and devices or to make changes to
those user accounts or device configurations. In addition, failover considerations for these types of
operations include scenarios in which portions of the functionality can be handled by a redundant
operation or management application that allows administrators to continue to facilitate some
configuration changes in the event of certain failures.
High availability considerations are also a concern for operations and serviceability applications that
provide services such as voice and video quality monitoring or application and operations fault
monitoring. Interrupted network connectivity or server or application failures will typically result in a
reduced ability to monitor and/or alert, and in some cases complete loss of such functionality. For voice
and video quality monitoring, this can mean that quality measurements for some call flows or devices
will be unavailable. For operations and fault monitoring services, high availability considerations
include the potential for loss of operational change tracking data or fault alerts and indications.
Capacity Planning
Network, call routing, and call control infrastructures as well as Unified Communications and
Collaboration applications and services must be designed and deployed with an understanding of the
capacity and scalability of the individual components and the overall system. Similarly, deployments of
operations and serviceability components and services must also be designed with attention to capacity
and scalability considerations. When deploying various operations and serviceability applications and
components, not only is it important to consider the scalability of these applications themselves, but you
must also consider the scalability of the underlying infrastructures. Certainly the network infrastructure
must have available bandwidth and be capable of handling the additional traffic load these operations
will create. Likewise, the call routing and control infrastructure must be capable of handling required
inputs and outputs as facilitated by the various operations and serviceability components in use. For
example, with operational applications and services such as voice quality monitoring and alerting and
operations and fault monitoring, there are capacity implications for each of these individual applications
or services in terms of the number of devices and call flows that can be monitored at a given time, but
just as important is the scalability of the underlying infrastructure and monitored applications to handle
the added network traffic and connections required for monitoring and alerting. While the monitoring
and alerting application or service itself may be able to support the monitoring of many network devices
and call flows, the underlying network or devices might not have available capacity to handle the probing
connections or the alarm messaging load generated by these monitoring and alerting services.
24-3
Chapter 24
Capacity Planning
For operation applications or services that provide user or device provisioning capabilities, capacity
planning considerations include things such as ensuring that the provisioning application can handle the
requested load and also that user or device provisioning operations not only do not exceed the number
of support devices or users for a particular underlying unified communications application or service,
but also that provisioning or configuration change transactions do not exceed either the capacity of the
underlying network or the rate at which a particular application can handle transactions. In most cases
additional capacity can be added by increasing the number of operational provisioning application
servers or by increasing the size or number of underlying Unified Communications and Collaboration
applications or service instances, assuming the underlying network and call routing and control
infrastructures are capable of handling this additional load.
For a complete discussion of system sizing, capacity planning, and deployment considerations related to
sizing, refer to the chapter on Collaboration Solution Sizing Guidance, page 25-1.
24-4
CH A P T E R
25
This chapter describes system sizing for Cisco Collaboration products and systems. Sizing involves
providing an accurate estimate of the required hardware platforms for the system, based on the number
of users, traffic mix, traffic load, and features that the system will provide.
Accurate sizing is critical to ensure that the deployed system will meet the expected service quality for
call volumes and throughput. For standalone products, manual calculation of the system size may be
feasible (as covered in the section on Sizing for Standalone Products, page 25-48). However, there are
many sizing factors to consider in a complex system deployment. For example, multiple products may
be distributed across different locations and may include video endpoints, call centers, and voice/video
conferencing. Cisco Systems provides a set of sizing rules to handle the resulting complexity.
This chapter provides a general introduction to system sizing methodology and the factors that affect
sizing, and also provides information about how to use the sizing tools.
Note
This chapter should be read in conjunction with the product descriptions and design and deployment
considerations covered in other chapters of this document. A good understanding of both of these aspects
is required for a successful deployment.
This chapter includes the following major sections:
Note
For simplified sizing guidance without the use of the Collaboration Sizing Tool, refer to the latest version
of the Cisco Preferred Architecture for Enterprise Collaboration CVD, available at
http://www.cisco.com/go/cvd/collaboration.
25-1
Chapter 25
Described in:
Revision Date
Performance Testing
Each product performs a set of functions, and each function utilizes a number of resources (such as CPU
and memory). Cisco defines and executes performance tests that allow us to measure resource usage
accurately for each function at different usage levels.
Most systems exhibit linearity within a certain range, beyond which the system performance can become
unpredictable. Cisco sets the usage levels for each performance test to identify and confirm the linear
range of the resource usage for each function. The results for each test can be graphed using a minimal
number of data points. If required, additional data points (at intermediate load levels) are obtained in
order to define the actual system behavior.
The slope of the linear section of the graph defines the resource usage and/or cost for each incremental
addition of work. The R2 value is used to estimate the closeness of the fit. If the R2 value is close to 1,
the formula is a close match for the data.
For example, Figure 25-1 shows the results of a test conducted to determine the memory requirements
for configuring single-line IP phones. It shows the memory consumed by configuring 1500, 4500, and
7500 single-line IP phones in Unified CM. The graph shows that the equation of the trend line is linear
and can be used to predict the dependent variable (in this case, memory) based on the control variable
(the number of phones).
25-2
Chapter 25
In this particular test, the R2 value is extremely close to 1. From the equation, we can compute that the
memory consumed with configuration of 7,500 one-line phones is approximately 519,000 Kbytes and
that each additional line configured for an endpoint in the system consumes an additional 8.91 Kbytes.
Figure 25-1
520000
y = 8.9135x + 452394
R2 = 0.9991
510000
500000
490000
480000
470000
460000
0
2000
4000
6000
8000
284680
System Modeling
Cisco uses the performance test results to create a system model. A system model is a mathematical
model that calculates the maximum resource usage for a specified set of features, endpoints, and traffic
mix, which are provided as inputs to the model.
To develop a system model for a given product, Cisco performs the following steps:
1.
Itemize all of the functions that the product performs. Identify variations of the function that need
to be tested. For example, each type of call will potentially use a different amount of the measured
resources.
2.
Determine the resources of interest. Generally this includes memory and CPU. Specific products
may have additional resources that impact system sizing.
3.
Run the performance tests (as described in the previous section) to determine the resource usage for
each function.
4.
For each function, use the linear range to define the formula for resource usage.
We may need to repeat these steps a number of times because other factors (such as software release,
call mix, and types of endpoints) can impact resource usage.
25-3
Chapter 25
The system model for the product consists of aggregating the formulas for each function supported by
the product. The model can be fairly simple for some products, but it can be very complex for a product
that supports multiple functions, multiple endpoint types, and multiple call types.
Specific considerations for memory and CPU resource types are described in the following sections.
Static memory
Static memory is consumed even when there is no traffic on the system. Static memory usage includes
the data for system configuration and the data for registered endpoints. Static memory also includes
configuration for the dial plan (which covers items such as partitions, translation patterns, route lists and
groups). In addition, static memory includes the memory allocated for CTI and other applications. In a
large system, static memory is mainly a function of the number of configured endpoints and the size of
the dial plan.
Note that each type of endpoint may consume a different amount of memory. Memory usage may also
depend on the device protocol (SIP or SCCP), the number of line appearances, security capabilities, and
other factors. Each of these variants must be measured and incorporated into the model.
Dynamic memory
Dynamic memory is used for transient activities, such as saving the context of each active call. In a large
system, dynamic memory is primarily a function of the number of concurrent calls.
The number of concurrent calls is determined by the average call holding time (ACHT). A longer ACHT
results in more dynamic memory use because there will be a larger number of concurrent active calls.
Memory usage may vary considerably for different types of calls and different protocols (such as SCCP
and SIP).
System memory
System memory is required by the operating system (OS) and by other processes and services. In
addition, some memory may be reserved for transient spikes in usage. System memory reduces the
amount of memory available for applications running on the platform.
25-4
Chapter 25
CPU usage will vary substantially depending on the actual hardware platform. Therefore, the same
performance tests must be repeated on all platforms that are supported for each product.
CPU usage is also affected by CPU-intensive call operations such as call transfers, conferences, and
media resource functions such as MTP or music on hold. Shared lines consume additional CPU
resources, because each call to a shared line is offered to all of the phones that share the line.
Traffic Engineering
Cisco uses industry-standard traffic engineering models to estimate the dynamic load on the system.
Traffic engineering provides mathematical models that calculate the maximum traffic level expected for
a set of users. The models also determine the amount of a shared resource (such as PSTN trunks) that is
required to support a given traffic load.
The following sections describe traffic engineering considerations for different types of traffic:
Definitions
Traffic engineering defines the following terms:
Maximum Simultaneous Calls
The maximum number of simultaneous active calls that the system can handle at one time.
Calls per Second
The number of new call attempts that arrive at the system in one second, plus the number of existing calls
torn down during that same one second interval. This unit can be used to define the average calls per
second that the system expects to handle during a busy hour. (This number is equivalent to the busy hour
calls divided by 60.)
This unit can also be used to define the maximum burst of traffic that the system needs to handle.
Busy Hour
The hour in a given 24-hour period during which the maximum total traffic occurs. This hour varies
depending on the organization and the type of traffic. For business voice traffic, the busy hour is
traditionally assumed to be during morning hours (for example, 10 AM to 11 AM).
Busy Hour Call Attempts (BHCA)
The user BHCA represents the average number of calls that a user initiates or receives during the busy
hour. Typically, BHCA will be calculated as the average of the busy hour call attempts from the busiest
30 days of the year). System BHCA is the User BHCA multiplied by the number of users.
25-5
Chapter 25
Blocking Factor
Indicates a grade of service, expressed as the probability that a call will be blocked during the busy hour
due to lack of resources. For example, a blocking factor of 1% indicates that one out of every 100 calls
may be blocked due to lack of resources required to process the call.
Average Call Hold Time
This is the average period of time that the resource is busy. For example, on a voice call the ACHT is the
period of time between call setup and call tear-down when there is an open speech path between the two
parties. A hold time of 3 minutes (180 seconds) is an industry average used for traffic engineering of
voice systems.
Erlang
The Erlang is a measure of traffic load on a system. To calculate Erlangs, multiply calls per hour by the
average holding time (in hours). Resource requirements can be derived from Erlangs by using the
appropriate Erlang model.
The number of Erlangs handled by a resource (such as a trunk group) is equal to the number of
simultaneous calls. The Erlang value is usually averaged over a one-hour period of time.
Erlang B Model
The Erlang B model can determine the number of trunks required to handle a traffic load (in Erlangs)
with a specified blocking factor. The Extended Erlang B model includes the modeling of retries (for calls
that are blocked). The retry percentage is an additional input to the Extended Erlang B model.
Erlang C Model
The Erlang C model incorporates queuing of incoming calls, and is therefore very useful for modeling
call center traffic.
Bursty Traffic
Traffic models assume a fairly steady arrival rate for the call attempts, which is a valid assumption for a
large number of subscribers acting independently. However, in a real system, a number of calls could
arrive over a very short period of time. Such a traffic burst will consume the system resources very
quickly, and can result in a high number of blocked calls. Products may specify the size and duration of
traffic bursts that they can handle.
Voice Traffic
Standard voice traffic is characterized by specifying the busy hour call attempts (BHCA) and the average
call holding time (ACHT). For example, if the system BHCA is 200 and the average call duration is
3 minutes, the system is being used for a total of 600 minutes, which is 10 Erlangs.
To calculate the usage of a shared resource (such as a PSTN trunk group), the blocking factor must also
be specified. For example, given an Erlang value and the blocking factor, we can use an Erlang calculator
or lookup tables to calculate the number of voice circuits that will be required on PSTN gateways.
Table 25-2 illustrates the relationship between number of trunks, blocking probability, and Erlangs of
traffic.
25-6
Chapter 25
Table 25-2
Number of
Erlangs
Blocking Probability
0.05%
1%
2%
3%
4%
5%
10
19
18
17
16
15
15
20
32
30
28
27
26
26
30
44
42
39
38
37
36
Given an Erlang requirement of 20 and a blocking factor of 1%, the system will need 30 circuits.
Additional circuits are required to provide a lower blocking factor (such as 1%) than to provide a
higher blocking factor (such as 5%).
Note
For additional information about Cisco Unified Contact Center deployments, refer to the Cisco Unified
Contact Center Enterprise SRND, available at
http://www.cisco.com/en/US/products/sw/custcosw/ps1844/products_implementation_design_guides_l
ist.html.
25-7
Chapter 25
Video Traffic
Point-to-point video traffic demonstrates similar characteristics to its voice equivalents for call arrival
rates, peak usage times, and call durations. Also, signaling for call setup and take-down is similar to
voice calls.
Video traffic requires significantly higher network bandwidth than voice because the payload in video
packets is much larger than in voice packets. Also, video traffic can be much burstier than voice. Voice
packet sizes are usually fairly consistent (specifics depend on the encoding algorithm in use), whereas
video frames can vary considerably in size, depending on how much change has occurred since the
previous frame. The resulting RTP packet stream can therefore exhibit bursts of traffic.
Implications for video conferencing are covered in the next section.
Call arrivals
A traditional traffic model assumes a Poisson distribution of busy-hour call arrivals throughout the
busy hour. However, most participants join their conference call within 5 to 10 minutes of the
meeting start time, and most conference calls are scheduled to start at the beginning of the hour.
Therefore, the call arrival rate will exhibit a single burst at the top of the hour rather than a Poisson
distribution throughout the hour.
Peaks
Business voice traffic typically has a distinct peak in the morning (between 10:00 and 11:00 AM)
and another peak in the afternoon (between 1:00 and 2:00 PM). However, conference facilities are
generally a limited resource, resulting in meetings that are distributed more evenly throughout the
business day, with less of a pronounced peak at peak times.
Call durations
The average business voice call duration is 3 minutes. The average conference call duration may be
closer to 50 minutes (depending on the mix of 30 minute, 60 minute, and longer meetings).
Video conferencing
Specialized equipment is required to provide the switching or combining of video streams.
Therefore, expected usage of video endpoints is an important factor in the model.
Sizing a deployment for conferencing primarily involves deciding how many concurrent connections are
required. For example, sizing for TelePresence Servers would include the following considerations:
Geographical location Each region served by Unified CM should have dedicated conferencing
resources.
25-8
Chapter 25
Conference resources are generally dedicated to a region in order to keep as much of the conference
media on the regional network; therefore, sizing can be considered on a region-by-region basis.
Cluster sizes
A major design decision is whether to create a large centralized Cisco Unified CM cluster or to
create a cluster at each major location. The central cluster may have a higher utilization, but you
may be forced into a second cluster if a cluster limit is exceeded.
Some system limits are not absolute and can change dynamically based on the sizing of other
services configured in the system.
Server capabilities
Each type of server or router supports different capabilities. For example, more powerful servers
might have a higher number of network ports compared to Cisco Business Edition 6000 platforms
or a Cisco Integrated Services Router (ISR).
As another example, different models of Cisco Integrated Services Routers have restrictions on the
number and types of network modules or Cisco Unified Computing System (UCS) E-Series blade
servers they can host.
25-9
Chapter 25
System release
System resource usage can vary between system releases. Sometimes, new capabilities in a release
can cause an increase in resource usage. In other cases, software improvements can result in a
decrease in resource usage.
Because of all the factors and possible variations, the accurate sizing of a large system deployment is a
complex undertaking. For this reason, Cisco strongly recommends using the system sizing tools
described in the following sections.
25-10
Chapter 25
Tool-specific sections below contain explanations of the inputs required for the tool and how the inputs
can best be collected from an existing system or estimated for a system still in the design stage.
Obviously, the sizing recommendations generated by the tools are only as accurate as the input data you
provide.
Cisco provides the following sizing tools:
Cisco Unified Communications Manager Session Management Edition (SME) Sizing Tool
This is a specialized tool that focuses on the specific functions of a Unified CM Session
Management Edition deployment.
For more information on these tools and their access privileges, refer to the Unified Communications
Sizing Tool Frequently Asked Questions (FAQ), available at
http://tools.cisco.com/cucst/help/ucst_faq.pdf
Caution
If any parameter of your system design exceeds the range of values that the above sizing tools allow you
to enter, consult your Cisco account team or a Cisco Systems Engineer (SE) about your design before
proceeding further.
In addition to these sizing tools, a Virtual Machine Placement Tool is available to Cisco partners and
customers with a valid login account. The Virtual Machine Placement Tool is a graphical tool that allows
you to select Tested Reference Configurations (TRC) or specifications-based hardware, and to drag and
drop the various Cisco Collaboration application virtual machines on those servers. Some placeholders
representing third-party application virtual machines are also available when deploying Cisco
Collaboration applications co-resident with third-party applications. The sizing tools determine how
large the servers need to be and how many virtual machines are necessary. This information can then be
entered as an input to this Virtual Machine Placement Tool in order to determine how to place the various
virtual machines and to determine how many servers would need to be deployed. Even though some of
the co-residency rules are implemented in the tool, Cisco recommends verifying the rules by using the
guidelines documented at
http://docwiki.cisco.com/wiki/Unified_Communications_Virtualization_Sizing_Guidelines
The Virtual Machine Placement Tool is available (with proper login authentication) at
www.cisco.com/go/vmpt
25-11
Chapter 25
The various types of trunk interfaces that the cluster services. The following trunk protocols are
supported by the SME; however, Cisco recommends SIP trunks as the preferred protocol:
SIP
H.323
MGCP (Q.931)
SIP (Q.SIG)
H.323 Annex M1
MGCP (Q.SIG)
The number of users that access SME cluster services through each type of trunk interface
BHCA per user for each trunk interface to leaf clusters for intercluster calls
BHCA per user for each trunk interface to leaf clusters for off-net (PSTN) calls
The type of trunk interface used by the SME cluster to connect to the PSTN
If the SME acts as a service aggregation point, you must consider the following additional sizing
parameters:
For centralized voice messaging, the percentage of calls that are sent to voice mail
For mobility, the number of users and the remote destinations per user
The performance of the SME is measured as calls-per-second across each pair of protocols. There are
variations across the hardware platforms and software versions.
25-12
Chapter 25
25-13
Chapter 25
Each cluster can support configuration and registration for a maximum of 40,000 secured or
unsecured SCCP or SIP phones.
Two TFTP server nodes are required, in addition to a dedicated publisher, if the number of endpoints
in the cluster exceeds 1,250.
Support for CTI connections has improved over the last several releases, and each cluster can
support a maximum of 40,000 CTI connections.
The number of call processing subscribers in a cluster cannot exceed 4, plus 4 standby, for a total of
8 call processing subscriber nodes. Also, the total number of server nodes in a cluster, including the
publisher, TFTP, and media servers, may not exceed 21 servers as the maximum allowed in a cluster.
The name of a Unified CM virtual machine (VM) configuration corresponds to the maximum
number of users, assuming that on average, each user has one phone. If this is not the case, the VM
configurations would indicate the maximum number of endpoints registered to a Unified CM node.
For example, a 10k-user VM configuration supports a maximum of 10,000 users, assuming one
device per user. However, if you plan to deploy multiple devices per user, then the maximum number
of supported users is reduced. For example, if you have 2 devices per user, then the 10k-user VM
configuration would support a maximum of 5,000 users with 10,000 devices. This same principal
applies for the smaller Unified CM VM configurations as well.
Deployment Options
The following deployment options are overall settings that affect all operations in the system, and they
are independent of how many endpoints are registered or how many calls are in progress.
Database Complexity
The CPU usage is considerably higher when the configuration database in Unified CM is considered to
be complex. There is no one metric to determine whether the database is simple or complex. As a general
rule, the database is complex if you have configured more than a few thousand endpoints and more than
a few hundred dial plan elements such as translation and route patterns, hunt pilots, and shared lines.
Number of Regions and Locations
Configuration of regions and locations in the Unified CM cluster requires both database and static
memory. The number of gateways that can be defined in the cluster is also tied to the number of locations
that can be defined. Table 25-3 lists these limits for some of the Unified CM VM configurations.
Table 25-3
VM Configuration
Maximum Number
of Regions
Maximum Number
of Locations
Maximum Number of
Trunks and Gateways
1,000
1,000
1,100
2,000
2,000
2,100
25-14
Chapter 25
Whether or not you can actually define the maximum number of locations and regions in a cluster
depends on how "sparse" your codec matrix is. If you have too many non-default values in the
inter-region codec setting, you might not be able to scale the system to its full capacity for regions and
locations. As a general rule, the change from default should not exceed 10% of the maximum number.
Call Detail and Call Management Records
Generation of call detail records (CDR) and call management records (CMR) places a heavier burden on
the CPU.
High Availability
After you determine the minimum number of nodes required for the specified deployment, add the
desired number of additional subscriber nodes to provide redundancy. Redundancy options are described
in the chapter on Call Processing, page 9-1.
Number of Virtual Server Nodes per Cluster
You can configure a regular cluster with up to four subscriber pairs. In a distributed topology, there may
be multiple clusters even when none of the clusters has reached the maximum.
For a centralized topology, there is generally one cluster unless the capacity limit is reached. Note that
other system limits might force a new cluster even if the per-node utilization is not at the limit.
Choice of VM Configurations and Hardware Platforms
Cisco provides Open Virtualization Archive (OVA) VM configurations that can be loaded onto a
hypervisor. Different templates specify different capacities. For example, the 10,000 Users template
defines a virtual machine that has a maximum capacity of 10,000 endpoints. There are also templates
defined to support a maximum of 1,000, 2,500, and 7,500 endpoints.
The formal definitions of the VM configurations for Unified CM and other Unified Communications
products are available at the following location:
http://docwiki.cisco.com/wiki/Unified_Communications_Virtualization_Sizing_Guidelines
Specific information for Unified CM is available at the following location:
http://docwiki-dev.cisco.com/wiki/Virtualization_for_Cisco_Unified_Communications_Manager_
(CUCM)
With Unified CM, some of the VM configurations are not supported on the low-end hardware platforms.
To verify which VM configuration is supported on a hardware platform, refer to the documentation at:
http://www.cisco.com/go/uc-virtualized
Hardware and Virtualization Software Requirements
The following requirements are common to all applications. See each application's product
documentation for additional requirements or restrictions.
25-15
Chapter 25
Note
Choice of placement of virtual machines running Unified CM and other Unified Communications
products can have an impact on performance and availability. For a discussion of these and other
considerations for Unified Communications on UCS deployments, refer to the documentation at
http://www.cisco.com/go/uc-virtualized.
Endpoints
The number of endpoints is an important part of the overall load that the system must support. There are
different types of endpoints, and each type imposes a different load on Unified CM. Endpoints can be
differentiated by:
Software-based or hardware
Each endpoint configured in the system uses system resources (such as static memory) just by being
defined and registered. The endpoint consumes CPU and dynamic memory based on its call rate.
An endpoint can also place additional load on the Unified CM by running applications such as CTI that
interact with services running in the Unified CM.
Table 25-4 shows the maximum number of endpoints supported by different VM configuration types.
Note that these values are guidelines only. A given system may support less than these maximum
amounts because of other applications included in the deployment.
Table 25-4
VM Configuration
10,000 Users
10,000
7,500 Users
7,500
2,500 Users
2,500
1,000 Users
1,000
1. These limits represent the maximum number of endpoints that can be configured in the database and registered per virtual
subscriber node. All other registered devices such as media termination points (hardware of software) or SIP trunks do not
count against these limits.
For Cisco Collaboration System Release (CSR) 11.x, the Unified CM deployments require all virtual
nodes to increase their vRAM by 2 GB of memory for the following VM configuration templates:
1,000 users
2 vCPU
6 GB vRAM
80 GB vDisk
25-16
Chapter 25
2,500 users
4 vCPU
6 GB vRAM
80 GB vDisk
7,500 users
2 vCPU
8 GB vRAM
110 GB vDisk
10,000 users
4 vCPU
8 GB vRAM
110 GB vDisk
Device Configuration
When configured in softphone mode, a Jabber Desktop Client configuration file is downloaded
through TFTP or HTTP to the client for Unified CM call control configuration information. In
addition, any application dial rules or directory lookup rules are also downloaded through TFTP or
HTTP to Jabber Desktop Client devices.
25-17
Chapter 25
The Jabber Desktop Client uses the Cisco Unified CM Cisco IP Phone (CCMCIP) service or UDS
service to gather information about the devices associated with a user, and it uses this information
to provide a list of IP phones available for control by the client in deskphone control mode. The
Jabber Desktop Client in softphone mode uses the CCMCIP or UDS service to discover its device
name for registration with Unified CM.
Deskphone Mode
When configured in deskphone mode, the Jabber Desktop Client establishes a CTI connection to
Unified CM upon login and registration to allow for control of the IP phone. Unified CM supports
up to 40,000 CTI connections. If you have a large number of clients operating in deskphone mode,
make sure that you evenly distribute those CTI connections across all Unified CM subscribers
running the CTIManager service. This can be achieved by creating multiple CTI Gateway profiles,
each with a different pair of CTIManager addresses, and distributing the CTI Gateway profile
assignments across all clients using deskphone mode.
Voicemail
When configured for voicemail, the Jabber Desktop Client updates and retrieves voicemail through
an IMAP or REST connection to the mailstore.
Authentication
Client login and authentication, contact profile information, and incoming caller identification are
all handled through a query to the LDAP directory, unless stored in the local Jabber Desktop Client
cache.
Contact Search
There are several contact sources that can be used with the Jabber Desktop Client. For example, the
UDS service can be used by clients to search for contacts in the Unified CM User database.
Alternatively, LDAP integration can be used. If the requested contact cannot be found in the local
Jabber Desktop Client cache, UDS or LDAP contact searches take place.
Client scalability
The Cisco IM and Presence Service VM configuration template determines the number of users a
cluster can support. The Cisco Jabber Client deployment must balance all users equally across all
nodes in the cluster. This can be done automatically by setting the User Assignment Mode Sync
Agent service parameter to balanced. The recommended maximum number of contacts in the
contact list (buddy list) should not be more than 100 per user. The higher the number of contacts per
user, the greater the system impact.
IMAP scalability
The number of IMAP or IMAP-Idle connections is determined by the messaging integration
platform.
25-18
Chapter 25
The Cisco Jabber Clients interface with Unified CM. Therefore, the following guidelines for the current
functionality of Unified CM apply when Cisco Jabber Client voice or video calls are initiated:
CTI scalability
In Desk Phone mode, calls from Cisco Jabber Clients use the CTI interface on Unified CM.
Therefore, observe the CTI limits as defined in the chapter on Call Processing, page 9-1. You must
include these CTI devices when sizing Unified CM clusters.
Codec selection
Cisco Jabber Client voice and video calls utilize codec selection through the Unified CM regions
configurations.
Cisco Unified CM 10.x provides the Security Assertion Markup Language Single Sign-On (SAML SSO)
feature, which enhances the end user experience by allowing users to log in only once to access all
applications within the Cisco Collaboration solution.
SAML SSO provides secure mechanism to use credentials and relevant information of the end user to be
leveraged across multiple Unified Communications applications (such as Unified CM, Cisco Unity
Connection, and IM and Presence). For the SAML Single Sign-On feature to work as expected, the
network architecture must scale to support the number of users for each cluster.
For a Unified Communications deployment across multiple applications (such as Unified CM, Cisco
Unity Connection, and IM and Presence), all SAML requests must authenticate with the Identity
Provider (IdP) for Cisco Jabber clients to login successfully.
Note
25-19
Chapter 25
Cisco Jabber with SAML SSO logins should also be factored into system sizing because the numbers of
users logging into the system in a typical day at the same time could have an impact on the time it takes
for user(s) to log in. This is expected due to the limiting factor of how many requests the system can
process at one time. The current maximum login rate for Jabber users is 2.7 logins per second (about
166 logins per minute) or 10,000 logins within one hour. This is assuming that all users and devices are
evenly distributed across all nodes and that Cisco Jabber is in softphone mode.
There are many interdependent variables that can affect Unified CM cluster scalability (such as regions,
locations, gateways, media resources, and so forth). Therefore it is vital to determine the number of
users, endpoints, and calls per user per hour, to deploy efficiently so that resources are available to
handle the required load.
As an example, consider a deployment with redundant subscriber pairs supporting 5,000 users, each
associated with two devices (desk phone and soft phone). This deployment would require the following
number of virtual machines and VM configurations (assuming high availability and redundancy):
The IM and Presence 5k-user VM configuration pair would support the 5,000 users, and a pair of
Unified CM 10k-user VM configurations would support the 10,000 devices.
Cisco WebEx Messenger service deployment network requirements are available at:
http://www.webex.com/webexconnect/orgadmin/help/17161.htm
25-20
Chapter 25
There are two parameters that are key to Unified CM's capacity to support single number reach (formerly
Mobile Connect) and enterprise two-stage dialing (Mobile Voice Access and Enterprise Feature Access).
For these functions to work appropriately, users must be enabled for mobility and remote destinations
with shared lines must be defined for the users. Table 25-5 shows the limits for users and remote
destinations and mobility identities in a cluster consisting of each class of Unified CM VM
configurations.
Table 25-5
Note
Cluster Nodes
40,000
30,000
10,000
4,000
A mobility-enabled user is defined as a user that has a remote destination profile and at least one remote
destination or a dual-mode device and a mobility identity configured.
Each remote destination and mobility identity defined in the system affects Unified CM in several ways:
The remote destination or mobility identity occupies static memory and configuration space in the
database.
Each occurrence uses a shared line with the users primary device, and hence calls to that line use
more CPU resources.
If the remote destination or mobility identity is an external number (such as the user's cell phone or
home), then gateway resources will be used to extend the call.
25-21
Chapter 25
Call Traffic
The quantity and quality of call traffic is a very significant factor in sizing Unified CM.
It is important to differentiate between call types because call origination and termination are considered
as distinct events in the half-call model. For endpoints registered on the same subscriber node, that
subscriber handles both call halves for calls between these endpoints. For calls made between two
subscriber nodes in the same cluster, each of the participating subscribers will handle either the call
origination or call termination. For calls made between endpoints registered on different clusters, each
cluster will handle only half of each call. For calls made between an endpoint in a cluster and the PSTN,
a PSTN gateway will handle half of the call, and these call types form the basis for sizing the gateways.
For accurate sizing of call traffic, you must consider the following factors:
BHCA from and to the PSTN using MGCP, H.323, and SIP protocols
BHCA from and to other clusters using H.323 intercluster trunks or SIP protocols
Each different type of call takes a different amount of CPU resources to set up. The number of busy hour
call attempts determines the CPU usage. CPU requirements vary directly with the call placement rate.
The ACHT determines the dynamic memory requirements to sustain calls for their duration. A longer
ACHT means that more dynamic memory must remain allocated, thus increasing the memory
requirement.
Call traffic can arise from other sources as well. Each time a call is redirected in a transfer or to
voicemail, it requires processing by the CPU. If a directory number is configured on multiple phones, an
incoming call to that number needs to be presented to all of those phones, thus increasing CPU usage at
call setup time. If advanced features are being used, calls made using this technology, and the percentage
of these calls that need to be redirected to the PSTN because of call quality, must also be accounted for.
Dial Plan
The dial plan in Unified CM consists of configuration elements that determine call routing and
associated policies. In general, dial plan elements occupy static memory space in Unified CM. The
following dial plan elements impact the amount of memory required:
Directory numbers
Shared directory numbers and the average number of endpoints that share the same DN
There are no hard limits enforced by Unified CM for any of the dial plan elements, but there is a fixed
amount of shared system memory available.
Most of the dial plan elements do not have a direct effect on CPU usage. The exception is shared lines,
such as hunt lists and line groups. Each shared line multiplies the CPU cost of a call setup because the
call is presented to all of the endpoints that share a particular directory number.
25-22
Chapter 25
The maximum number of CTI controlled and/or monitored endpoints that can be registered to a
Unified CM subscriber node.
The maximum number of endpoints that a Unified CM subscriber node running the CTI Manager
service can monitor or control.
The maximum number of TAPI/JTAPI application instances that can connect to a Unified CM
subscriber node running the CTI Manager service. The TAPI/JTAPI application instances that can
connect to a Unified CM subscriber node running the CTI Manager service are sometimes referred
as CTI connections.
Note that the maximum number of CTI resources for a VM configuration corresponds to the endpoint
capacity of that VM configuration.
In addition to native applications provided by Unified CM, third-party applications may also be
deployed that use Unified CM CTI resources. When counting CTI ports and route points, be sure to
account for the third-party applications as well.
Table 25-6
VM Configuration
1,000 Users
1,000
2,500 Users
2,500
7,500 Users
5,000 or 7,5001
10,000 Users
10,000
1. 7,500 CTI resources supported with Unified CM 10.5 and later releases; 5,000 CTI resources supported with Unified CM
releases prior to 10.5.
In addition to the maximum number of connections and devices, CTI limits are also influenced by:
The number of lines on each of the controlled devices (up to 5 lines per controlled device)
The number of shared occurrences of a line controlled by CTI (up to 5 per line)
The CTI resources available on Unified CM are reduced if any of these values is exceeded.
25-23
Chapter 25
Step 2
1 to 5 lines
1.0
6 lines
1.2
7 lines
1.4
8 lines
1.6
9 lines
1.8
10 lines
2.0
Note
Step 3
If there are multiple line factors for the devices within a cluster; determine the average line factor
across all CTI devices in the system.
Step 4
1 to 5 applications
1.0
6 applications
1.2
7 applications
1.4
8 applications
1.6
9 applications
1.8
10 applications
2.0
Calculate the required number of CTI resources according to the following formula:
Required Number of CTI Resources = (Total CTI Device Count) (The greater of {the CTI Line Factor
or the CTI Application Factor})
25-24
Chapter 25
IP Phone Services
Cisco Unified IP Phone Services are applications that utilize the web client and/or server and XML
capabilities of the Cisco Unified IP Phone. The Cisco Unified IP Phone firmware contains a
micro-browser that enables limited web browsing capability. These phone service applications provide
the potential for value-added services and productivity enhancement by running directly on the users
desktop phone.
Cisco Unified IP Phone Services act, for the most part, as HTTP clients. In most cases they use
Unified CM only as a redirect server to the location of the subscribed service. Because Unified CM acts
only as a redirect server, there typically is minimal performance impact on Unified CM unless there is a
large number of requests (hundreds of requests per minute or more).
With the exception of IP Phone Services for the integrated Extension Mobility and Unified CM Assistant
applications, IP Phone Services must reside on a separate web server. Running phone services other than
Extension Mobility and Unified CM Assistant on a Unified CM node is not supported.
Creation of EM profiles requires both disk database space and static memory.
The rate at which users may log into their EM accounts affects both CPU and memory usage.
Unified CM nodes have bounds on the maximum number of logins per minute that they can support.
Extension Mobility Cross Cluster (EMCC) has a higher impact on resources. There is a limit on the
number of EMCC users that a Unified CM node can support. The maximum EMCC login rates
supported are lower than those supported for EM. In addition, there is a trade-off between EM and
EMCC login rates. If both are occurring at the same time, then the maximum capacity for each will
be reduced.
25-25
Chapter 25
EM and EMCC login rates per cluster are not simply the login rate of each node multiplied by the
number of nodes in the cluster, because profiles in a shared database have to be accessed. The
maximum login rate in a cluster consisting of more than one call processing subscriber should be
limited to 1.5 times that of a single node.
Table 25-9 shows the maximum number of EM and EMCC logins per minute for each type of VM
configuration.
Table 25-9
VM Configuration
Maximum EM
Login Rate (per
Node)
Maximum EM
Login Rate (Dual
Nodes)
Maximum EMCC
Login Rate (Per
Node)
Maximum EMCC
Login Rate (Dual
Nodes)
Maximum
Concurrent
EMCC Devices
1,000 Users
200
300
60
70
333
2,500 Users
235
352
71
80
833
250
375
75
90
2,500
Cisco Extension Mobility login and logout functionality can be distributed across a pair of subscriber
nodes to increase login/logout cluster capacity. For example, when the EM load is distributed evenly
between two virtual machines with the 7,500-user VM configuration, the maximum cluster-wide
capacity is 375 sequential logins and/or logouts per minute.
Note
The Cisco Extension Mobility service can be activated on more than two nodes for redundancy purposes,
but Cisco supports a maximum of two subscriber nodes actively handling logins/logouts at any given
time.
Note
25-26
Chapter 25
CTI line instance from the CTIManager. When you configure Unified CM Assistant, the number of
required CTI lines or connections must be considered with regard to the overall cluster limit for CTI lines
or connections.
The following limits apply to Unified CM Assistant:
A maximum of 33 Managers can be configured for a single Assistant (if each Manager has one
Unified CM Assistant-controlled line).
A maximum of 3,500 Assistants and 3,500 Managers (7,000 total users) can be configured per
cluster using the 7,500-user or 10,000-user virtual machines
A maximum of three pairs of primary and backup Unified CM Assistant nodes can be deployed per
cluster if the Enable Multiple Active Mode advanced service parameter is set to True and a second
and third pool of Unified CM Assistant server nodes are configured.
In order to achieve the maximum Unified CM Assistant user capacity of 3,500 Managers and 3,500
Assistants (7,000 users total), multiple Unified CM Assistant server pools must be defined. (For more
information, see Unified CM Assistant, page 18-19.)
Cisco WebDialer
Cisco WebDialer provides a convenient way for users to initiate a call. Its impact on Unified CM is fairly
limited because extra resources are required only at call initiation and are not tied up for the duration of
the call. Once the call has been established, its impact on Unified CM is just like any other call.
The WebDialer and Redirector services can run on one or more subscriber nodes within a Unified CM
cluster, and they support the following capacities:
Each WebDialer service can handle up to 4 call requests per second per node.
The following general formula can be used to determine the number of WebDialer calls per second (cps):
(Number of WebDialer users) ((Average BHCA) / (3600 seconds/hour))
When performing this calculation, it is important to estimate properly the number of BHCA per user that
will be initiated specifically from using the WebDialer service. The following example illustrates the use
of these WebDialer design calculations for a sample organization.
Example: Calculating WebDialer Calls per Second
Company XYZ wishes to enable click-to-call applications using the WebDialer service, and their
preliminary traffic analysis resulted in the following information:
50% of all calls are dialed outbound, and 50% are received inbound.
Projections estimate 30% of all outbound calls will be initiated using the WebDialer service.
Note
These values are just examples used to illustrate a WebDialer deployment sizing exercise. User
dialing characteristics vary widely from organization to organization.
25-27
Chapter 25
10,000 users each with 6 BHCA equates to a total of 60,000 BHCA. However, WebDialer deployment
sizing calculations must account for placed calls only. Given the initial information for this sizing
example, we know that 50% of the total BHCA is for placed or outbound calls. This results in a total of
30,000 placed BHCA for all the users enabled for click-to-call using WebDialer.
Of these placed calls, the percentage that will be initiated using the WebDialer service will vary from
organization to organization. For the organization in this example, several click-to-call applications are
made available to the users, and it is projected that 30% of all placed calls will be initiated using
WebDialer.
(30,000 placed BHCA) 0.30 = 9,000 placed BHCA using WebDialer
To determine the number of WebDialer server nodes required to support a load of 9,000 BHCA, we
convert this value to the average call attempts per second required to sustain this busy hour:
(9,000 call attempts / hour) (hour/3,600 seconds) = 2.5 cps
Each WebDialer service can support up to 4 cps, therefore one node can be configured to run the
WebDialer service in this example. This would allow for future growth of WebDialer usage. In order to
maintain WebDialer capacity during a server node failure, additional backup WebDialer server nodes
should be deployed to provide redundancy.
Attendant Console
The integration of Cisco Unified CM with the Attendant Console utilizes CTI resources. The
server-based attendant console monitors the last 2,000 users to whom the attendant sent calls, thus
increasing CTI resource usage. In addition, each call uses a number of CTI route points and ports for
greetings, queuing, and so forth.
Media Resources
Unified CM offers the Cisco IP Voice Media Streaming Application (IPVMS), which provides certain
media functions that are performed in software only and do not require hardware resources. Unified CM
can act as a media termination point (MTP), as a conference bridge, as an annunciator (for playing
announcements), or as a source of music-on-hold streams. Although the capabilities of Unified CM are
limited compared to similar functions provided by Cisco Integrated Service Routers (ISRs), they are
generally the key source of music-on-hold streams (both unicast and multicast).
The Cisco IP Voice Media Streaming Application may be deployed in one of two ways:
Co-resident deployment
In a co-resident deployment, the streaming application runs on any server node (either publisher or
subscriber) in the cluster that is also running the Unified CM software.
Note
The term co-resident refers to two or more services or applications running on the same
server node or virtual machine.
Standalone deployment
In a standalone deployment, the streaming application runs on a dedicated server node within the
Unified CM cluster. The Cisco IP Voice Media Streaming Application service is the only service
enabled on the server node, and the only function of the server node is to provide media resources
to devices within the network.
25-28
Chapter 25
The Cisco IP Voice Media Streaming Application can provide MTP, annunciation, and conferencing
capabilities, but a more scalable design is to place these functions on external Cisco Integrated Service
Routers (ISRs). The music-on-hold functionality of this application is, however, not so easily placed on
external sources. Table 25-10 lists the maximum values that may be configured for each of these
services.
Table 25-10
Default Quantity
Maximum Number of
Streams or Devices
Supported Codecs
Annunciator
48
750
48
256
G.711, L16WB
Music on Hold
250
1,000
48
512
Note
All values represent the number of callers supported per media device. For instance, 48 software
conference bridges can support 16 three-party conferences.
These devices can be co-resident with the call processing nodes when using default settings or near
to default settings.
When increasing capacities to the maximum values, Cisco recommends deploying the media devices
on standalone nodes (not with call processing).
If MoH audio sources are used with initial (greeting) announcements, Cisco recommends keeping
the initial announcements less than 15 seconds in duration, otherwise you might need to reduce the
maximum number of MoH streams per MoH server node to between 500 and 700 due to extra file
I/O.
Each media device may be disabled/enabled via the IPVMS Service Parameter (MoH is on the MoH
device configuration page). It is possible to configure an MoH-only Unified CM node, and so forth.
To calculate the capacities of each of the media functions on the DSPs supported by each individual ISR,
refer to the Cisco ISR product data sheets or to the chapter on Media Resources, page 7-1.
25-29
Chapter 25
Music on Hold
Table 25-11 11 lists the VM configurations and the maximum number of simultaneous music-on-hold
(MoH) streams each node can support. You should ensure that the actual usage does not exceed these
limits, because once MoH maximum stream capacity has been reached, additional load could result in
poor MoH quality, erratic MoH operation, or even loss of MoH functionality. Add additional MoH nodes
(co-resident or dedicated) to increase Unified CM cluster MoH stream capacity.
Table 25-11
Unified CM
OVA Template
1,000 User
500
750
500
500
1,000
1,000
1,000
1,000
2,500 User
7,500 User
750
1,000
10,000 User
1. All capacities based on non-sRTP streams.
As shown in Table 25-12, beginning with Unified CM 10.5(2) you can define a maximum of 500 unique
sources of audio for Music on Hold in a Unified CM cluster. The maximum audio source capacities
shown in Table 25-12 are per-cluster based on the VM configuration size and MoH server type
(co-resident or standalone) used in the cluster. Adding MoH nodes to a Unified CM cluster increases
only MoH stream capacity but does not increase audio source capacity. Audio source capacity can be
increased only by moving from co-resident to standalone MoH nodes, increasing the cluster-wide node
VM configuration size, or adding additional Unified CM clusters.
Table 25-12
Unified CM
OVA Template
1,000 User
100
250
50
250
500
2,500 User
7,500 User
10,000 User
The capacity limits described in Table 25-11 and Table 25-12apply to any combination of unicast,
multicast, or simultaneous unicast and multicast streams.
Performance Considerations
To maximize the number of MoH audio sources and streams, you must reduce the number of some other
media devices, such as disabling software MTPs and/or software conference bridges. The Cisco IP Voice
Media Streaming Application service does not support maximum settings for all the media devices
simultaneously. Oversubscribing the system resources (for example, CPU usage and disk I/O) with
media devices would impact the overall system performance. An IPVMS alarm is issued if a media
device is unable to meet provisioned capacity.
25-30
Chapter 25
For low-end configurations (1,000-user or 2,500-user VM configuration) and MoH co-resident with
moderate call processing, MoH is limited to a maximum of 500 streams, 100 MoH audio sources, and
48 to 64 annunciator streams with MTPs and conference bridges set at default values or disabled.
A dedicated 1,000-user or 2,500-user VM configuration MoH node is required to support 750 MoH
streams with 250 MoH audio sources and 250 annunciator streams.
To support a maximum of 1,000 MoH streams, 500 MoH audio sources, and 750 annunciators, the
minimum requirement is a 7,500-user OVA dedicated standalone MoH server.
Use of sRTP for MoH and/or annunciator will reduce the maximum number of MoH callers by 25%, and
a dedicated IPVMS server for MoH and annunciator is highly recommended in this case.
The Unified CM MoH server supports four codecs: G.711 ulaw, G.711 mulaw, G729a, and Wideband
audio. With unicast MoH, because the codec is negotiated during call setup, the number of MoH streams
depends not on the number of MoH codecs enabled but on the number of endpoints that are on hold with
unicast MoH. In the case of multicast MoH, each multicast-enabled audio source generates one MoH
steam for each MoH codec enabled. For example, if 2 codecs are enabled and all 500 MoH sources are
multicast-enabled, then 1,000 multicast MoH streams would be active even if no endpoints are on hold.
In this scenario, if any endpoints are placed on unicast MoH, then additional MoH streams capacity
would be required.
Impact on Unified CM
Whether deployed in co-resident or standalone mode, the Cisco IP Voice Media Streaming Application
consumes CPU and memory resources. This impact must be considered in the overall sizing of
Unified CM.
In general, usage of media resources can be considered to add to the BHCA that needs to be processed
by Unified CM.
25-31
Chapter 25
The maximum number of users for a Unified CM cluster is limited by the maximum size of the internal
configuration database that gets replicated between the cluster members. Currently the maximum
number of users that can be configured or synchronized is 160,000. To optimize directory
synchronization performance, Cisco recommends considering the following points:
Note
Directory lookup from phones and web pages may use the Unified CM database or the IP Phone
Service SDK. When directory lookup functionality uses the Unified CM database, only users who
were configured or synchronized from the LDAP store are shown in the directory. If a subset of users
is synchronized, then only that subset of users is seen on directory lookup.
When the IP Phone Services SDK is used for directory lookup, but authentication of Unified CM
users to LDAP is needed, the synchronization can be limited to the subset of users who would log
in to the Unified CM cluster.
If only one cluster exists, if the LDAP store contains fewer than the maximum number of users
supported by the Unified CM cluster, and if directory lookup is implemented to the Unified CM
database, then it is possible to import the entire LDAP directory.
If multiple clusters exist and if the number of users in LDAP is less than the maximum number of
users supported by the Unified CM cluster, it is possible to import all users into every cluster to
ensure directory lookup has all the entries.
If the number of user accounts in LDAP exceeds the maximum number of users supported by the
Unified CM cluster and if the entire user set should be visible to all users, it will be necessary to use
the Unified IP Phone Services SDK to off-load the directory lookup from Unified CM.
If both synchronization and authentication are enabled, user accounts that have either been
configured or synchronized into the Unified CM database will be able to log in to that cluster. The
decision about which users to synchronize will impact the decision on directory lookup support.
Cisco supports the synchronization of user accounts up to the limit mentioned above, but it does not
enforce this limit. Synchronizing more user accounts can lead to starvation of disk space, slower
database performance, and longer upgrade times.
Note
IM and Presence does not count toward the 21-server limit for a megacluster deployment.
Cisco IM and Presence has introduced a VM configuration template to align with megacluster
deployments using a 25,000-user VM configuration.
A Unified Communications deployment can be simplified in certain cases with a Unified CM
megacluster. The following limits increase with such a deployment:
Maximum number of endpoints supported is twice the number of a normal cluster (8 call processing
subscriber pairs).
25-32
Chapter 25
However, some cluster-wide constants do not increase. Chief among these are:
Note
Maximum number of LDAP synchronized or provisioned end users (160,000 users per cluster)
Due to the many potential complexities surrounding megacluster deployments, customers who wish to
pursue such a deployment must engage their Cisco Account Team, Cisco Advanced Services, or their
certified Cisco Unified Communications Partner.
Analyze the given system description as an aggregation of the identified services and arrive at a net
system cost.
Determine the number of required servers based on system cost and deployment options.
For IM and Presence, the following system variables in the system under analysis are relevant and must
be considered for accurate sizing:
chat room
State changes per user (both call related and user initiated)
Deployment model
Whether intercluster presence is supported
Whether federation is supported
Whether high availability is desired
Server preferences
The desired VM configuration size
System options
Whether compliance recording is required
Once the system requirements are quantified, the number of required virtual machines can be determined
from the data in Table 25-13.
25-33
Chapter 25
Table 25-13
VM Configuration
500 Users
1,500
1,000 Users
1,000
2,000 Users
6,000
5,000 Users
15,000
15,000 Users
45,000
25,000 Users
75,000
1.
For additional information on Cisco IM and Presence, refer to the latest version of Compatibility
Information for Cisco Unified Communications Manager, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-device-support-tables-list.html
The formal definitions of the VM configurations for Cisco IM and Presence are available at
http://docwiki.cisco.com/wiki/Virtualization_for_Unified_CM_IM_and_Presence
Impact on Unified CM
The Cisco IM and Presence Service influences the performance of Unified CM in the following ways:
In general, the impact of user synchronization (except for a one-time hit) and that of presence
information through the SIP trunk are negligible. The effect of CTI control of phones, however, must be
counted against CTI limits.
IM and Presence VM configurations differ from Unified CM VM configurations. IM and Presence
templates are user based while Unified CM templates are device based. For example, a 5k-user IM and
Presence VM configuration used with a Unified CM 10k-user VM configuration would support 5,000
users with 2 devices each. All IM and Presence nodes within the same cluster must use the same type of
VM configuration.
Emergency Services
The Cisco Emergency Responder tracks the locations of phones and the access switch ports to which
they are connected. The phones may be discovered automatically or entered manually into the
Emergency Responder. Table 25-14 shows the VM configurations that support the Emergency
Responder and their maximum capacities.
Note
These limits apply to standalone Emergency Responder deployments, and they assume that Native
Emergency Services are not being used.
25-34
Chapter 25
Table 25-14
VM Configuration
Maximum
Number of
Automatically
Tracked
Phones
Maximum
Number of
Manually
Configured
Phones
Maximum
Number of
Roaming
Phones
Maximum
Number of
Switches
Maximum
Number of
Switch Ports
Maximum
Number of
Emergency
Response
Locations
12,000 Users
12,000
2,500
1,200
500
30,000
3,000
20,000 Users
20,000
5,000
2,000
1,000
60,000
7,500
30,000 Users
30,000
10,000
3,000
2,000
120,000
10,000
40,000 Users
40,000
12,500
4,000
2,500
150,000
12,500
The formal definitions of the VM configurations for Cisco Emergency Responder and other Unified
Communication products are available at the following location:
http://docwiki.cisco.com/wiki/Virtualization_for_Cisco_Emergency_Responder
There can be only one Emergency Responder active per Unified CM cluster. Therefore, choose an VM
configuration that has sufficient resources to provide emergency coverage for all of the phones in the
cluster.
For more details on network hardware and software requirements for Emergency Responder, refer to the
Cisco Emergency Responder Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps842/prod_maintenance_guides_list.html
Cisco Expressway
Cisco Expressway deployments rely on Cisco Unified CM as the component for call control, including
remote endpoint registration. When sizing such a system, consider the function it performs as well as its
impact to Unified CM.
When sizing Cisco Expressway mobile and remote access, you typically must consider the following
parameters to determine the required number of Cisco Expressway-C and Expressway-E node pairs:
Number of endpoint registrations through each pair of Expressway-C and Expressway-E nodes
during peak usage time
Expected number of simultaneous voice-only and video calls traversing each pair of Expressway-C
and Expressway-E nodes
25-35
Chapter 25
Table 25-15 lists the Cisco Expressway mobile and remote access registrations and call capacities for
Cisco Expressway-C and Expressway-E server node pairs and clusters.
Table 25-15
Cisco Expressway-C and Expressway-E Mobile and Remote Access Node and Cluster
Capacities
Platform
Proxy Registrations
Video Calls
Audio-only Calls
1. Cisco Expressway-C and Expressway-E can be clustered across multiple Business Edition 6000 nodes for redundancy
purposes; however, there is no increased capacity when clustering with Business Edition 6000.
Note
The capacity numbers shown in Table 25-15 apply only to Cisco Expressway mobile and remote access
connections.
The following guidelines apply when clustering Cisco Expressway for mobile and remote access:
Note
All nodes of the Expressway-E and Expressway-C cluster pairs must use identical VM
configurations. For example, an Expressway-E node using the large VM configuration must not be
deployed if other nodes in the Expressway-E cluster or in the corresponding Expressway-C cluster
are using the medium size VM configuration.
Expressway peers should be deployed in equal numbers across Expressway-E and Expressway-C
clusters. For example, a three-node Expressway-E cluster should be deployed with a three-node
Expressway-C cluster.
An Expressway-E and Expressway-C cluster pair can be formed by a combination of nodes running
on an appliance or running as a virtual machine, as long as the node capacity is the same across all
nodes.
The Expressway node VM configurations or Expressway Appliances must match across and within
Expressway Series cluster pairs.
There is a dependency between Cisco Expressway clusters and Cisco Unified CM clusters. Expressway
capacity planning must also consider the capacity of the associated or dependent Unified CM cluster(s).
For more information about Cisco Expressway capacity planning considerations, including sizing limits,
capacity planning, and deployment considerations, refer to the Cisco Expressway product
documentation available at
http://www.cisco.com/c/en/us/support/unified-communications/expressway-series/tsd-products-su
pport-series-home.html
25-36
Chapter 25
Gateways
PSTN gateways handle traffic between the Unified Communications system and the PSTN. The amount
of traffic determines the resource usage (CPU and memory) and the number of PSTN DS0 circuits
required for the gateways.
PSTN traffic is generated by the endpoints registered to Unified CM, but there may be other sources such
as interactive voice response (IVR) applications and other parts of a contact center deployment.
Gateways can also perform other functions that require resources (such as CPU, memory, and DSP).
These functions include media processing such as media termination point (MTP), transcoding,
conference bridge, and RSVP Agents.
Gateways, especially those based on the Cisco Integrated Service Routers (ISRs), can provide other
functions such as serving as VXML processing engines, acting as border elements, doubling as Cisco
Unified Communications Manager Express or Survivable Remote Site Telephony (SRST), or performing
WAN edge functions. All of these activities need to be taken into account when calculating the gateway
load.
Gateway Groups
When considering the number of gateways, you also need to consider the geographical placement of
physical gateway servers. In a deployment model where PSTN access is distributed, you need to size
those gateways as a group by themselves and assign the appropriate amount of load to each such group.
A grouping might also be appropriate if certain gateways are expected to be dedicated for certain
functions and share common characteristics.
Therefore, to accurately estimate the number of gateways required, the following information is
required:
Groups of gateways that share a common group profile. The common profiles will depend on the
complexity of the deployment.
For each group, the traffic patterns, platform, blocking probability, and so forth, that make up the
profile.
The individual gateway platform that makes up the group. In deciding on a particular gateway
model, ensure that the model can support the capabilities and the capacity that is expected of it. Note
that more than one gateway might be required in a gateway group, depending on the ability of the
selected platform to meet the performance requirements.
PSTN Traffic
PSTN circuits are shared by all users of the system, and there are usually many more users than PSTN
circuits. The number of circuits required is estimated by using the traffic management principles
described in the section on call traffic (Call Traffic, page 25-22).
The amount of external traffic received and generated by your business determines the number of PSTN
circuits required. When converting from a TDM-based system, many customers will continue to use the
same number of circuits for their IP-based communications system as they had used for the previous
system. However, you may want to perform a new traffic analysis, which will detect if the system is
over-provisioned for the current levels of traffic (and, therefore, the customer is paying for circuits that
are not needed). If the system is under-provisioned, users will experience an unacceptable number of
blocked and/or lost calls, in which case increasing the number of circuits will remedy the situation.
25-37
Chapter 25
The number of PSTN circuits determines the DSP requirements for the gateways. DSP resources are
required to perform conversion between IP and TDM voice (PSTN circuits use TDM encoding).
One key input is the blocking factor, which determines the percentage of call attempts that may not be
serviced at peak traffic levels. A lower blocking factor means that more call attempts will succeed, but
the system will require more circuits than for a higher blocking factor.
PSTN gateway calls are made with ISDN PRI trunks using H.323.
The Real Time Control Protocol (RTCP) timer is set to the default value of 5 seconds.
Dedicated voice gateway configurations are used, with Ethernet (or Gigabit Ethernet) egress and no
QoS features. (Using QoS-enabled egress interfaces or non-Ethernet egress interfaces, or both, will
consume additional CPU resources.)
No supplementary call features or services are enabled such as general security (for example,
access control lists or firewalls), voice-specific security (TLS, IPSec and/or SRTP), AAA lookups,
gatekeeper-assisted call setups, VoiceXML or TCL-enabled call flows, call admission control
(RSVP), and SNMP polling/logging. Such extra call features use additional CPU resources.
25-38
Chapter 25
savings that VAD can provide varies with the call flow, the application, and the nature of speaker
interactions, it tends to use 10% to 30% fewer packets than would be sent for a call made with VAD
turned off.
VAD is most often turned off in endpoints and voice gateways deployed in Unified CM networks; VAD
is most often turned on in voice gateways in other types of network deployments.
Codec
Both G.711 and G.729A use as their default configuration a 20 ms sampling time, which results in a
50 packets-per-second (pps) VoIP call in each direction. While a G.711 IP packet (200 bytes) is larger
than a G.729A packet (60 bytes), this difference has not proven to have any significant effect on voice
gateway CPU performance. Both G.711 and G.729 packets qualify as "small" IP packets to the router,
therefore the packet rate is the salient codec parameter affecting CPU performance.
Performance Overload
Cisco IOS is designed to have some amount of CPU left over during peak processing, to handle
interrupt-level events. The performance figures in this section are measured with the processor running
at an average load of approximately 75%. If the load on a given Cisco IOS gateway continually exceeds
this threshold, the following results will occur:
The deployment will not be supported by Cisco Technical Assistance Center (TAC).
The Cisco IOS Gateway will display anomalous behavior, including Q.921 time-outs, longer
post-dial delay, and potentially interface flaps.
Cisco IOS Gateways are designed to handle a short burst of calls, but continual overloading of the
recommended call rate (calls per second) is not supported.
Note
With any gateway, you might be tempted to assign unused hardware ports to other tasks, such as on a
Cisco Communication Media Module (CMM) gateway where traffic calculations have dictated that only
a portion of the ports can be used for PSTN traffic. However, the remaining ports must remain unused,
otherwise the CPU will be driven beyond supported levels.
Performance Tuning
The CPU utilization of a Cisco IOS Voice Gateway is affected by every process that is enabled in a
chassis. Some of the lowest level processes such as IP routing and memory defragmentation will occur
even when there is no live traffic on the chassis.
Lowering the CPU utilization can help to increase the performance of a Cisco IOS Voice Gateway by
ensuring that there are enough available CPU resources to process the real-time voice packets and the
call setup instructions. Table 25-16 describes some of the techniques for decreasing CPU utilization.
25-39
Chapter 25
Table 25-16
Technique
CPU Savings
Description
Up to 20%
Up to 2%
Disable other non-essential functions such
as: Authentication, Authorization, and
Accounting (AAA); Simple Network
Management Protocol (SNMP); and logging
Additional Information
A full discussion of every gateway, its capabilities, and call processing capacities is not possible in this
chapter. For more information on Cisco Voice Gateways, refer to the following documentation:
Gateway protocols supported with Cisco Unified Communications Manager (Unified CM):
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/admin/8_0_1/ccmsys/a08gw.html
Interfaces and signaling types supported by the following Cisco Voice Gateways:
Cisco 3900 Series Integrated Services Routers
http://www.cisco.com/en/US/products/ps10536/products_relevant_interfaces_and_modules.ht
ml
Cisco 2900 Series Integrated Services Routers
http://www.cisco.com/en/US/products/ps10537/products_relevant_interfaces_and_modules.ht
ml
25-40
Chapter 25
Skinny Client Control Protocol (SCCP) feature support with FXS gateways:
http://www.cisco.com/en/US/prod/collateral/voicesw/ps6790/gatecont/ps2250/ps5516/product_dat
a_sheet09186a00801d87f6.html
Gateway capacities and minimum releases of Cisco IOS and Unified CM required for conferencing,
transcoding, media termination point (MTP), MGCP, SIP, and H.323 gateway features:
http://www.cisco.com/en/US/prod/collateral/routers/ps259/product_data_sheet0900aecd8057f2e0.
pdf
Voice Messaging
Voice messaging is an application that needs to be sized not only by itself but also for its effect on other
Unified Communications components, mainly Unified CM.
Total number of users is the key factor for sizing the voice messaging system. Other factors that affect
sizing for voice messaging are:
Number of calls during the busy hour that the application has to handle
Number of users who check their messages during the busy hour
Ports on the voice messaging system are analogous to the DS0s on a gateway and are shared
resources that need to be optimized. The same considerations of probabilistic arrival and the need
for blocking apply to both types of resources.
Table 25-17 shows the applicability of the various voice messaging solutions to the scalability
requirements of the deployment.
Table 25-17
Maximum Number of
Users Supported in a
Digital Networking
Solution
Maximum Number of
Users Supported in an
HTTPS Networking
Solution
Solutions
500
1,000
15,000
20,000
100,000
100,000
Yes
No
No
No
Yes
No
Yes
Yes
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Table 25-18 shows the maximum limits of various functions of different VM configurations running
Cisco Unity Connection.
25-41
Chapter 25
Table 25-18
VM Configuration
Maximum Number of
Ports
Maximum Voice
Recognition Sessions
Maximum Text to
Speech Sessions
Maximum Number of
Voicemail Users
100 Users
100
500 Users
16
16
16
500
1,000 Users
24
24
24
1,000
5,000 Users
100
100
100
5,000
10,000 Users
150
150
150
10,000
20,000 Users
250
250
250
20,000
The formal definitions of the VM configurations for Cisco Unity Connection are available at
http://docwiki.cisco.com/wiki/Virtualization_for_Cisco_Unity_Connection
Impact on Unified CM
The impact of a voice messaging system on Unified CM can be gauged by considering the extra
processing that Unified CM needs to do. These extra call flows add to the sizing load of Unified CM as
follows:
Calls that need to be forwarded to the voice messaging system when the user is not present or if the
user deliberately forwards the calls using Do Not Disturb (DND) or other features.
Calls from users who dial the voice messaging pilot number to access their voice messages go
through Unified CM, and these calls must be added to the calls being handled by Unified CM,
including both the number and the duration of these calls.
Collaborative Conferencing
Cisco Collaborative Conferencing systems include Cisco Unified CM as a component for call control.
When sizing such a system, the function it performs as well as its impact to Unified CM should be
considered.
When sizing such conferencing systems, you typically have to consider the following parameters to
determine the type and number of nodes:
Number of users who could use the system at any one time
Number of audio, video, and web users on the system at the peak usage time
25-42
Chapter 25
Table 25-19
20,000 to 50,000
1,500
15 to 35
50,000 to 500,000
2,000
25 to 250
500,000 to 1,000,000
3,000
165 to 335
1,000,000 to 2,000,000
3,500
285 to 570
2,000,000 to 8,000,000
4,000
500 to 2,000
When migrating from an "operator-scheduled" model to a user-scheduled model, you might need to
add another 20% to the baseline.
The default average meeting size is 4.5 callers per meeting. Use the value that is applicable to your
case if it is different than the default.
If the largest single meeting exceeds 20% of the estimated capacity, increase the estimate
accordingly.
If there are continuous meetings with dedicated ports, then you must add those additional ports
((Meetings) (Dedicated callers)) to the baseline.
The total number of ports will include all the above factors in addition to the baseline. Plan for
conferencing system capacity expansion if the total estimated port capacity exceeds 80% of the
maximum supported ports.
25-43
Chapter 25
Impact on Unified CM
The impact to Unified CM can be analyzed based on the extra call traffic that the conferencing system
generates. The most impact occurs when conference users dial into their meetings that are typically
scheduled at the top of the hour or half-hour. A large amount of call traffic within a few minutes of
conference start times increases the load on Unified CM for just those few minutes and must be designed
in appropriately. In addition, if conference users include callers from the PSTN or from other clusters,
those parameters must also be considered to gauge their impact on the gateways.
25-44
Chapter 25
Sizing Factors
The sizing tool uses the following inputs to calculate system capacity:
Video
The percent of conferences with video and high-quality video will impact the network bandwidth
required by the system. Up to 50% of the users can be using high-quality video.
Traffic mix
Different call types require different Unified CM resources. For accurate assessment of the
Unified CM impact, the tool requires estimates of the following call types:
Percent of conference calls incoming via enterprise IP phones. This call leg is handled by
Disaster recovery
For disaster recover, you can configure a cold-standby system in a second data center. If the primary
system is configured for high availability, you can optionally choose to configure high availability
for the disaster recovery system.
High availability
The system can be configured in non-redundant mode or in high-availability (HA) mode. In HA
mode, the cluster is provisioned with one or more backup servers (the specific configuration depends
on the system size).
25-45
Chapter 25
System Capacities
Cisco WebEx Meetings Server is offered in four system sizes, as listed in Table 25-20. System size is
expressed as the maximum number of concurrent users of the system. Maximum concurrent users
defines the maximum number of users who can participate in conference calls at any given time.
Table 25-20
250 Concurrent
Users
800 Concurrent
Users
2,000
Concurrent
Users
250
800
2,000
25
125
400
1,000
50
100
100
100
63
200
500
Recordings of meetings in
progress
13
40
100
125
400
1,000
20
Maximum
50 Concurrent
Users
Note that the following optional capabilities can be used without any impact on system capacity:
Low-resolution video
Recordings
Meetings for up to 5% of the ports (or 10% of meetings) can be recorded. You need to provision an
NFS-mounted hard drive of sufficient size to store the recorded meetings. One meeting will generate a
file with a size of 50 to 100 MB.
Network Bandwidth
To estimate the bandwidth required on the LAN and WAN, the sizing tool makes the following
assumptions:
The user mix will be 80% internal to the enterprise and 20% external.
Therefore, the required bandwidth (in Mbps) on the LAN is 0.8 * (Number of ports), and on the WAN
is 0.2 * (Number of ports)
25-46
Chapter 25
25-47
Chapter 25
BE6000H Maximum capacity of 1,000 users; 2,500 devices; and 100 contact center agents.
Supports nine collaboration application options in a single virtualized server platform. Maximum of
5,000 BHCA.
BE6000M Maximum capacity of 1,000 users; 1,200 devices; and 100 contact center agents.
Supports five collaboration application options in a single virtualized server platform. Maximum of
5,000 BHCA
BE6000S Maximum capacity of 150 users and 300 devices. Supports five fixed collaboration
applications in a single integrated router/gateway/virtualized blade server platform. Maximum of
750 BHCA.
To learn more about Cisco Business Edition 6000 solutions, visit http://www.cisco.com/go/be6000.
Cisco Business Edition 7000 is available in two hardware platform options:
BE7000H This high-density model typically supports five to ten collaboration applications in
deployments sized for 1,000 to 5,000 users with 3,000 to 15,000 devices and multiple sites.
BE7000M This medium-density model typically supports four to six collaboration applications
in deployments sized for 1,000 to 5,000 users with 3,000 to 15,000 devices and multiple sites.
To learn more about Cisco Business Edition 7000 solutions, visit http://www.cisco.com/go/be7000.
25-48
Chapter 25
25-49
Chapter 25
For example, assume there are two classes of users with the following characteristics:
100 phones at 8 BHCA = 800 BHCA
150 phones at 4 BHCA = 600 BHCA
Also assume 10 shared lines for each group, which would add the following BHCA values:
10 shared lines in the group at 8 BHCA = 0.5108 = 40 BHCA
10 shared lines in the group at 4 BHCA = 0.5104 = 20 BHCA
The total BHCA for all phone devices in this case is the sum of the BHCA for each phone group added
to the sum of the BHCA for the shared lines:
800 + 600 + 40 + 20 = 1,460 total BHCA
Note that the total BHCA in each example above is acceptable because it is below the system maximum
of 5,000 BHCA.
If you are using Cisco Unified Mobility for single number reach (SNR) on Business Edition 6000, keep
in mind that calls extended to remote destinations and mobility identities or off-system phone numbers
affect BHCA. In order to avoid oversubscribing the appliance, you have to account for this SNR remote
destination or off-system phone BHCA. To calculate the BHCA for these SNR features, see Capacity
Planning for Cisco Unified Mobility, page 21-60, and add that value to your total BHCA calculation.
Note
Media authentication and encryption using Secure RTP (SRTP) impacts the system resources and affects
system performance. If you plan to use media authentication or encryption, keep this fact in mind and
make the appropriate adjustments. Typically, 100 IP phones without security enabled results in the same
system resource impact as 90 IP phones with security enabled (10:9 ratio).
Another aspect of capacity planning to consider for Cisco Business Edition 6000 is call coverage.
Special groups of devices can be created to handle incoming calls for a certain service according to
different rules (top-down, circular hunt, longest idle, or broadcast). This is done through hunt or line
group configuration within Cisco Business Edition 6000. BHCA can also be affected by this factor, but
only as it pertains to the line group distribution broadcast algorithm (ring all members). For Business
Edition 6000, Cisco recommends configuring no more than three members of a hunt or line group when
a broadcast distribution algorithm is required. Depending on the load of the system, doing so could
greatly affect the BHCA of the system and possibly oversubscribe the platform's resources. The number
of hunt or line groups that have a distribution algorithm of broadcast should also be limited to no more
than three. These are best practice recommendations meant to prevent oversubscription of the system
BHCA. Exceeding these recommendations within a deployment is supported as long as the overall
BHCA capacity of the system is not exceeded.
25-50
Chapter 25
For example:
Assuming a system of 300 users at 5 BHCA each, with each user having one remote destination or
mobility identity (total of 300 remote destinations and mobility identities), the calculation for the
total remote destination and mobility identity BHCA would be:
Total remote destination and mobility identity BHCA =
0.5 (300 users) (1 remote destination or mobility identity per user) (5 BHCA per user) =
750 BHCA
Total user BHCA in this example is [(300 users) (5 BHCA per user)], which is 1,500 total user
BHCA. By adding the total remote destination BHCA of 750 to this value, we get a total system
BHCA of 2,250 (1,500 total user BHCA + 750 total remote destination and mobility identity
BHCA).
If other applications or additional BHCA variables are in use on the system in the example above, the
capacity might be limited. (See the preceding sections for further details.)
For more information on Cisco Business Edition 6000 capacity planning as well as other product
information, refer to the following product documentation for Cisco Business Edition 6000:
http://docwiki.cisco.com/wiki/Cisco_Unified_Communications_Manager_Business_Edition_6000
http://www.cisco.com/en/US/products/ps11369/tsd_products_support_series_home.html
25-51
Chapter 25
25-52
CH A P T E R
26
This chapter provides recommendations for administrators to manage migrations from traditional PBX
systems to IP telephony as well as from previous Unified Communications solution versions (8.x, 9.x,
and 10.x) to the latest Cisco Collaboration System Release (CSR) 11.x.
Note
Cisco 7800 Series Media Convergence Servers (MCS) are not supported with Cisco Collaboration
System Release 11.x.
For more information on minimum hardware and software requirements for Open Virtualization Archive
(OVA) templates, VMware, ESXi Hypervisor, and Unified Communications applications, refer to the
following documentation:
For Cisco Collaboration System Release 11.x, most Cisco Collaboration applications require
virtualization deployments and may not be installed directly on a server without a hypervisor. VMware
vSphere ESXi is currently the only supported hypervisor, and it is mandatory for all virtualized
deployments of Cisco Collaboration Systems. Cisco Collaboration System Releases do not support
VMware vSphere ESX or any other VMware server virtualization products besides ESXi.
This chapter discusses the following types of migrations:
Cisco 7800 Series MCS servers to Cisco Unified Communications Manager (Unified CM) on Cisco
Unified Computing System (UCS) servers
Pre-9.x license migration, from device-based licenses to user-based licensing utilizing Cisco Prime
License Manager
26-1
Chapter 26
In order to ensure a successful migration, Cisco recommends using the following resources to validate
that all requirements have been met prior to migration:
The validation will ensure a successful migration using supported upgrade paths. For example, some
early application software versions might require a multi-step upgrade in order to migrate successfully.
Similarly, server hardware along with software compatibility might require a combination of multi-step
hardware and software upgrades.
For details on Cisco Collaboration System products, refer to the documentation available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
For a list of all supported system hardware, refer to the Unified Computing Products documentation
available at
http://www.cisco.com/en/US/products/ps10265/products.html
Described in:
Revision Date
26-2
Chapter 26
Migration Prerequisites
Before implementing any collaboration migration steps, all administrators should ensure their IP
infrastructure is "collaboration ready," including redundancy, high availability, power consumption,
Quality of Service (QoS), in-line power, ethernet ports, and so forth. For further details, refer to the
chapter on Network Infrastructure, page 3-1.
If this is a first deployment of Cisco Unified Communications on Cisco Unified Computing System
(UCS), follow the guidance provided in the Cisco UCS Site Preparation Guide, available at
http://www.cisco.com/en/US/docs/unified_computing/ucs/hw/site-prep-guide/ucs_site_prep.html
Business needs of the users play an important role in identifying key system requirements to ensure that
the features and functionality are reserved or translated during migration to provide equivalent behavior.
A list of features and the versions of various devices and software helps in understanding what is
supported. Typically some kind of survey of the site and users should be performed to ensure that all
requirements (for example, fax/modems, environmental control systems, and so forth) are appropriately
identified and accounted for.
Phased Migration
This method typically starts with a small trial focused around Cisco Unified Communications Manager
(Unified CM). Once the customer is familiar with Unified CM, the system administrator can initiate the
steps to migrate and move groups of users at a time to the production system, with the new Unified CM
release.
Parallel Cutover
This method begins similar to the phased approach; however, once the customer is satisfied with the
progress of the trial, then a time and date are chosen for cutting-over all the users at once to the new
Cisco Collaboration System.
A parallel cutover has the following advantages over a phased migration:
If something unexpected occurs, the parallel cutover provides a back-out plan that allows you to
revert, with minimal effort, to the previous system, which is essentially still intact. For example,
with phased migration from a PBX, service can be restored to the users simply by transferring the
inbound PSTN trunks from the Cisco Collaboration System gateway(s) back to the PBX.
Parallel cutover allows for verification of the configuration of the collaboration services before the
system carries live traffic. This scenario can be run for any length of time prior to the cutover of the
collaboration services, thereby ensuring correct configuration of all user information such as
phones, gateways, the dial plan, mailboxes, and so forth.
26-3
Chapter 26
The parallel cutover allows for verification of the configuration of the Unified Communications
service before the system carries live traffic. This scenario can be run for any length of time prior to
the cutover of the Unified Communications service, thereby ensuring correct configuration of all
user information such as phones, gateways, the dial plan, mailboxes, and so forth.
Training can be carried out at a more relaxed pace by allowing subscribers to explore and use the
collaboration services at their own pace prior to the cutover.
The system administrator does not have to make special provisions for "communities of interest."
With a phased approach, you have to consider maintaining the integrity of features such as call
pickup groups, hunt groups, shared lines, and so forth. These associations can be easily accounted
for when moving the complete service in a parallel cutover.
One disadvantage of the parallel cutover is that it requires the collaboration services, including the
supporting infrastructure, to be fully funded from the beginning because the entire system must be
deployed prior to bringing it into service. With a phased migration, on the other hand, you can purchase
individual components of the system as and when they are needed, and this approach does not prevent
you from starting with a small trial system prior to moving to full deployment. Neither method is right
or wrong, and both depend upon individual customer circumstances and preferences to determine which
option is most suitable.
This approach typically entails a small Cisco Collaboration System trial that is connected to the main
corporate PBX. The choice of which signaling protocol to use is determined by the required features and
functionality as well as by the cost of implementation. Cisco Unified Communications Manager
(Unified CM) can support either regular PSTN-type PRI or QSIG PRI as well as H.323 and SIP. Of these
options, QSIG PRI typically provides the highest level of feature transparency between any two systems.
PSTN-type PRI provides for basic call connectivity as well as Automatic Number Identification (ANI).
In some instances, the protocol also supports calling name information. This level of connectivity is
available to all PBXs and therefore is considered to be the least costly option; that is, if the PBX can
connect to the public network through PRI, then it can connect to Unified CM because Unified CM can
be configured as the "network" side of the connection.
With either PSTN-type PRI or QSIG, the process for a phased migration is similar: move users from the
PBX to Unified CM in groups, one group at a time, until the migration is complete.
The Cisco San Jose campus, consisting of some 23,000 users housed in approximately 60 buildings, was
migrated to a Cisco Collaboration System in this manner and took just over one year from start to finish
at the rate of one building per weekend. All users in the selected building were identified, and their
extensions were deleted from the PBX on a Friday evening. At the same time, additions were made to
the PBX routing tables so that anyone dialing those extension numbers would then be routed over the
correct PRI trunk for delivery to Unified CM. During the weekend, new extensions were created in
Unified CM for the users, and new IP phones were delivered to their appropriate office locations, ready
for use by Monday morning. This process was repeated for each building until all users had been
migrated.
26-4
Chapter 26
For this approach, all IP phones and gateways are fully configured and deployed so that users have two
phones on their desk simultaneously, an IP phone as well as a PBX phone. This approach provides the
opportunity not only to test the system but also to familiarize users with their new IP phones.
Outbound-only trunks can also be connected to the Cisco Collaboration System, giving users the
opportunity to use their new IP phones to place external as well as internal calls.
Once the Cisco Collaboration System is fully deployed, you can select a time and date for bringing the
new system into full service by transferring the inbound PSTN trunks from the PBX to the Cisco
Collaboration System gateways. You can also leave the PBX in place until such time as you are confident
in the operation of the Cisco Collaboration System, at which point the PBX can then be decommissioned.
The Cisco San Jose campus voicemail service was originally provided by four Octel 350 systems serving
some 23,000 users. Cisco Unity servers were installed and users' mailboxes were configured. Users had
access to the their Cisco Unity mailbox by dialing the new access number, in order to allow them to
record their name and greeting(s) as well as to familiarize themselves with the new Telephony User
Interface (TUI). Approximately two weeks later, a Unified CM Bulk Administration Tool (BAT) update
was carried out on a Friday evening to change the Call-Forward Busy and No-Answer (CFB/CFNA)
numbers as well as the Messages button destination number for all users to the Unity system. Upon
returning to work on Monday morning, users were serviced by Cisco Unity. The Octel 350 systems were
left in place for one month to allow users to respond to any messages residing on those systems before
they were decommissioned.
Centralized Deployment
In the case of an enterprise that has chosen a centralized deployment model for its Cisco Collaboration
System, two options exist:
Start from the outside and work inward toward the central site (that is, smallest to largest).
Start from the central site and work outward toward the edges.
26-5
Chapter 26
The majority of customers choose the first option because it has the following advantages:
It provides the opportunity to fully deploy all the Cisco Collaboration services and then conduct a
small trial prior to rolling the services out to the remote locations.
The rollout of Cisco Collaboration services can be done one location at a time, and subsequent
locations can be migrated whenever convenient.
This option is the lowest cost to implement once the core Cisco Collaboration services are deployed
at the central site.
IT staff will gain valuable experience during migration of the smaller sites prior to migrating the
central site.
The remote sites should be migrated by the parallel cutover approach, whereas the central site can be
migrated using either the parallel or phased approach.
Ensure that the technical functionality (for example, codecs or the ability to do content sharing) is
fully supported so that the migration will not result in the loss of any features.
Phased migration is the most commonly used method for this purpose because it enables users
to become familiar with the new devices while still having their existing phones as backup for
some time.
Provide adequate network capacity to ensure a good experience for users. As the video
resolution increases, higher bandwidth is needed when compared with audio-only calls.
Migrate the dial plan and associated gateways (for example, ISDN H.320 gateways) and
For endpoints, consider any additional licenses needed if endpoint versions will be upgraded or if
some devices need different licenses.
System management tools can be a big help when there is a large number of endpoints or if the
26-6
Chapter 26
Organizations can then assess the types of devices, the feasibility, and the scope of the tasks needed so
that the migration of video devices to Unified CM is as efficient and effective as possible.
Note
If your existing Enterprise License Manager (ELM) resides on a physical server, then in order to upgrade
to 11.x you must move to a virtual environment, and in doing so your existing licenses must be rehosted
to the new virtual machine and reinstalled. In a virtual environment, just as with ELM, the new Prime
License Manager also requires a static MAC address to function properly.
Customers who have already deployed Cisco Unified Communications can use the Cisco Global
Licensing Operations (GLO) process to migrate existing licenses to Cisco Prime License Manager.
Note
You must have a valid Cisco.com login ID and password in order to access the Product License
Registration tool at the above link.
You may also get assistance from the Cisco Global Licensing Operations (GLO) team by either opening
a case using the Support Case Manager https://tools.cisco.com/ServiceRequestTool/scm/mgmt/case or
submitting the form at https://survey.opinionlab.com/survey/s?s=10422.
For all license migrations, follow the guidelines presented in this section.
The license migration process has been simplified to make the migration to Prime License Manager
much easier. Customers wanting to upgrade to Cisco Collaboration System Release 11.x may contact the
Cisco Global Licensing Operations (GLO) team directly for all migration needs. GLO processes the
request and sends the license file, a statement describing any user quantity changes, and the migration
policy to the requester's email address. Cisco adjusts the current software service contract product
records to reflect the number of Release 11.x users that have been licensed. The requester also receives
an email about any contract record updates to the user quantities. If it is a license entitlement inquiry,
then GLO sends the current entitlement information directly to the requester.
Make sure you register any unused Product Activation Keys (PAKs) for the system being migrated. If
the customer had planned for growth in the previous license model, take that into consideration for the
current migration. As an example, if there is a need to have some clusters on a pre-11.x release while
upgrading the rest to 11.x, then the license migration request to the GLO team must clearly state exactly
how many DLUs need to be migrated and how many will still be utilized in the old format when the
request is made.
26-7
Chapter 26
Ensure that you have analyzed all licensing needs in detail, and clearly state the needs in the request. If
migration is from a pre-9.x release with DLUs, once those DLUs are migrated there is no way to revert
back to the old schema because all migrated DLUs will be in the "revoked" state.
Note
The licensing process is subject to change with each new release. Always confirm the process with Cisco
GLO before submitting your license request to license@cisco.com.
You can contact GLO for license migration during the following stages:
After the upgrade process to obtain Cisco Unified Communications Manager Release 11.x migration
licenses.
If you are upgrading from a pre-9.x system, then use License Count Utility (LCU) output that was
run on the Cisco Unified Communications Manager publisher node.
MAC address of the Cisco Unified Communications Manager publisher node. If available, include
all previous publisher or license MAC addresses.
Note
If you are upgrading from a pre-9.x system, then use License Count Utility (LCU) output that was
run on the Cisco Unified Communications Manager publisher node.
MAC address of the Cisco Unified Communications Manager publisher node. If available, include
all previous publisher or license MAC addresses.
Site information for the contract update (for example: name-all name permutations, city, state,
country)
(Optional) Email addresses to send license and software support contract updates.
(Optional) If a User Connect License (UCL) customer, how the customer wants to allocate unused
DLUs.
For all pre-9.x systems moving to Collaboration System Release 11.x, customers must decide if they
want to use their unused DLUs or drop them at the time of migration. There are no refunds for dropped
DLUs; however, customers will save on future service charges. Note the differences between current
Cisco Unified Communications Software Subscription (UCSS) users on contract and estimate the
change, if any, in their UCSS and Essential Operate Services (ESW) costs at renewal.
At the time of migration, customers may choose how to use their existing licenses. The options are:
26-8
Chapter 26
After the upgrade process is complete, the information is locked in and becomes the customer
entitlement record moving forward. There are no further modifications to the license migration
information.
For more information, refer to the latest version of the following documents:
Cisco Unified CM
For migration of other applications not supported by Cisco Prime Collaboration Deployment, follow the
specific product migration guidelines provided in the documentation available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
Used by Cisco Unified CM and Unity Connection; Cisco Unified Workspace License (CUWL) only
Used by version 10.x of Cisco Unity Connection (non-CUWL) and Emergency Responder
26-9
Chapter 26
Using Cisco Prime Collaboration Deployment for Migration from Physical Servers to Virtual Machines
Migration without required Cisco Support; electronic fulfillment through Cisco Prime LM
Migration via email request to the Cisco GLO Team for migration assistance
Migration via the self-serve option for users experienced in the licensing process and functions
Install and configure Cisco Prime License Manager (Prime LM), and add product inventory.
PAKs are required for all migration license requests submitted to the Cisco Global Licensing Operations
(GLO) team by email to licensing@cisco.com. When it fulfills a request, GLO provides the license
file(s) to the requester via email, and the migration can proceed. Existing customers using Cisco Unified
Communications 9.x release with Cisco Enterprise License Manager may upgrade directly to release
11.x to get all the benefits of Cisco Prime LM. To fulfill any new licenses, please proceed to the Cisco
Product License Registration portal at
http://www.cisco.com/go/license
Install Cisco Option Package (COP) files (locales or device packs) on a cluster (8.6.1 and later
releases)
Switch versions
Reboot
26-10
Chapter 26
Deploy the Cisco Prime Collaboration Deployment virtual machine (delivered as a virtual
appliance).
Download the Open Virtualization Archive (OVA) files for the Cisco Collaboration applications, and
create target virtual machines using the OVA.
Download Cisco ISO images for the target release and upload them to Cisco Prime Collaboration
Deployment.
Install Cisco Collaboration System Release 11.x nodes on the virtual machines.
Simple Migration
With this type of migration, the IP addresses and hostnames are not changed during the migration. The
following procedure describes the recommended steps for migration task configuration within Cisco
Prime Collaboration Deployment to ensure a highly available migration.
Cisco Prime Collaboration Deployment first exports the data of all the existing nodes. Then it powers
down the existing publisher, installs the new publisher running as a virtual machine, and imports the
publishers data.
After the publisher migration is done, Cisco Prime Collaboration Deployment migrates the TFTP and
backup call processing nodes of the cluster. First the existing TFTP and backup call processing nodes
are powered down. Then Prime Collaboration Deployment installs the new TFTP and backup call
processing nodes, and imports the backup data.
Once the backup call processing nodes are migrated, the Cisco Prime Collaboration Deployment
migration task pauses and the administrator should configure all phones to re-register to the backup call
processing nodes by changing the order in the Unified Communications Manager group, or by using
device pools.
Finally, Cisco Prime Collaboration Deployment migrates the primary call processing nodes. Once this
is done, phones can be re-registered to their primary call processing server.
26-11
Chapter 26
For more details, refer to the latest version of the Migration to Cisco Unified Communications Manager
Using Prime Collaboration Deployment guide, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-installation-guides-list.html
Network Migration
Cisco Prime Collaboration Deployment can also be utilized for a migration requiring network
configuration changes for parameters such as server IP address and hostname. If the source Unified CM
cluster is running release 8.x or later, the phone Initial Trust List (ITL) file should be updated during the
migration by using the Bulk Certificate Management export, consolidate, and import functions in order
to avoid having to delete the ITL file manually on each phone.
For more information, refer to the latest version of the Cisco Prime Collaboration Deployment
Administration Guide, available at
http://www.cisco.com/c/en/us/support/unified-communications/unified-communications-managercallmanager/products-maintenance-guides-list.html
Ensure that the technical functionality (for example, codecs or the ability to do content sharing) is
fully supported so that the migration will not result in the loss of any features.
Provide adequate network capacity to ensure a good experience for users. As the video resolution
increases, higher bandwidth is needed when compared with audio-only calls.
Migrate the dial plan and associated gateways or trunks (for example, ISDN H.320 gateways) and
application servers (for example, conferencing servers and bridges).
For endpoints, consider any additional licenses needed if endpoint versions will be upgraded or if
some devices need different licenses.
System management tools can be a big help when there is a large number of endpoints or if the
endpoints need more back-end administration and support.
Customers should assess the types of devices, the feasibility, and the scope of the tasks needed so
that the migration of video devices to Unified CM is as efficient and effective as possible.
26-12
Chapter 26
26-13
Chapter 26
migration, only SIP specific default values in the phone report get migrated; other values in the template
are not migrated. Migrating a phone from SCCP to SIP does not require a manual reset because the
migration itself handles the reset of the phones.
For more information and detailed migration steps, refer to the latest version of the Cisco Unified
Communications Manager Bulk Administration Guide, available at
http://www.Cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
If the SIP request carries a user=phone tag, the SIP URI is interpreted as a numeric SIP URI and
Unified CM assumes that the user portion of the SIP URI is a directory number. If no user=phone is
present, the decision is based on the dial string interpretation setting in the calling device's (endpoint or
trunk) SIP profile. This setting either defines a set of characters that Unified CM will accept as part of
numeric SIP URIs (0-9, *, #, +, and optionally A-D) or it enforces the interpretation as a directory URI.
Note
If you do not specify a port, the default SIP port (5060) is assumed. If you have changed the default SIP
port to something else, then you must specify it in the SIP URI.
The following URI and directory number (DN) considerations apply to Unified CM and supported
endpoints:
URIs can be dialed from Cisco Unified IP Phone 9900 Series, Cisco Unified IP Phone 8961, Cisco
Jabber, Cisco DX Series, and third-party endpoints.
The primary URI can be synchronized directly from LDAP in Unified CM.
For more information, see the section on Implementing Endpoint SIP URIs, page 14-77.
26-14
Chapter 26
Live or fixed audio source feeds from a USB audio device are not supported on Unified CM.
Instead, a Cisco IOS router may be used to deliver multicast MoH feeds from fixed or live audio
sources. This requires configuration of multicast MoH on the Cisco IOS router using Survivable
Remote Site Telephony (SRST) or Enhanced SRST.
Multicast must be enabled on the network to enable the Cisco IOS router to stream audio to
endpoints and gateways.
While the Cisco UCS B-Series Blade Servers and C-Series Rack-Mount Servers do support a local
keyboard, video, and mouse (KVM) cable connection that provides a serial port, a Video Graphics Array
(VGA) monitor port, and two Universal Serial Bus (USB) ports, the Unified CM VMware virtual
application has no access to these USB and serial ports. Therefore, Unified CM no longer supports the
Cisco Messaging Interface (CMI) service for Simplified Message Desk Interface (SMDI) integrations,
fixed MoH audio source integration for live MoH audio feeds using the audio cards, or flash drives to
these servers.
26-15
Chapter 26
26-16
CH A P T E R
27
Network Management
Revised: July 30, 2015
Network management is a service consisting of a wide variety of tools, applications, and products to
assist network system administrators in provisioning, operating, monitoring and maintaining new and
existing network deployments. A network administrator faces many challenges when deploying and
configuring network devices and when operating, monitoring, and reporting on the health of the network
infrastructure and components such as routers, servers, switches and so forth. Network management
helps system administrators monitor each network device and network activity so that they can isolate
and investigate problems in a timely manner for better performance and productivity.
With the convergence of rich media and data, the need for unified management is greater than ever. The
Cisco Prime Collaboration (Prime Collaboration) offers a set of integrated tools that help to test, deploy,
and monitor Cisco Unified Communications and TelePresence systems. Prime Collaboration implements
the various management phases to strategically manage the performance and availability of Cisco
Unified Communications applications including voice, video, contact center, and rich media
applications. The network management phases typically include: plan, design, implement, and operate
(PDIO). Table 27-1 lists the PDIO phases and the major tasks involved with each phase.
Table 27-1
Implement
Operate
27-1
Chapter 27
Network Management
This chapter provides the design guidance for the following management tools and products that fit into
the implementation and operation phases of Cisco Unified Communications Management:
Cisco Prime Collaboration manages provisioning of initial deployments and ongoing operational
activation for Unified Communications and TelePresence services. Cisco Prime Collaboration
provides comprehensive monitoring with proactive and reactive diagnostics for the entire Cisco
Unified Communications system. It also provides a reliable method of monitoring and evaluating
voice quality in Cisco Unified Communications systems. For details, refer to the related product
documentation available at
http://www.cisco.com/en/US/products/ps11480/index.html
Cisco TelePresence Management Suite (TMS) offers visibility and centralized control of your
telepresence videoconferencing network, including remote systems. For details, refer to the related
product documentation available at
http://www.cisco.com/en/US/products/ps11338/index.html
For information on which software versions are supported with Cisco Unified Communications Manager
(Unified CM), refer to the Cisco Unified Communications Manager Software Compatibility Matrix,
available at
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/compat/ccmcompmatr.html
Described in:
Revision Date
Real-time service troubleshooting and diagnostics for Cisco TelePresence systems and phones
27-2
Chapter 27
Network Management
Cisco Prime Collaboration
Note
Diagnostics tests using Cisco IP Service Level Agreements (IP SLA) and Video SLA Assessment
Agent (VSAA)
Prime Collaboration Assurance Advanced also includes Prime Collaboration Analytics. If you have
purchased the Prime Collaboration Analytics license, you can access the Prime Collaboration Analytics
dashboards. Prime Collaboration Analytics helps you identify traffic trends, technology adoption trends,
and over/under-utilized resources in your network. You can also track intermittent and recurring network
issues and address service quality issues.
Prime Collaboration Provisioning application provides:
Standard services (phone, line, and voicemail, for example) to be ordered for subscribers (the owner
of the individual phone, voicemail, or other service)
Configuration templates that provide the ability to auto-configure the Cisco Unified
Communications voice infrastructure in a consistent way
Easy addition of the Provisioning application to an existing Cisco Unified Communications network
Simplified policy-driven Day 2 provisioning interface to move, add, delete, or change phone users
A Self-Care feature that enables end users to modify personal options quickly and easily
For information on the benefits and key features of Prime Collaboration and guidelines for deployment
(white papers), refer to the Cisco Prime Collaboration documentation available at
http://www.cisco.com/go/primecollaboration
27-3
Chapter 27
Network Management
Assurance
Cisco Prime Collaboration Assurance is a comprehensive video and voice service assurance and
management system with a set of monitoring, troubleshooting, and reporting capabilities that help
ensure end users receive a consistent, high-quality video and voice collaboration experience. Prime
Collaboration Assurance is available in two modes: Standard and Advanced.
Prime Collaboration Advanced provides all the features that enable integrated assurance management of
applications and the underlying transport infrastructure. This includes real-time monitoring and
troubleshooting of Cisco TelePresence solutions and the entire Unified Communications system.
Prime Collaboration Standard provides basic assurance features that help you manage Unified
Communications and TelePresence components. The features include:
Support for Unified Communications components including voicemail and IM and Presence
Fault monitoring for core Unified Communications components (Cisco Unified CM and Cisco Unity
Connection)
Pre-configured and customizable performance metrics dashboards that display term trends for core
Unified Communications components
Support for TelePresence components including Cisco TelePresence Video Communication Server
(VCS), Cisco TelePresence MCU, TelePresence Server, and Cisco TelePresence Conductor
27-4
Chapter 27
Network Management
Assurance
Prime Collaboration Standard also includes the following features to help you manage the Unified
Communications and TelePresence components:
Cisco Prime Collaboration Assurance provides a unified view of the entire Cisco Unified
Communications infrastructure and presents the current operational status of each element of the Cisco
Unified Communications network. Prime Collaboration also provides diagnostic capabilities for faster
problem isolation and resolution. In addition to monitoring Cisco gateways, routers, and switches, Prime
Collaboration continuously monitors the operational status of various Cisco Unified Communications
elements such as:
Cisco Unified Contact Center Enterprise (Unified CCE), Unified Contact Center Express
(Unified CCX), and Unified Customer Voice Portal (Unified CVP)
27-5
Chapter 27
Network Management
Assurance
Note
Note
Cisco Prime Collaboration Service Level View does not support multiple Cisco Unified
Contact Center Enterprise (Unified CCE) deployments.
Cisco Prime Collaboration supports Unified Communications and TelePresence applications running in
a virtualized environment but does not provide monitoring of VMware or hardware. Use vCenter for
managing VMware hosts. For Unified Computing System (UCS) B-series Blade servers, UCS Manager
provides unified, embedded management of all software and hardware components in the Cisco UCS. It
controls multiple chassis and manages resources for thousands of virtual machines. For UCS C-series
servers, the Cisco Integrated Management Controller provides the management service.
For more information on the supported products (particularly Cisco endpoints) and versions supported
by Prime Collaboration, refer to the Cisco Prime Collaboration data sheet available at
http://www.cisco.com/en/US/products/ps11480/index.html
One protocol that Prime Collaboration uses to monitor the Unified Communications elements is Simple
Network Management Protocol (SNMP). SNMP is an application-layer protocol using UDP as the
transport layer protocol. There are three key elements in SNMP managed network:
Managed devices Network devices that have an SNMP agent (for example, Unified CM, routers,
switches, and so forth).
Agent A network management software module that resides in a managed device. This agent
translates the local management information on the device into SNMP messages.
Manager Software running on a management station that contacts different agents in the network
to get the management information (for example, Prime Collaboration).
The SNMP implementation supports three versions: SNMP v1, SNMP v2c, and SNMP v3. SNMP v3
supports authentication, encryption, and message integrity. SNMP v3 may be used if security is desired
for management traffic. Prime Collaboration supports all three versions of SNNP. SNMP v1 and v2c
read/write community strings or SNMP v3 credentials must be configured on each device for agent and
manager to communicate properly. Prime Collaboration needs only SNMP read access to collect network
device information.
For more information on SNMP, refer to the Cisco Prime Collaboration documentation available at
http://www.cisco.com/en/US/products/ps11480/index.html
Simple Network Management protocol (SNMP) to manage all Cisco Unified Communications
servers, gateways, and switches.
Administrative XML Layer (AXL) to manage Unified CM. AXL is implemented as a Simple Object
Access Protocol (SOAP) over HTTPS web service.
27-6
Chapter 27
Network Management
Call Quality Monitoring (Service Experience)
HTTP to the IP phone to collect serial number and switch information. HTTP must be enabled on
the IP phones.
Enhanced event processing with Cisco Unified CM remote syslog integration, and leveraging the
Cisco Real-Time Monitoring Tool (RTMT) interface for pre-collected Unified CM cluster-wide data
Skinny Client Control Protocol (SCCP) and Session Initiation Protocol (SIP) to Cisco Unified
IP Phones for synthetic tests.
Internet Control Message Protocol (ICMP) or Ping Sweep for Cisco IOS routers and switches, and
for other voice as well as non-voice devices.
Figure 27-1 shows the system-level overview of how Prime Collaboration leverages multiple interfaces
with Unified CM to gather performance counters and alarms.
Figure 27-1
Cisco Prime
Collaboration
Cisco Unified
CM
M
SNMP RESPONSE
UDP
162
UDP
161
SNMP TRAP
ICMP REQUEST
ICMP RESPONSE
HTTPS Query
HTTPS Response
TCP
8443
345068
AXL Query to
Unified CM
(SOAP over
HTTPS)
SNMP Query
(Getting all the
management data
from Unified CM)
27-7
Chapter 27
Network Management
Note
A set of global call quality thresholds can be defined as one per supported codec type. Different
thresholds can be grouped together based on the Unified CM cluster being monitored.
Protocol: SFTP
Directory path on the Prime Collaboration virtual machine to which CDR and CMR files are
transferred
Prior to Cisco Collaboration System Release (CSR) 11.0, the emphasis was on using the Cisco Voice
Transmission Quality (CVTQ) algorithm as one means to monitor voice quality. CVTQ is based on the
Klirrfaktor (K-factor) method to estimate the MOS value of voice calls. With Cisco CSR 11.0 and later
releases, packet counts, concealment ratios, and concealment second counters represent primary
statistics because they can alert the network operator before network impairment has an audible impact
or is visible through MOS. Table 27-3 describes these counters as well as metrics computed from them.
Table 27-3
Counter or Metric
Description
Concealment
CS concealed seconds
27-8
Chapter 27
Network Management
Call Quality Monitoring (Service Experience)
As noted in Table 27-3, SCSR represents a measure of voice quality and is used by Prime Assurance to
grade calls. For calls less than 20 seconds in duration, the following SCSR values are used to estimate
call quality:
Grade
SCSR Value
Good
Acceptable
Poor
For calls of 20 seconds or longer in duration, the following SCSR values are used to estimate call quality:
Grade
SCSR Value
Good
Acceptable
Poor
27-9
Chapter 27
Network Management
The Cisco NAM provides voice quality statistics every 60 seconds. Unified CM provides voice
quality statistics after the call is completed (ended).
Unified CM monitors only the call segment within its own cluster.
Unified CM voice quality monitoring is best used to gauge the overall voice call quality in the
network.
Even if Unified CM call quality metrics are not used, Prime Collaboration uses Unified CM CDR
information to correlate with the NAM report for the following information:
Device types
Interface through which the call flowed in the case of a call to or from a gateway
Exact Unified CM server (not just the Unified CM cluster) to which the phone is connected
Trunk Utilization
Cisco Prime Collaboration provides real-time Unified CM trunk utilization performance graphs. It is
also tightly integrated with Cisco Prime Analytics in order to provide the call information it collects to
Analytics for long-term trending and reporting purposes. The call information is provided from the CDR
and CMR records Prime Collaboration gathers from Unified CM.
Prime Collaboration automatically selects and gathers voice quality information (via CDR and CMR
files) for all Cisco Unified IP Phones configured in a given Unified CM cluster. There is no
configuration option to monitor only certain IP phones in the cluster.
27-10
Chapter 27
Network Management
Call Quality Monitoring (Service Experience)
Note
When Cisco Prime Collaboration is operating at full capacity, its projected database growth (for Syslog,
CDR, and CMR files) is estimated to be about 2.4 GB per day.
Note
Protocol
Port
Service
UDP
161
SNMP Polling
UDP
162
SNMP Traps
TCP
80
HTTP
TCP
443
HTTPS
TCP
1741
UDP
22
SFTP
TCP
43459
Database
UDP
514
Syslog
TCP
8080
TCP
8443
The Cisco NAM is accessed remotely over HTTPS with a non-default port. Prime Collaboration will
authenticate with each Cisco NAM and maintain the HTTP/S session.
All the management traffic (SNMP) originating from Prime Collaboration or managed devices is marked
with a default marking of DSCP 0x00 (PHB 0). The goal of network management systems is to respond
to any problem or misbehavior in the network. To ensure proper and reliable monitoring, network
management data must be prioritized. Implementing QoS mechanisms ensures low packet delay, low
loss, and low jitter. Cisco recommends marking the network management traffic with an IP Precedence
of 2, or DSCP 0x16 (PHB CS2), and providing a minimal bandwidth guarantee. The DSCP value must
be configured in the Windows Operating System.
If managed devices are behind a firewall, the firewall must be configured to allow management traffic.
Prime Collaboration has limited support in a network that uses Network Address Translation (NAT). It
must have IP and SNMP connectivity from the Prime Collaboration server to the NAT IP addresses for
the devices behind the NAT. Prime Collaboration contains static NAT support.
27-11
Chapter 27
Network Management
Analytics
Bandwidth Requirements
Prime Collaboration polls the managed devices for operational status information at every configured
interval, and it has the potential to contain a lot of important management data. Bandwidth must be
provisioned for management data, especially if you have many managed devices over a low-speed WAN.
The amount of traffic varies for different types of managed devices. For example, more management
messages may be seen when monitoring Unified CM as compared to monitoring a Cisco Voice Gateway.
Also, the amount of management traffic will vary if the managed devices are in a monitored or partially
monitored state and if any synthetic tests are performed. Prime Collaboration has a Bandwidth Estimator
that is available at
http://www.cisco.com/web/applicat/ombwcalc/OMBWCalc.html
Analytics
Cisco Prime Collaboration Analytics provides many additional benefits to Prime Assurance. It provides
trending to identify degradation over time. It can also utilize trending to provide capacity planning and
quality of service (QoS) information. The capacity planning feature allows administrators to plan for
growth and also to identify over- or under-utilized resources (for example, TelePresence endpoints) in
their network. Analytics can generate automated reports that provide actionable information to CIOs and
IT planners. Reports can be customized to meet unique business needs.
Analytics supports the following predefined dashboards:
Technology Adoption
Asset Usage
Traffic Analysis
Capacity Analysis
Service Experience
27-12
Chapter 27
Network Management
Provisioning
Note
Provisioning
Prime Collaboration Provisioning is available in the following forms:
Prime Collaboration Provisioning Standard Available with Cisco Collaboration Systems 10.x
releases (Cisco Unified CM 10.x and Cisco Unity Connection 10.x).
27-13
Chapter 27
Network Management
Provisioning
Table 27-5
Features
Standard
Advanced
Batch provisioning
Allows you to deploy a large number of Provides advanced batch options such as
services by combining them into a single importing users and services, and adding or
modifying users and services across multiple
batch.
clusters. You can also batch-import
Note: Batch Provisioning is available for
infrastructure settings across multiple
a single cluster only.
clusters.
Infrastructure templates
API
Support for North Bound Interface (NBI) Support for North Bound Interface (NBI) is
is not available.
available.
Cisco Prime Collaboration provides a simplified web-based provisioning interface for both new and
existing deployments of Cisco Unified Communications Manager (Unified CM), Cisco Unified
Communications Manager Express (Unified CME), Cisco Unity, Cisco Unity Connection, and Cisco
Unity Express. Prime Collaboration provides provisioning for both the infrastructure and subscribers for
Day 1 and Day 2 needs. Day 1 needs include configuring new deployments and adding more sites or
locations; Day 2 needs include services for ongoing moves, adds, and changes on various components
of the Cisco Unified Communications solution.
Cisco Prime Collaboration also exposes northbound APIs to allow Cisco and third parties to integrate
with external applications such as HR systems, custom or branded user portals, other provisioning
systems, and directory servers.
For details on Prime Collaboration system requirements and installation steps, provisioning users and
the infrastructure of supported components, and capacity information, refer to the Cisco Prime
Collaboration documentation available at
http://www.cisco.com/en/US/products/ps11480/index.html
To provide a better understanding of how Prime Collaboration can be used as a network management
solution for provisioning various Cisco Unified Communications components, the next section presents
some of the basic concepts of Prime Collaboration.
27-14
Chapter 27
Network Management
Provisioning
Provisioning Concepts
Cisco Prime Collaboration serves as a provisioning interface for the following components of a Cisco
Unified Communications system:
Call processors
Cisco Unified Communication Manager (Unified CM)
Cisco Unified Communications Manager Express (Unified CME)
Message processors
Cisco Unity
Cisco Unity Connection
Cisco Unity Express
Presence processors
Cisco IM and Presence
Cisco Voice Gateways
Cisco VG224, VG204, and VG202 Analog Voice Gateways
Note
For more information on component version compatibility, refer to the Prime Collaboration information
at http://www.cisco.com/en/US/products/ps11480/index.html.
The following sections describe some of the Prime Collaboration concepts involved in configuring those
components.
Domain
Domains are used for administrative purposes to create multiple logical groups within a system.
Domains have the following characteristics:
One domain can contain multiple call processors and multiple optional message processors.
Service Area
Service areas represent offices. Service areas determine the dial plans and other voice-related
configuration settings in the domain. In reality, each office may have multiple service areas. The service
area determines attributes such as device group, route partition, and calling search space used within
Unified CM. Service areas have the following characteristics:
Each service area is assigned to a single call processor and one optional message processor.
27-15
Chapter 27
Network Management
Provisioning
When deploying a new site or making moves, adds and changes to an existing site, users make all
changes to the underlying systems through a two-stage process of creating an order and then processing
that order. You can set policies for both of these stages. For example, you can configure the system so
that one group of users can only create and submit orders, while another group of users can view and
perform processing-related activities. Prime Collaboration contains an automation engine that performs
the order processing, including service activation and business flow, based on how Prime Collaboration
is configured.
The workflow coordinates activities of the ordering process (approval, phone assignment, shipping, and
receiving).
Configuration Templates
Prime Collaboration enables you to configure Unified CM, Unified CME, Cisco Unity, Cisco Unity
Express, and Cisco Unity Connection in a consistent way through the use of configuration templates.
You can use these templates to configure any of these products, to perform an incremental rollout on
these existing products, and to deploy a new service across existing customers.
Batch Provisioning
Creating users and provisioning their services can also be done automatically through batch provisioning
for rolling out a new office or transitioning from legacy systems.
Best Practices
The following best practices and guidelines apply when using Prime Collaboration to provision Cisco
Unified Communications components for any new and/or existing deployments:
Managed devices must be up and running before using Prime Collaboration for further day-one
activities such as rolling out a new site and day-two activities such as moves, adds, and changes.
Pre-configuration is required for Cisco Unified CM, Cisco Unity, Unified CME, Survivable Remote
Site Telephony (SRST), Cisco Unity Express, and Cisco IM and Presence Service.
Consider the use of Subscriber Types, Advanced Rule settings, and other configuration parameters.
Add call processors such as Unified CM, and/or Unified CME and message processors such as Cisco
Unity, Unity Connection, and/or Unity Express.
Create domains and assign call processors and message processors to the created domains.
Provision the voice network by creating and using templates to configure Unified CMs or
Unified CMEs, or import current voice infrastructure configurations from an existing deployment.
Set up the deployment by creating service areas for each domain (typically one per dial plan) and
assigning subscriber (user) types to each service area.
27-16
Chapter 27
Network Management
Provisioning
processors.
Create a domain for each site, consisting of one call processor and zero or more optional
message processors.
Create multiple domains if different administrators are required to manage a subset of the
subscribers.
Add only the Unified CM publisher as the call processor for Prime Collaboration. Any changes
made to the Unified CM publisher through Prime Collaboration will be synchronized to all the
Unified CM subscriber servers.
Use configuration templates for Unified CM, Unified CME, or Cisco Unity Express.
Use Cisco IOS commands for Unified CME and Cisco Unity Express configuration templates.
Add Cisco Unified CM infrastructure data objects for Unified CM configuration templates.
Change and modify the existing configuration templates for batch provisioning for large quantities
of phones and lines (DNs).
Create multiple domains if you want different domain administrators to manage different sets of
subscribers for Day 2 moves, adds, and changes of services (such as phones, lines, and voicemail),
even for a single-site deployment.
Create multiple service areas if multiple dial plans are required for the device pools, location, calling
search space, and phones.
Collaboration user configuration interface allows users to enter only IPv4 IP addresses because
Unified CM has SOAP AXL interfaces in IPv4 only. Therefore, Prime Collaboration must use
IPv4 addresses to communicate with the AXL interfaces on Unified CM.
Prime Collaboration handles the IPv6 addresses contained in SIP trunk AXL response
messages.
Support of IPv6-aware functions does not affect support for current Cisco Unified
Communications Manager Express, Cisco Unity, Cisco Unity Express, and Cisco Unity
Connection devices.
27-17
Chapter 27
Network Management
Protocol
Port
Service
TCP
80
HTTP1 2
TCP
8443
HTTPS2
TCP
22
SSH3
SSH
23
Telnet3
TCP
1433
Database4
27-18
Chapter 27
Network Management
Cisco TelePresence Management Suite (TMS)
Calendaring Options
Calendaring integration gives users the ability to schedule video conferences and invite participants
directly from their calendaring application while viewing availability information of resources
regardless of where meetings are created. Calendaring options include:
Cisco TelePresence Management Suite Extension for IBM Lotus Notes (TMSXN)
Allows conference organizers to schedule conferences using their IBM Lotus Notes client. For more
information on Cisco TMSXN, refer to the product documentation available at
http://www.cisco.com/en/US/products/ps11472/tsd_products_support_series_home.html
Cisco highly recommends integrating your corporate calendaring application with the scheduling and
management platform chosen by your organization. However, you may also choose to schedule
conferences using the TMS web interfaces.
When deploying Cisco TMS calendar integration as your corporate calendaring application, choose the
appropriate extension for your environment. For example, if Microsoft Exchange is the existing
calendaring application, use TMSXE. TMSXE is installed on a standalone server, and TMSXN is
installed on the Lotus Domino server. The integration software is installed separately from Cisco TMS
and communicates with your calendaring server using HTTP or HTTPS.
Cisco recommends having your video conferencing resources (Cisco TelePresence Video
Communication Server or Cisco MCU) dedicated for either scheduled or permanent/instant conferences.
This is because permanent or instant conferences could consume scheduled resources, which would
result in undesirable consequences on the scheduled conferences, such as scheduled video participants
being unable to join or joining as audio-only due to lack of resources on the server.
Reporting
Cisco TMS provides various types of reporting and analysis functionality, including:
Asset management reports: ticket logs, device events, device alarms, and connectivity
Scheduling activity reports, including user-based, scheduling interface used, conference event logs,
and conference reports
27-19
Chapter 27
Network Management
However, some of these functions work only in certain deployments. For example, when an endpoint
such as the Cisco TelePresence TX9000 or Cisco TelePresence System EX90 is registered to Cisco
Unified Communications Manager (Unified CM), Cisco TMS cannot generate reports for that endpoint.
Cisco TMS can generate only call history and call detail record (CDR) reports for an endpoint registered
to the Cisco TelePresence Video Communication Server (VCS). For those endpoints that are registered
to Unified CM, CDRs can be downloaded from Unified CM.
Organizations that require more customized reports, business knowledge, and integration with Business
Intelligence Applications can use the Cisco TelePresence Management Suite Analytics Extension
(TMSAE), which is an online analytical processing system for Cisco TMS that provides advanced
reporting functionality for your video network. For more information on Cisco TMSAE, refer to the
product documentation available at
http://www.cisco.com/en/US/products/ps11472/tsd_products_support_series_home.html
Management
The main functions of management in the TelePresence environment include: provisioning, monitoring,
maintenance, and resource management. Cisco TelePresence Management Suite (Cisco TMS) enables
management of the TelePresence environment, along with a scheduling interface that it supports in a
TelePresence environment.
27-20
Chapter 27
Network Management
Cisco TelePresence Management Suite (TMS)
Provisioning
The Cisco TMS Provisioning Extension (TMSPE) is a provisioning application for Cisco TMS and
Cisco VCS. Cisco TMSPE enables video conferencing network administrators to create and manage
large deployable video conferencing solutions. Cisco TMSPE is an add-on replacement for the TMS
agent on the Cisco TMS server, and it provides the following main features:
Ability to import users from Microsoft and generic LDAP sources (LDAP, LDAPS, AD)
User personalization and administrative device configuration control for devices supported by
Cisco TMS Provisioning Extension (for example, Jabber video, Cisco IP Video Phone E20, and
Cisco TelePresence System EX Series and MX Series)
End-user FindMe portal on Cisco TMS using Microsoft Active Directory (AD) login instead of
Cisco VCS Web user interface
For further information, refer to the latest version of the Cisco TelePresence Management Suite
Provisioning Extension Deployment Guide, available at
http://www.cisco.com/en/US/products/ps11338/products_installation_and_configuration_guides_l
ist.html
Phone books
Phone books help users maintain their contacts and dial them. Cisco TMS phone books can be created
and populated from different sources such as Microsoft Active Directory (AD), Cisco Unified CM, an
H.350 server, and gatekeepers.
There are two types of phone books: local phone books and global phone books. Local phone books (also
called favorites) are a file stored on an endpoint specific to the end user. Contacts can be added,
modified, and deleted as desired by the user.
Global or corporate phone books are pushed from Cisco TMS to the endpoints. They cannot be modified
from the endpoint because they are automatically populated from AD, an H.350 server, or the local
Cisco TMS database. Administrators can select the phone books for specific users and push them to the
appropriate endpoints.
27-21
Chapter 27
Network Management
Cisco Unified Communications Manager (Unified CM) including Cisco IM and Presence, which
is licensed through Unified CM, and Cisco Unified Communications Manager Session Management
Edition (Unified CM SME)
When these applications are deployed as part of Cisco Business Edition 6000 or 7000, they also use
Cisco Prime LM.
Following license purchase, the licenses are registered (through electronic fulfillment on Cisco
Prime LM or manually at the Product License Registration portal at www.cisco.com/go/license), and
then installed on Cisco Prime License Manager. Cisco Prime LM is connected to the application
instances under license management, and it polls the applications. When polled, a subscribing
application sends its license requirements to Prime LM, and Prime LM compares the application's
requirements to the available licenses. If the application's requirements, totaled for all application
instances, are within the available license count, then Prime LM returns a status of in compliance.
Similarly, if license requirements exceed available licenses, then Prime LM returns a status of not in
compliance.
An application is allowed 60 days of non-compliance, during which administrators can make changes if
there are insufficient licenses or if the Prime LM node has lost communication with the application node.
After 60 days of non-compliance, the Unified Communications Manager application(s) will no longer
allow administrative changes; however, the application(s) will continue to function (call control) with
no loss of service. After 60 days of non-compliance, the Unity Connection application(s) will allow
administrative changes, but the application(s) will not continue to function (users will not have access
to voice messaging).
For more information on Cisco Unified Communications licensing, refer to the information at
http://www.cisco.com/go/uclicensing
Deployment Scenarios
Prime LM is installed automatically on the same virtual machine as the Unified CM (including SME)
and Unity Connection applications when they are installed. You may choose to use Prime LM on one of
these virtual machines in a co-resident configuration, as the active managing Prime LM, or you may opt
to run Prime LM in a standalone configuration where Prime LM is installed on a dedicated virtual
machine.
In the co-resident configuration, Prime LM consumes only a very small amount of resources and hence
is considered to have no impact to the virtual machine sizing. For example, no additional vCPUs would
need to be added to the virtual machine configuration because of the Prime LM service. In the
co-resident configuration, Cisco Prime LM is supported for use on any of the applications OVA virtual
machine configurations.
27-22
Chapter 27
Network Management
Cisco Prime License Manager
In a standalone configuration the Prime LM resides on a separate virtual machine created and managed
specifically for, or dedicated to, the Prime LM application. In the standalone configuration, the
Prime LM is installed as a separate virtual machine using the Prime LM OVA.
The main considerations for choosing between co-resident and standalone deployments center around
administration and management. The main benefits of deploying Prime LM in a standalone
configuration are as follows:
Upgrading a standalone Cisco Prime LM is done independently from upgrading the applications
(Unified CM or Unity Connection). Whereas, upgrading a co-resident configuration of Cisco
Prime LM is done by upgrading the co-resident application, which upgrades the application and
Cisco Prime LM at the same time
Administrative changes required on a standalone Prime LM will not impact the application servers.
For example, on a co-resident configuration, having to upgrade or reboot the Prime LM would
require an upgrade or reboot of the application.
The trade-off, however, with a standalone configuration is that it requires a separate virtual machine to
be created and managed.
Deployment Recommendations:
If you are installing only a single application on a single node or cluster, run Prime LM co-resident.
If you are installing a very small number of application instances, you may:
Run Prime LM on a separate virtual machine. This approach provides more administration and
management flexibility but requires a separate virtual machine for Prime LM.
Run a single Prime LM co-resident with one application virtual machine if you want license
pooling and/or centralized management but you are unwilling to dedicate a virtual machine for
running Prime LM.
Run a different Prime LM on each application instance if you do not need license pooling and
With Cisco Business Edition 6000, Prime LM would typically be co-resident with one of the
application servers. However, if desired, it is possible to run Prime LM as a standalone virtual
machine, but it would need to be counted against the maximum number of applications allowed on
a Cisco Business Edition 6000 server.
If you have a medium to large deployment, run Prime LM on a separate virtual machine. The
incremental impact on the number of required virtual machines is minimal in this case, and the
trade-off between operating expenses and capital expenditures is favorable.
Enterprise or global
As the description implies, one Prime LM instance can support an entire enterprise or global
deployment. This model provides the most simplicity by utilizing one common centralized license
pool for all the Unified Communications applications connected to the Prime LM.
27-23
Chapter 27
Network Management
Redundancy
Prime LM is deployed as a non-redundant application. In the event that the Prime LM application
becomes unavailable (for example, if the Prime LM virtual machine is experiencing operating system
issues and it cannot boot up), the customer has 60 days to restore the Prime LM application before
license enforcement occurs. The applications will run for a period of 60 days without communication
with the Prime LM.
For restoration of the Prime LM application, another installed Prime LM application (such as a
co-resident instance) can be created by re-adding the product inventory to the new Prime LM
application. Since the MAC address of the virtual machine running the new Prime LM application would
be different, transferring the registration of the license file to this new Prime LM would be required.
Alternatively, Prime LM can be restored from a Disaster Recovery System (DRS) backup. In this case,
configure the same MAC address on the new and original Prime LM virtual machines, otherwise the
license registration will have to be transferred to the new virtual machine. If license additions or changes
have been made since the DRS backup, a new license file will have to be requested. A Prime LM
co-resident backup can be restored only to a co-resident Prime LM application, and a standalone backup
can be restored only to a standalone Prime LM.
27-24
Chapter 27
Network Management
Additional Tools
Additional Tools
In addition to the network management tools mentioned above, the following tools also provide
troubleshooting and reporting capabilities for Cisco Unified Communications systems:
Cisco IOS Voice Gateways (3700 Series, 2800 Series, 3800 Series, 5350XM, and 5400XM)
Unified Analysis Manager provides the following key features and capabilities:
Supports setting and resetting of trace level across Unified Communications elements.
Supports collection and export to a define FTP server of log and trace files from Unified
Communications elements.
Supports analysis of the call path (call trace capability) across Unified Communications elements.
For more details on the report options, refer to the information about the Cisco Unified Analysis
Manager in the Cisco Unified Real-Time Monitoring Tool Administration Guide, available at
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/service/8_5_1/rtmt/RTMT.html
27-25
Chapter 27
Network Management
Additional Tools
Unified CM Device Counts Summary Provides the number of devices by model and protocol that
exist in the Cisco Unified Communications Manager database.
Unified CM Database Status - Provides a snapshot of the health of the Unified CM database. This
report should be generated before an upgrade to ensure the database is healthy.
For more information on the report options, refer to the latest version of the Cisco Unified Reporting
Administration Guide, available at
http://www.cisco.com/en/US/products/sw/voicesw/ps556/prod_maintenance_guides_list.html
27-26
Chapter 27
Network Management
Integration with Cisco Unified Communications Deployment Models
Campus
In the campus model, Cisco Unified Network Management applications, along with call processing
agents, are deployed at a single site (or campus) with no telephony services provided over an IP WAN.
An enterprise would typically deploy the single-site model over a LAN or metropolitan area network
(MAN). Figure 27-2 illustrates the deployment of Cisco Unified Network Management applications in
the single-site model.
Figure 27-2
Campus Deployment
Data Center
MCU's
Applications
M
M
H.320
gateway(s)
M
M
Voice gateway
(Cisco Unified
Border Element)
V
PSTN
IP WAN
Voice gateway
(Cisco Unified
Border Element)
348674
IP
27-27
Chapter 27
Network Management
The following design characteristics and recommendations apply to the single-site model for deploying
Prime Collaboration:
Cisco recommends deploying Unified CM voice quality monitoring to monitor overall voice quality
in the network.
Cisco recommends deploying the Cisco NAM to monitor key IP phone devices, gateway devices,
and application servers in the network and to investigate and troubleshoot voice quality issues.
Prime
Collaboration
Cisco Unified
CM cluster
Data Center
H.320
gateway
M
M
IP
M
M
H.320
gateway
Branch A
PSTN
Applications
IP
MCU's
IP WAN
IP
Branch B
27-28
348675
Headquarters
Chapter 27
Network Management
Integration with Cisco Unified Communications Deployment Models
The following design characteristics and recommendations apply to the multisite model for deploying
Prime Collaboration with centralized call processing:
Multiple Prime Collaborations can be deployed, with each instance managing multi-site and
multi-cluster Unified Communications environments. In this deployment scenario, Cisco
recommends that you deploy a Manager of Managers (MoM). Each Prime Collaboration can provide
real-time notifications to the higher-level MoM using SNMP traps, syslog notifications, and email
to report the status of the network being monitored.
Cisco recommends deploying Unified CM voice quality monitoring to monitor overall voice quality
in the network.
Cisco recommends using the Service Level Agreement (SLA) feature and Synthetic test feature to
check for network infrastructure status.
Cisco recommends deploying the Cisco NAM to monitor key IP phone devices, gateway devices,
and application servers in the network and to investigate and troubleshoot voice quality issues.
27-29
Chapter 27
Network Management
Prime
Collaboration
M
M
M
Media
Resources
Applications
IP
H.320
gateway
Branch A
H.320
gateway
ISDN
network
Applications
Media
Resources
H.320
gateway
V
V
IP WAN
M
M
IP
Prime
Collaboration
PSTN
Applications
Headquarters
IP
Branch B
348676
Media
Resources
27-30
Chapter 27
Network Management
Integration with Cisco Unified Communications Deployment Models
A multisite WAN deployment with distributed call processing has many of the same requirements as a
single site or a multisite WAN deployment with centralized call processing in terms of deploying Prime
Collaboration. Follow the best practices and recommendations from these other models in addition to
the ones listed here for the distributed call processing model:
If only one Cisco Unified Network Management system is deployed to manage multiple Unified CM
clusters, Cisco recommends deploying Prime Collaboration along with the Unified CM cluster that
has the highest call volume and the most endpoints.
Multiple Prime Collaborations can be deployed, with each instance managing multi-site and
multi-cluster Unified Communications environments. In this deployment scenario, Cisco
recommends that you deploy a Manager of Managers (MoM). Each Prime Collaboration can provide
real-time notifications to the higher-level MoM using SNMP traps, syslog notifications, and email
to report the status of the network being monitored.
Cisco recommends deploying Unified CM voice quality monitoring to monitor overall voice quality
in the network.
Cisco recommends deploying the Cisco NAM to monitor key IP phone devices, gateway devices,
and application servers in the network and to investigate and troubleshoot voice quality issues.
Site 1
Unified CM Cluster
Site 2
Publisher
M
M
M
Cisco Unified
OM
M
M
Prime
Collaboration
IP
Note
IP
Manager of
Managers
IP
IP
348677
Manager of
Managers
There is no native high-availability or redundancy support for Prime Collaboration with this model.
27-31
Chapter 27
Network Management
The following design characteristics and recommendations apply when deploying Prime Collaboration
with clustering over the WAN:
Cisco recommends deploying Prime Collaboration in the headquarter site where Unified CM
publisher is located.
Multiple Prime Collaborations can be deployed, with each instance managing multi-site and
multi-cluster Unified Communications environments. In this deployment scenario, Cisco
recommends that you deploy a Manager of Managers (MoM). Each Prime Collaboration can provide
real-time notifications to the higher-level MoM using SNMP traps, syslog notifications, and email
to report the status of the network being monitored.
Cisco recommends deploying Unified CM voice quality monitoring to monitor overall voice quality
in the network.
Cisco recommends deploying the Cisco NAM to monitor key IP phone devices, gateway devices,
and application servers in the network and to investigate and troubleshoot voice quality issues.
27-32
GLOSSARY
A
AA
Automated attendant
AAD
AAR
AC
ACD
ACE
ACF
Admission Confirm
ACL
ACS
AD
ADAM
ADFS
ADPCM
ADUC
AES
AFT
AGM
ALG
ALI
AMI
AMIS
AMWI
GL-1
Glossary
ANI
AP
Access point
APDU
API
ARJ
Admission Reject
ARP
ARQ
Admission Request
ASA
ASP
ASR
ATA
ATM
AVC
AXL
B
BAT
BBWC
BES
BFCP
BFD
BGP
BHCA
BHCC
BIB
Built-in bridge
BLF
BOSH
BPDU
GL-2
Glossary
bps
BRI
BTN
Bill-to number
C
CA
Certificate Authority
CAC
CAM
Content-addressable memory
CAMA
CAPF
CAPWAP
CAR
CAS
CBWFQ
CCA
CCD
CCS
CDP
CDR
CER
CGI
CIF
CIR
CKM
CLEC
CLID
CM
CMC
GL-3
Glossary
CME
CMI
CMM
CNG
CO
Central office
Co-located
Two or more devices in the same physical location, with no WAN or MAN connection between them
COM
COP
COR
Class of restriction
Co-resident
Two or more services or applications running on the same server or virtual machine
CoS
Class of service
CPCA
CPI
CPN
CRM
CRS
cRTP
CSF
CSTA
CSUF
Cross-Stack UplinkFast
CSV
Comma-separated values
CTI
CTL
CUBE
Cisco Unified Border Element, formerly the Cisco Multiservice IP-to-IP Gateway (IP-IP Gateway)
CUE
CUMI
CUPI
CUSP
GL-4
Glossary
CVTQ
CWA
D
DC
Domain controller
DDNS
DDR
DFS
DHCP
DID
DIT
DMVPN
DMZ
Demilitarized zone
DN
Directory number
DNIS
DNS
DoS
Denial of service
DPA
DRS
DSCP
DSE
DSP
DTIM
DTLS
DTMF
DTPC
DUC
GL-5
Glossary
E
E&M
EAP
EAPOL
EC
Echo cancellation
ECM
ECS
EI
Enhanced Image
EIGRP
ELCAC
ELIN
EM
Extension Mobility
EMCC
ER
ERL
ESF
E-SRST
F
FAC
FCC
FCoE
FECC
FIFO
First-in, first-out
FQDN
FR
Frame Relay
FTP
FWSM
GL-6
Glossary
FXO
FXS
G
GARP
GC
Global catalog
GDPR
GKTMP
GLBP
GMS
GPO
GPRS
GSB
GSM
GSS
GUI
GUP
H
H.225D
H.225 daemon
HDLC
HMS
HP
Hewlett-Packard
HSRP
HTTP
HTTPS
HTTP Secure
HVD
Hz
Hertz
GL-7
Glossary
I
Internet Assigned Numbers Authority
IANA
IAPP
ICA
ICCS
ICE
ICMP
ICS
ICT
Intercluster trunk
IdP
Identity Provider
IE
Information Element
IETF
IGMP
IIS
ILS
IM
Instant messaging
IMAP
IMS
IP Multimedia Subsystem
IntServ
Integrated Services
IntServ/DiffServ
IOPS
IP
Internet Protocol
IPCC
IPMA
IPPM
IPSec
IP Security
IP SLA VO
IPVMS
GL-8
Glossary
ISO
ISR
ITEM
ITU
IVR
J
JCF
JID
Jabber Identifier
JTAPI
K
kbps
KPML
L
LAN
LBM
LBR
Low bit-rate
LCD
LCF
Location Confirm
LCS
LDAP
LDAPS
LDIF
LDN
LEAP
LEC
GL-9
Glossary
LFI
LHS
Left-hand side
LLDP
LLDP-MED
LLQ
Low-latency queuing
LRG
LRJ
Location Reject
LRQ
Location Request
LSC
LUN
LWAP
LWAPP
M
MAC
MAN
Mbps
MCM
MCS
MCU
MDN
MDS
MFT
Multiflex trunk
MGCP
MIB
MIC
MIME
MIPS
GL-10
Glossary
MISTP
MITM
Man-in-the-middle
MLA
MLP
MLPP
MLPPP
MLTS
MMoIP
MMP
MOC
MoH
Music on hold
MOS
MPLS
MRG
MRGL
ms
Millisecond
MSI
MSP
MSP
MTP
mW
Milli-Watt
MWI
MXE
N
NAS
NAT
NDR
Non-delivery receipt
GL-11
Glossary
NENA
NFAS
NIC
NOC
NPA
NSE
NSF
NTE
NTLP
NTP
O
OBTP
OCS
ORA
OSPF
OU
Organizational unit
OVA
OWA
P
PAC
PAK
PBX
PC
Personal computer
PCAP
PCI
PCM
GL-12
Glossary
PCoIP
PC over IP
PCTR
PD
Powered device
PHB
Per-hop behavior
PIN
PINX
PIX
PKI
PLAR
PoE
POTS
PPP
Point-to-Point Protocol
pps
PQ
Priority Queue
PRACK
PRI
PSAP
PSE
PSK
Pre-Shared Key
PSTN
PVC
Q
QBE
QBSS
QoS
Quality of Service
QSIG
Q signaling
GL-13
Glossary
R
RADIUS
RAS
RBAC
RCC
RCP
RDNIS
REST
RF
Radio frequency
RFC
RHS
Right-hand side
RIM
Research In Motion
RIP
RIS
RMA
RMTP
RoST
RSNA
RSP
Route/Switch Processor
RSSI
RSTP
RSVP
RTCP
RTMP
RTMT
RTP
GL-14
Glossary
RTSP
RTT
Round-trip time
S
S1, S2, S3, and S4
SaaS
Software-as-a-Service
SAF
SAML
SAN
SBC
SCCP
SCSI
SDI
SDK
SDL
SDP
SE
SF
Super Frame
SFTP
SI
Standard Image
SIMPLE
SIP
SIS
SIW
Service Inter-Working
SLA
SLA VO
SLB
SLDAP
Secure LDAP
GL-15
Glossary
SMA
SMDI
SME
SMS
SMTP
SNMP
SOAP
SPA
SPAN
SQL
SRE
SRND
SRST
SRSV
SRTP
SRV
Server
SS7
Signaling System 7
SSID
SSL
SSO
Single Sign-On
STP
STUN
SUP1
SUP2
SUP2+
SUP3
GL-16
Glossary
T
TAC
TAPI
TCD
TCER
TCL
TCP
TCS
TDD
TDM
Time-division multiplexing
TEHO
Tail-end hop-off
TFTP
TIP
TKIP
TLS
TMS
ToD
Time of day
ToS
Type of service
TPC
TRaP
TRC
TRP
TSP
TTL
Time to live
TTS
Text-to-speech
TTY
Terminal teletype
TUI
TURN
GL-17
Glossary
U
UAC
UAS
UCCN
UCS
UDC
UDLD
UDP
UDPTL
UDS
UMTS
UN
UNC
UP
User Priority
UPS
URI
USB
UTIM
UTP
UUIE
V
V3PN
VAD
VAF
Voice-Adaptive Fragmentation
VATS
VCS
VDI
GL-18
Glossary
VDS
VIC
VLAN
VMO
VoIP
Voice over IP
VoPSTN
VoWLAN
VPIM
VPN
VRRP
VUI
VWIC
VXC
VXI
VXME
W
WAN
WebDAV
WEP
WFQ
WINS
WLAN
WLC
WLSM
WMM
Wi-Fi Multimedia
WMM TSPEC
WPA
GL-19
Glossary
X
XCP
XML
XMPP
GL-20
INDEX
Symbols
! in route patterns
@ in route patterns
+ dialing
AA
14-27
19-22
AAR
14-26
14-57
14-74
5-31
Numerics
AC
5-11
14-79, 21-41
8-8
14-19
8-34
8-34
4-13
8-9, 8-16
8-9, 8-16
14-26, 14-27
14-19
AD
8-9, 8-16
4-22, 4-28
addresses
21-22
MAC
4-7
security
4-5
security issues
4-4
14-69, 15-1
16-12, 16-30
flat
3-12
3-4, 3-7
4-21
ADAM
3-4
802.1X authentication
911 calls
8-34
21-44
21-52
8-34
802.3af PoE
8-5
8-5
8-8
802.1w
4-21
3-4
802.1s
19-7
8-10
8-17
10-21
18-42
access codes
8-17
14-70, 14-78
AD LDS
3-71, 4-11
16-22
27-6
3-58
IN-1
Index
AFT
application users
15-24
agent desktop
architecture
23-14
10-23
11-42
call processing
25-6
ALI
Cisco Jabber
alias normalization
all trunks busy
12-2
9-2
8-25, 20-6
8-29, 25-20
15-24
15-14
analog
connection types
8-6
endpoints
8-5
gateways
5-2, 8-5
collaboration system
8-6
standalone gateways
Analysis Manager
Analytics
interface modules
Android
directories
27-25
IP Phone Services
media resources
7-16
answer supervision
presence
21-88
21-64
trunks
21-52
6-3
WebDialer
18-42
area code
Extension Mobility
for mobile users
18-7, 18-28
21-1
IP Manager Assistant
IP Phone Services
18-19
18-2
18-34
ARP
3-71, 4-11
ASA
4-22, 4-28
ASR
11-42
Assurance
18-19
3-61
14-79
Assistant Console
4-34
3-61
18-34, 18-37
wireless LAN
18-1
18-32
27-4
10-57
21-45
applications
Attendant Console
24-2
20-18
8-38
WebDialer
21-54
security
7-2
described
18-2
21-93
AnyConnect VPN
21-54
18-8
Apple iOS
8-2
Extension Mobility
AnyConnect
10-3
annunciator
25-20
16-7
endpoints
27-12
22-2
2-1
deployment models
8-6
AP
17-2
14-74
ANI
16-7
17-1
IN-2
Index
18-42, 25-28
audio conferencing
11-15
for WebEx
audio on computer
8-27
general rule
audio sources
authentication
3-34
provisioning
3-63
mechanisms
of phones
16-34
16-16
bearer-caps command
bearer traffic
14-70, 14-78
5-14
3-53
best-effort bandwidth
5-31
3-72
9-34
19-22
15-5, 15-24
Cisco Unity
15-3
19-31
25-6
LDAP synchronization
27-6
music on hold
voice messaging
WAN design
3-6
bandwidth
best-effort
3-35
3-58
for conferencing
for contact center
BGP
11-46
10-23
16-19
10-10
19-31
3-33
BHCA
BHCC
25-5
23-5
19-31
11-49
blade servers
22-16
11-46
BIB
BFD
19-45
7-39
single-site deployment
advanced formulas
19-31
15-11
BackboneFast
10-14
16-16
3-35
10-21
19-7
21-79
21-79
beacons
4-17
BDI
13-45
3-48
16-9, 16-22
4-17, 8-35
auto-detection
13-1
16-39
AXL
10-42
management of
database
of users
11-49
guaranteed
7-40
13-67
3-59
BLF
11-46
15-5
10-54
20-16
IN-3
Index
blocking factor
blocking numbers
Bluetooth
25-6
21-52
for TelePresence
for video
11-46
links
3-6
BTN
4-25
3-51
bursty traffic
Business Edition
business-to-business communications
busy hour
13-46, 13-47
SIP trunks
13-61
topologies
13-74
weights
25-6
13-52
13-41
call anchoring
21-56
callback
10-35
25-5
25-5
15-11, 15-17
22-1
21-75
caller ID matching
caller ID transformations
20-44
music on hold
13-1
bandwidth requirements
13-45
effective path
elements
21-72
call handoff
21-72, 21-80
14-70
21-72, 21-80
14-28
13-41
13-40
13-40
call hand-in
call hand-out
13-74
13-75
enhanced locations
7-25, 7-28
13-1
7-23, 7-26
7-23, 7-26
13-40
design considerations
21-59
call flows
described
10-57
3-57, 3-60
components
15-11, 15-17
20-16
15-15, 21-14
13-76
replication network
23-5
13-72
13-41
regions
13-81
15-5
bursting
13-41, 13-42
MPLS cloud
10-54, 10-56
BYOD
21-75
18-47
13-60, 13-79
3-6
13-83
13-67, 13-79
locations
7-44
broadcast messages
7-41
22-17
13-63
in 911 calls
15-5
localization
14-62
calling privileges
calling restrictions
14-41
14-41
IN-4
Index
signaling
20-17
call processing
CAMA
agents
campus
10-23
architecture
capacity planning
15-6
3-3
deployment model
9-21
design considerations
25-5
4-7
access switch
9-2
centralized
5-14
10-8, 27-27
infrastructure requirements
3-1
capacity planning
9-25
distributed
guidelines
9-1
Attendant Console
hardware platforms
high availability
redundancy
Business Edition
9-4
by product
9-12
subscriber server
call routing
12-1
12-3
call traffic
25-22
25-33
21-68
21-95
25-47
dual control
forwarding
14-45
7-20
codecs
5-29
monitoring
outbound
recording
routing
5-9
14-41
22-2
23-1
21-38
2-4
22-19
CTI applications
9-30, 25-23
deployment models
10-4
25-1
25-22
emergency services
endpoints
25-17
25-42
contact center
21-37
queuing
25-39
conferencing
5-30
privileges
25-14
collaboration system
7-18
preservation of
20-58
23-1
music on hold
25-20
20-17
inbound
25-20
10-38
14-69, 15-1
25-47
14-28
emergency
history
25-47
15-1
classification of
hold
call routing
23-17
calls
911
9-21
15-22
21-67
outbound
9-24, 25-48
9-6
architectural layer
18-45, 25-28
25-13
call processing
5-3, 9-13
17-4
25-34
8-44, 25-16
IN-5
Index
Extension Mobility
factors to consider
gateways
18-17, 25-25
25-9
25-37
deployment model
18-6
megacluster
migration to
performance tuning
CER
24-3
CFUR
25-39
25-14
servers
25-14
sizing tools
CIR
21-88
9-21, 21-60, 25-51
Unified CM servers
Cisco Expressway
18-26, 25-26
4-5
25-44
7-15
21-60, 25-21
videoconferencing
3-12
20-18, 25-33
25-43, 25-44
21-29, 25-35
9-22
Unified MeetingPlace
25-38
Cisco Jabber
Cisco LEAP
8-35
Cisco Mobile
25-41
21-82
11-49
wireless networks
3-67, 8-35
25-21
Cisco Prime
3-62
10-44
CCA
3-72, 11-46
CCD
10-57
CDP
4-5
CDR
27-9
26-10
18-47
27-1
CAR
7-16, 25-28
21-77, 21-82
18-40, 25-27
XMPP clients
3-51
Unified CM Assistant
voice messaging
3-68
25-13
Unified Mobility
4-19
14-70
25-10
Unified CM
CAPWAP
9-22, 25-10
WebEx
3-31
25-33
WebDialer
25-39
8-44
regions
tools
10-21
performance overload
19-11
26-5
centralized messaging
25-32
presence
7-30, 25-28
19-6
distributed messaging
25-31
25-14
media resources
phones
centralized messaging
20-43
25-47
25-47
25-47
27-7
7-9
IN-6
Index
Cisco Spark
8-29, 8-39
8-29, 25-20
calls
4-30, 23-11
15-4
CLID
14-28
clipping
10-53
22-1
cloud architecture
20-12
11-46
described
22-4
22-7
25-43, 25-44
27-26
10-19
local failover
10-45
music on hold
7-47
20-30
remote failover
10-52
troubleshooting
10-45
9-5
19-5
10-42
19-19
design guidelines
26-9
19-18
22-13, 27-31
WAN considerations
10-41
presence
9-29
Cisco Unity
19-14, 19-16
CTI applications
11-42
Cisco Unity
21-63
10-14
14-29
3-72
clients
9-34
3-4
10-23
9-27
14-28
traffic
27-1
25-48
11-50
classification of
27-25
25-20
capacity planning
27-8
for Unified CM
guidelines for
home
15-9, 15-22
20-19
9-5
9-11
18-14
home cluster
18-18
maximum capacity
25-14
IN-7
Index
redundancy
9-15
server nodes
9-5
services
9-5
visiting
18-14
Device Mobility
messaging system
presence
19-2
20-3
14-29
CMR
CMR Cloud
11-69
conferencing
audio
11-58
11-15
collaborative
codecs
25-42
conference bridges
capacity planning
25-39
complexity modes
7-4, 7-5
flex mode
7-5
described
11-1
hardware
9-35
7-39
resources
7-36
security
20-5
traffic
25-17
21-75
8-5
console
2-1
for attendants
described
19-13
3-51
4-19
18-32
gateway sizing
25-38
traffic patterns
25-7
Contact Sharing
8-24
contact sources
22-1
contact management
13-54
18-42
contact center
10-13, 10-22
3-25
16-4
9-34
simplified method
25-21
11-42
Communicator
4-35
8-28, 20-9
25-8
Unified CME
8-25, 20-6
collaborative conferencing
4-29, 11-23
Collaboration Cloud
11-1
25-42
contact management
11-29
11-15
rich media
collaboration
common locations
7-14
conferencing
19-22, 25-23
CMR Hybrid
clients
3-46,
3-48
CMC
COM
21-15
16-4
components of
15-4
8-28
22-8
8-33, 8-42
23-14
Context Service
22-8
4-7
IN-8
Index
control signaling
COP
3-57, 3-60
26-10
Core Layer
campus
co-resident
MoH
3-3
described
3-4
CPN
15-5
DHCP
cRTP
19-4
CTI Manager
for presence
22-10
9-27
media resources
7-14
27-27
20-26
19-22
19-22
CUE
20-10
10-56
4-19
CTI-QBE
20-31
3-46, 3-48
CSR
3-25
25-4
19-13
10-1
federation
25-5
CPU usage
27-31
7-31
CoS
cps
10-8, 27-27
3-26
core switch
4-36
deployment models
3-11
DHCP
20-22
9-36
7-36
19-22
22-20
26-1
CUWL
26-9
CVTQ
27-8
music on hold
single cluster
4-11
database
single site
site-based
10-4
replication
9-8
virtualized servers
redundant
security
4-39
4-26
server farm
single
3-11
Delayed Offer
delay of packets
6-19, 7-9
10-42, 10-44
10-21
10-4
25-9
4-38
10-53
10-53, 10-57
10-57
20-26
25-14
16-30
7-37, 7-46,
10-24
complexity
data centers
7-36, 7-43,
7-42
19-5
cutover
DAI
3-70
8-27
21-37
8-8
IN-9
Index
destination of a call
elements
14-79
device mobility
dial plan
21-20
14-13
21-15
14-83
21-15
operation flowchart
operation of
21-69
21-19
functions
21-19
21-17
fundamentals
Physical Location
21-15
globalized numbers
14-27
14-62
mobility
14-60
14-56
10-46, 10-52
route group
15-17
14-30
14-55, 14-62
21-18
18-28
14-3
international calls
21-18
transformations
18-14
DHCP
14-71
14-57, 14-58
Unified Mobility
3-68
8-32
14-1
parameter settings
pools
21-20
21-15
settings
15-12
21-55
binding information
4-11
deployment options
3-25
described
21-22
3-23
21-73, 21-83
lease times
3-24
21-86
Option 150
3-24
21-84
servers
DID
3-26
Snooping
15-5
4-8, 4-11
starvation attack
4-10
digital gateways
14-57
911 calls
15-1
digit manipulation
21-52
digit prefixing
14-3
14-58
14-41
25-22
design considerations
device mobility
14-70
21-20
21-20
21-53
15-5
directories
access
16-4, 21-79
architecture
14-22
capacity planning
19-28
call routing
5-3
digital networking
13-82
dial plan
+ dialing
14-52
16-7
authentication of users
filtering
16-9, 16-22
16-27
18-32
IN-10
Index
high availability
16-30
DVO
16-1
search base
16-13
7-30
21-73, 21-83
DVO-F
21-86
21-84
searches
8-29
DVO-R
security
16-19
sn attribute
synchronization
URI dialing
UserID
16-10
14-49
Distribution Layer
DNS
3-22
3-22
E911
15-1, 15-4
4-19
13-41
ELCAC
3-48
13-40, 13-86
conversion of
ELIN
7-8
gateway capabilities
on H.323 gateways
on SIP gateways
7-14
7-12
5-5, 7-13
14-13
15-10, 15-11
5-3
7-8
EMCC
4-19
15-23
14-69
15-12
3-71
15-10,
15-11
10-38
13-75
dual-mode
clients
3-71
6-20, 7-9
efficiency of links
7-30
DTMF
DTPC
E.164
effective path
7-4
3-70
Relay
3-35
ECC
DSP resources
DTIM
25-4
Early Offer
PVDM
3-23, 4-8,
described
3-68
3-9
3-35
4-36
4-11
distributed messaging
DSCP
16-16
3-69
DMZ
15-11
4-10, 4-11
DMVPN
8-11
16-10
directory URI
8-16
14-23, 14-50
7-30
16-1, 25-31
schema
16-6
21-63
Emergency Responder
21-77, 21-82
Cisco Collaboration System 11.x SRND
IN-11
Index
eMWI
bandwidth
19-37
encryption
3-57, 3-58
for phones
CPU usage
4-17
for security
for signaling
memory usage
3-58, 3-59
ERL
analog gateways
25-4
25-6
8-2
error rate
8-44, 25-16
design considerations
E-SRST
8-45
25-6
10-45
8-21, 10-14, 10-17
directory access
16-4
ESXi Hypervisor.
high availability
8-44
ettercap virus
immersive video
8-19
mobile
multipurpose video
off premises
end users
7-13
18-7
dial plan
14-83
8-15, 15-16
EMCC
16-7, 20-3
Energy conservation
18-10, 18-18,
13-74
3-12
EnergyWise Technology
3-12
13-40, 13-86
F
19-37
8-21, 10-14
10-17
FAC
21-70
14-28
25-9
failover
Cisco Unity
19-17, 19-18
18-28
25-25
3-64, 8-34
Enhanced SRST
described
25-25
8-1
wireless
8-35
20-50
capacity planning
8-24
supplementary services
types of
8-34
software-based
video
extensible messaging
8-5
4-14
telepresence
8-17
8-15
20-45
8-18
15-17
personal video
26-11
4-11
Expressway
8-38
7-32
Erlang
8-5
capacity planning
25-24
8-35
endpoints
architecture
25-4
4-17
25-49
scenarios
18-5
fallback mode
7-46
Fast Start
10-45, 10-52
7-13
fax
IN-12
Index
gateways
5-3, 5-34
911 services
8-6
15-13
additional documentation
10-54, 10-55
16-16
analog
20-31
10-54, 10-55
5-2, 8-5
blocking
capabilities
16-30
23-6
5-14
capacity planning
25-37
20-58
around gateways
4-28
configuration in Unified CM
4-25
centralized deployment
described
digital
4-24
digit manipulation
stealth mode
4-25
firewalls
with H.323
4-22, 4-28
security
5-3
5-9
4-27
selection of
7-5
5-3
service prefixes
14-28
SIP
15-7
types of
5-31
5-6, 5-11
standalone
14-45
8-6
5-2
voice applications
4-22, 4-28
VoiceXML
15-7
GDPR
21-47, 21-48
4-2
geographical diversity
GARP
geolocations
5-29
GLBP
4-11
GLO
gatekeeper
call admission control
10-7
14-90
3-9
26-7
10-23
5-1, 8-5
general security
G
gain settings
5-11
15-13
redundancy
7-44
placement
protocols
8-11
21-22
10-51
flat addressing
5-30
4-28
4-29
5-5
5-3
routed mode
4-25
5-13
25-38
4-36
4-22
transparent mode
5-31
15-14
call recording
firewalls
access control lists
15-14
20-31
25-40
3-9
14-55, 14-62
IN-13
Index
Business Edition
26-7
call processing
11-42, 11-46
call routing
1-1
9-12
12-3
8-6
groups for
contact center
call routing
CTI
14-30
15-18, 15-20
Unified CM redundancy
22-15
10-4
16-30
8-44
9-13
11-42, 11-46
guaranteed bandwidth
2-3
deployment models
endpoints
7-1
3-34
Extension Mobility
18-15
hardware platforms
9-12
IP Phone Services
media resources
H.245 Signal
music on hold
7-8
7-33, 7-35
network services
H.323
call hairpinning
3-4
phones
5-9
presence
7-13
20-21
4-29
requirements
gateways
5-3
supplementary services
hairpinning
transcoders
voice services
21-72, 21-80
WebDialer
hardware
media resource capacities
MTP resources
music on hold
WebEx
7-30
hold
9-4
3-53
high availability
applications and serviceability layer
Attendant Console
18-44
17-3
10-14
11-46
history of calls
7-31
9-20
18-39
wireless LAN
7-15
types of platforms
18-23
21-72, 21-80
9-15
9-13
Unified CM Assistant
21-72
hand-out of a call
21-45
7-35
Unified CM
9-34, 21-48
handoff of calls
10-5
7-13
6-4
hand-in of a call
24-3
8-44
firewalls
trunks
9-12
9-34
21-55
7-35
network connectivity
7-8
21-55
18-5
H.245 Alphanumeric
21-94
9-31
directories
25-37
media resources
GSB
4-11
9-21
3-65
20-17
7-18, 7-20
holdee
7-19
holder
7-19
home cluster
18-14, 18-18
3-9, 10-23
IN-14
Index
HSRP
3-9, 10-23
HTTPS
19-29
hub-and-spoke topology
introduction
3-3, 3-33
1-1
20-13
3-19, 10-53
4-28
7-15
IP Communicator
I/O modules
IButton
iPhone
10-55
IPMA
14-21
ICCS
ICMP
5-11
18-19
IP Precedence
16-32
16-32
IDS
4-28, 10-43
immediate start
3-4, 3-47
10-13, 10-22
security
8-6
5-29
InformaCast
18-47
3-18
ISDN
IVR
8-19
Intelligent Proximity
10-10
8-19
11-29
21-57
10-10
Jabber
8-6
call handoff
15-5
clients
3-69
14-27
inter-VLAN routing
5-3
international calls
10-14, 10-15
ISDN Link
20-1, 20-37, 20-43
9-34
3-12
Integrator C Series
19-42
25-28
interface modules
4-5
8-19
instant messaging
10-13, 10-22
IPv6
20-1, 25-33
interoperability
18-2, 25-25
IM and Presence
18-19
8-8
IP Phone Services
IPSec
IdP
8-24
IP phones
16-1, 16-32
5-11
4-4
identity management
inline power
9-8, 10-43,
10-47
8-21, 8-32
21-80
25-18
deployment models
Desktop Client Cache
5-11
desktop clients
desktop video
20-10
8-28
21-83
IN-15
Index
link efficiency
21-63
21-92
21-82
20-3
8-38, 21-77
13-41, 13-42
15-5
3-22
load balancing
3-30, 9-18
4-35
15-3, 15-13
14-62
14-62
10-45
20-52
3-46, 3-47
LMHOSTS file
9-19
14-60
14-31, 14-56
LAN infrastructure
3-4, 10-23
common
Layer 3
3-4
defined
LBM
13-54
13-41
enhanced
4-3
13-41, 13-48
LBM Hub
13-41, 13-52
13-40
13-81
maxiumum number
25-14
LBR
7-36
shadow location
LCR
5-33
shared
LDAP
LDN
logical partitioning
8-35
LWAPP
15-3, 15-13
Lync
7-36
3-59
3-46, 3-47
3-62
8-29
3-62
13-41, 13-52
8-6
5-33
13-41, 13-48
14-59, 14-90
3-24
loop start
13-54
15-5
leased lines
13-56
LEAP
13-57
locations
3-4
Layer 2
layers of security
LLQ
10-42
JTAPI
3-48
21-78
20-3
jitter
3-50
9-8, 16-1,
M
MAC address
3-71
4-7
20-39
14-24
IN-16
Index
failover
15-3
redundancy
25-5
5-3
7-33
7-33
10-51
high availability
7-30
MISTP
MLP
9-7
7-38
23-9
media transparency
MLTS
15-2
21-86, 21-88
21-33
architecture
6-25
described
11-58, 11-69
8-38
hairpinning
25-4
19-22
21-52
21-54
functionality
21-47
21-48
messaging
bandwidth management
redundancy
19-31
deployment models
21-47
21-52
21-55
Mobility
19-1
7-16
access numbers
6-7
Cisco Unity
MLPP
centralized
26-1
3-46
mobile endpoints
7-7
memory usage
26-1
3-4
described
7-14
7-15
megacluster
13-72
Mobile Connect
types
conference bridges
19-5
21-39, 21-71
to IP Telephony
7-33, 7-35
4-27
described
20-52
migration
7-30
MediaSense
20-52
8-29, 25-20
mid-call features
voice quality
design guidelines
server
7-36
7-1
security
5-3
Microsoft Lync
7-30, 25-28
deployment models
PVDM
20-39
7-2
capacity planning
described
19-2
7-2
media resources
architecture
MFT
MGCP
7-33
19-17
system components
27-7
19-17, 19-18
19-4
19-13
applications
21-1
21-63
21-1, 21-55
IN-17
Index
dial plan
21-69
emergency services
21-70
21-60
multiserver certificates
20-47
MOS
23-1
19-22
8-18
15-7
NAM
7-33
MRGL
MRM
5-3, 5-34
27-7
MPLS cloud
MRG
27-30
27-28
monitoring calls
4-19
8-18
21-80
21-42
MoH
10-22
7-33
NAT
7-2
MSAG
5-6, 7-8
4-25
15-10,
15-24
15-3
MTP
conference bridges
described
7-14
13-79
network hold
6-7
core layer
18-47
LAN
7-16
3-4
3-4
requirements
3-46
roles
14-34
3-9
network management
15-2
3-69
3-11
high availability
16-22
3-4
distribution layer
22-7
27-9
7-20
access layer
3-70
4-25
network infrastructure
19-32
15-10, 15-24
multipath distortion
7-15
7-15
multichannel support
15-23
NENA
7-15
software resources
with SIP trunk
7-7
hardware resources
types
27-9
3-4
27-4
3-1
3-3
3-7
4-4
21-65
3-33
IN-18
Index
wireless LAN
WLAN
21-65
3-61
network management
network services
packets
22-21, 27-1
delay
3-22
headers
3-32
5-29
jitter
3-19
7-44
10-42
10-42
Paging Server
18-47
paging systems
8-7
PAK
normalization
of aliases
3-53
loss of
4-37
10-42, 10-44
26-10
parallel cutover
14-74
26-3
NPA
14-79
NTE
5-6, 7-8
NTLP
NTP
partitions
5-29
21-53
passive-interface command
3-32
number blocking
3-11
21-52
PC port on IP phone
14-79
call rate
4-14
9-1
designing for
modeling
25-9
25-3
OpenAM
20-52
of WebDialer
open authentication
persistent chat
25-2
20-43
OVA templates
phased migration
5-30
phone books
9-25
25-39
25-39
11-69
outbound calls
8-15
26-3
27-21
phones
overlap
of channels
sending
24-1
performance testing
3-23, 3-24
4-24
receiving
9-25
18-26
18-40
tuning of gateways
4-24
OSPF
20-25
overload on gateways
8-35
9-22
18-17
of Unified CM Assistant
20-11
20-5
Option 150
of Extension Mobility
of presence servers
15-17
13-41
performance
14-57, 14-58
14-49, 14-51
off-premises endpoints
21-17
3-68
14-27
14-27
oversubscription of a link
3-51
3900 Series
8-10
7800 Series
8-8
7900 Series
8-8
8800 Series
8-9, 8-16
Cisco Collaboration System 11.x SRND
IN-19
Index
8900 Series
8-9, 8-16
ping utility
9900 Series
8-9, 8-16
PIX
Attendant Console
polling model
PortFast
3-12
18-7
firmware upgrades
8-11
security
POTS
3-13
3-12, 8-12
3-13
3-13
secure mode
4-14, 4-35
services
18-2, 25-25
settings
4-16
software-based
14-16
Type-B
14-18
14-79
5-31
presence
calendar integration
Type-A
3-4, 3-47
prefixes
18-14
security
call history
8-24
20-44
20-17
capacity planning
25-33
4-16
WebDialer
18-34
18-19
clusters
described
end user
8-34
8-34
20-3
deployment models
8-34
20-30
20-19
components
physical security
15-7
14-15
wireless
4-6
21-38
3-68
user input
4-14
3-13
SIP
4-7
on the IP phone
4-14
roaming
20-49
21-39
20-17
3-6
access
18-2
4-2
ports
8-44
mid-call features
3-12, 8-12
for presence
8-8
IP Phone Services
9-4
21-63, 21-94
high availability
15-7
policy
Extension Mobility
SCCP
PoE
21-37
8-45
energy conservation
PC port
platforms
4-17
8-44
design considerations
4-22, 4-28
18-42
10-44
20-22, 20-26
20-1, 20-2
20-3
8-34
4-4
groups
20-45
20-31
20-17
guidelines
20-18
20-43
IN-20
Index
20-26
20-41
propagation of database
20-52
protocols
26-15
mobility integration
policy
20-47
20-17
polling model
presentity
20-48
20-16
server guidelines
20-50
server performance
server redundancy
20-25
20-21
20-18
server synchronization
SIP
20-19
20-16
20-17
20-19
20-48
20-14
BGP
11-46
3-62
4-5
3-46, 3-48
3-23, 4-8, 4-10, 4-11
GARP
4-11
GLBP
3-9
H.323
HSRP
3-9, 10-23
IPSec
10-13, 10-22
LDAP
15-5
primary extension
20-3
5-3
MISTP
3-4
MLP
3-46
NTP
3-32
RCP
4-12
25-47, 27-12
25-47
26-3
27-1
26-9, 27-22
26-9, 27-22
prioritization of traffic
3-47
4-22, 4-28
14-41
3-11
RSTP
3-4, 3-7
RSVP
3-33
10-23
5-3, 7-9, 7-23, 14-15, 20-16
SIMPLE
20-18
SIP 5-6, 5-11, 6-4, 6-6, 6-7, 7-17, 7-26, 8-43, 9-38, 10-23, 14-16,
14-18, 14-19, 20-14
SMTP
19-27
SNMP
15-7
SOAP
20-19
SRTP
3-53, 4-17
STP
TFTP
15-3
4-24
SCCP
25-47, 27-2
MGCP
RTP
15-5
Prime compliance
3-62
routing
5-9
15-14
9-8
DHCP
RIP
20-2
preservation of calls
Prime LM
11-46
LWAPP
20-14
speed dial
BFD
cRTP
20-50
servers
3-71, 4-11
CDP
20-2
protocol interfaces
SCCP
ARP
CAPWP
20-49
26-10
PRI
20-52
TLS
3-6
3-24, 3-27, 9-5, 9-19
4-17
IN-21
Index
UDP
10-23
3-73
VPIM
19-27
VRRP
3-9
provisioning servers
9-22
proxy
8-7
18-20
PSTN
for LAN
911 calls
destination number
traffic patterns
3-14
14-79
for security
25-37
7-41
4-20
for video
10-13, 10-22,
14-79, 15-2
9-5, 10-43
for WAN
7-30
3-33, 3-36
8-37
3-73
QBSS
22-2
QoS
3-17, 3-75
8-13
3-14
21-67
8-41
7-41
4-20
8-31
18-32
8-24, 13-79
8-35
for security
3-19
22-16
9-28, 19-22
8-7
7-38
3-60
queuing of calls
3-72, 3-76
18-32
8-20
8-31
8-24, 13-79
publisher server
21-67
8-41
10-21
PVDM
8-13
10-13, 10-22
3-19
22-16
15-2
QBE
3-72, 3-76
rate of error
RBAC
27-4
RBOC
15-3
10-45
RCC
20-18, 20-52
RCP
4-12
RDNIS
19-7
3-33, 3-36
8-37
3-4, 3-7
20-48
16-3
10-23
recording
IN-22
Index
replication network
22-7, 23-3
replication of database
23-1
MediaSense
23-2
9-8
23-9
SPAN method
18-35
resilience
20-48
21-55
9-1
redundancy
3-33
20-48
restrictions for
call processing
9-13
Extension Mobility
cluster configurations
Extension Mobility
for messaging
9-15
WebDialer
21-55
RF
10-14
21-45
18-23
IP Phone Services
TFTP services
3-30
8-35
5-6, 7-8
18-5
9-18
18-41
5-3, 5-9
load balancing
11-1
21-57
4-24
RMON
27-9
roaming
3-68
18-39
regions
15-3
DHCP server
4-8
network extensions
13-46, 13-47
25-14
20-18, 20-52
4-12
phone pickup
21-52
21-38, 21-51
3-3
root guard
3-6
10-44, 10-47
3-7
4-24
branch office
21-25
flash
22-20
9-27
27-4
routers
21-55
Remote Device
4-8
remote destination
caller ID matching
21-17
rogue
maximum number
18-28
23-14
RFC 2833
WebDialer
Rev
20-21
18-7
Unified CM Assistant
19-17
18-18
IP Phone Services
18-15
profile
13-52
27-9
10-14
re-packetization of a stream
7-7
10-52
4-21
7-44
7-44
3-3
15-4
routes
filters
14-27
group devices
14-30
IN-23
Index
groups
lists
presence
14-28, 14-30
14-29
patterns
20-16
schema
14-22, 14-26
selection of
SDK
14-81
routing
14-15
16-1
16-4
calling line ID
calls
14-28
14-22, 21-67
digit manipulation
inbound calls
least-cost
8-39
18-14
3-11
14-89
4-24
RSVP
3-33
4-19
4-29, 11-23
data center
4-35
4-26
DHCP Snooping
16-3, 27-25
4-8
RTP
10-23
RTT
10-44, 10-47
directories
4-14
Extension Mobility
11-42
architecture
described
SAML
SAN
firewalls
4-22, 4-36
gateways
4-27
infrastructure
SAF
in general
10-57
scalability of
IP Phone Services
18-6
4-1, 4-2
4-35
4-7
phones
SCCP
7-9
4-5
4-27
3-48
5-3
7-23
14-15
4-16
physical access
4-4
QoS
4-14
4-14
phone settings
policy
9-10
4-3
media resources
9-1
phones
4-4
IPv6 addressing
layers
4-19, 10-56
DTMF signaling
18-13
intracluster communications
10-57
Unified CM
4-10
16-19
endpoints
4-21
4-30
configuration example
WAN infrastructure
SaaS
3-4, 3-7
RTMT
4-17
security
5-30
time-of-day (ToD)
RSTP
8-5
5-33
protocols
Section 508
8-21, 8-32
outbound calls
8-5
14-28
5-29
inter-VLAN
Section 255
16-13
4-2
4-20
IN-24
Index
architecture
4-8
described
4-34
switch port
WebEx
for IP phones
4-16
prefix
20-57
16-32,
4-34
5-31
supplementary
4-34
18-47
Redirector
18-35
WebDialer
18-34
9-22, 25-14
9-5, 20-19
co-located
3-25
annunciator
7-17
CTI Manager
data center
3-26
early offer
7-9
5-6
3-26
trunks
7-31
20-18
paging server
performance
19-21
Type-A phones
14-16
Type-B phones
14-18
shadow location
20-21
shaping traffic
4-34
standalone
3-26, 7-31
subscriber
9-6
9-6, 9-19
4-16
13-56
3-50
shared
line appearances
20-19
23-14
9-5, 10-43
synchronization of
20-14
18-47
9-22, 20-25
redundancy
7-26
presence
7-1
10-23
5-11
TFTP
14-19
gateways
9-19
3-11
for DHCP
security
dial rules
7-31
3-11
publisher
7-9
co-resident DHCP
co-resident MoH
3-68, 3-71
delayed offer
clusters
9-5
servlet for
15-3, 15-4
capacity planning
5-5
14-81
farm
18-2
within a cluster
services
4-20
web access
21-78
4-15
VPN clients
10-57
service discovery
4-6
voice VLAN
10-57
3-59, 15-17
18-21
13-54
23-14
IN-25
Index
signaling encryption
signal strength
3-58, 3-59
5-29
site-based design
22-7, 23-3
20-18
15-7
21-75
sizing
methodology
4-38
tool
19-43
21-33, 21-36
Unified CM servers
messaging model
dial rules
SNMP
7-9
SNR
7-9
10-23
16-10
4-8
21-33, 21-36
SOAP
20-19
SocialMiner
gateways
software
5-11
5-6
7-11
route pattern
22-6
14-49
7-30
7-15
Type-A phones
14-16
Type-B phones
14-18
SPAN
11-42
16-4
23-2, 23-3
8-27
8-24
14-29
routing requests
trunks
7-26
14-15
15-7
snooping
14-19
Early Offer
20-16
19-27
sn attribute
7-17
7-23
SMTP
delayed offer
5-3
14-15
presence
SME
18-47
SIP
annunciator
7-9
19-4
Singlewire InformaCast
9-22
phones
deployment model
25-9
16-1
25-1
25-2
DTMF signaling
21-90
25-17
20-26
8-35
20-19
SIW
19-27
single inbox
10-4
3-6
20-16
21-88
IN-26
Index
SRST
7-44, 8-13, 8-21, 8-32, 8-37, 8-41, 8-43, 9-15, 10-13, 10-14,
10-17, 15-4
SRST Manager
10-19
SRSV
19-8
TAPI
9-19
SRTP
3-53, 4-17
TCS
23-14
SSID
3-68, 3-71
TEHO
SSO
11-63, 16-1, 16-32, 16-33, 20-5, 20-7, 20-36, 20-56, 21-90, 25-19
15-11
call routing
3-6
19-5
11-66
dial plan
14-52
4-19
endpoints
20-17
interoperability
9-6
13-60
14-52
16-10, 16-16
TelePresence Conductor
supplementary services
8-24, 13-79
11-30, 11-36
design considerations
for H.323 endpoints
on gateways
19-5
TelePresence
4-25
subscriber server
8-5
Telecommunications Act
25-4
stealth firewall
14-71
3-26, 7-31
8-6
14-71
9-37
7-13
5-3
5-5, 5-6
23-14
8-35
4-6
3-3
8-18
25-4
9-4, 10-53
IP phones
20-19
Unified CM database
15-17
third-party
16-9, 16-10
presence servers
8-19
7-4
synchronization of
system memory
19-8
switches
directories
23-2, 23-3
16-30
SIP phones
8-43
8-43
third-party CA certificates
Third-Party Open API
4-19
20-48
20-59
IN-27
Index
14-89
25-21
21-42
translation of digits
14-24
translation patterns
14-24
5-14
4-25
3-32
TRaP
4-17
4-17
19-5
TRC
3-68
9-4, 10-53
TMSBA
27-19
TMSPE
11-61, 27-21
TMSXE
11-61, 27-19
TRP
TMSXN
27-19
trunks
ToD
described
3-68
tracking domain
traffic
SIP
call control
TSP Audio
25-8
25-5, 25-6
prioritization
3-53
3-17, 3-75
shaping
3-50
11-49
3-47
provisioning for
TUI
voice calls
19-5
8-20
8-19
Type-A phones
14-16
Type-B phones
14-18
25-37
U
3-56
UCS
high availability
3-53, 25-6
QoS
25-6
UDLD
19-32
UDS
resources
7-7
UN
transformations
21-59
14-57, 14-58
10-53
3-6
3-48, 10-23
7-6
3-19
UDP
described
caller ID
9-20
virtualized servers
transcoding
Cisco Unity
21-50, 21-52
25-8
11-66
two-stage dialing
25-7
27-10
3-57, 3-60
classification
6-4
utilization of
3-53
6-4
6-1
features supported
15-21, 15-22
bearer traffic
6-3
13-74
10-45
architecture
14-89
5-6
unassigned DNs
14-68
7-25, 7-28
IN-28
Index
3-6
27-25
4-30, 23-11
Unified CCX
22-6
22-6
Unified CM
call recording and monitoring
capacity planning
23-10
Unified Mobility
18-19, 25-26
Unified SM
9-34
Unity
9-34
27-1
9-22
Unity Express
19-6, 19-17
19-22
UPS
24-1
10-53
19-39,
16-16
7-8
5-6
3-74
3-6
3-12
URI dialing
18-38
3-19
3-12
19-41
UplinkFast
1-1
virtualized servers
Unity Connection
UP
12-1
22-7
10-14
22-7
10-23
high availability
27-7
27-7
Unified WIM
9-27
27-26
25-48
27-13
19-1
27-13
10-23
25-43, 25-44
Unified Reporting
9-27
22-7
22-7
25-48
design considerations
22-4
22-7
Unified PM
capacity planning
QoS
22-7
Unified CME
introduction
Unified EIM
Unified MeetingPlace
9-22
Unified CM Assistant
22-4
20-14
sizing tool
Unified CVP
Unified IC
16-30
10-46, 10-52
presence
22-6
25-13
database synchronization
groups
22-3
22-3
Unified CCMP
22-1
21-43
3-48, 10-23
IN-29
Index
user hold
UserID
VLAN
7-20
4-5
16-10
dial plan
3-74
users
14-52
directory integration
application users
end users
security
16-13
14-73
4-32
16-7
input on phones
UTIM
16-7
16-31
videoconferencing
25-44
19-39, 19-41
27-18
19-5
virtualization
of call processing
9-3
V3PN
10-13, 10-22
VAD
25-38
virtualized servers
VAF
3-49
21-22
3-51
VBrick Rev
14-52
directory integration
virtual machine
26-10
virtual network
4-37
4-37, 10-13, 10-22
16-31
14-73
visiting cluster
4-32
3-4, 3-68
VCS
security
27-18
18-14
VLAN
4-21
video
bandwidth utilization
3-56
23-9
customer care
22-20
endpoints
8-15, 15-16
gateways
5-11
interoperability
13-67, 13-79
video
4-5
voice
4-5, 4-15
VMO
VMware
3-19, 10-53
voice
bearer traffic
migration to Unified CM
26-12
3-68
19-5
bandwidth requirements
traffic characteristics
3-4
13-67
3-9
3-19
23-3
bearer traffic
8-42
10-53, 26-10
23-14
dial plan
19-30
21-65
8-24, 13-79
25-8
3-16
gateways
3-53
5-1, 8-5
port integration
termination
traffic
VLAN
3-48
19-39, 19-41
7-4
25-6
4-5, 4-15
IN-30
Index
25-38
W
3-49
WebDialer
19-1
WebEx
19-22, 19-28
third-party systems
unified messaging
WEP
27-8, 27-10
10-21
23-1
22-2
10-1
14-2
endpoints
8-2
gateways
5-1
15-2
4-37
21-3
network infrastructure
3-3
network management
27-2
20-2
3-19
23-3
security
16-2
7-2
mobility applications
presence
3-9
vSphere
13-1
9-2
emergency services
21-47, 21-48
19-27
VRRP
call processing
dial plan
25-6
3-53
VRF
13-41
8-35
deployment models
15-14
VXI
19-27
7-38
VPN-less access
3-47
10-21
VoPSTN
20-54
3-53
11-50
21-42
19-1, 25-41
VoiceXML (VXML)
8-39, 21-93
19-1
11-42
25-20
WebEx Messenger
19-47
voice traffic
voice quality
4-16
18-34, 25-27
WebEx Meetings
19-43
3-33
WebEx Connect
19-27
voice messaging
3-3
10-51
21-42
single inbox
VPN
aggregation router
21-42
Cisco Unity
VPIM
10-13, 10-22
infrastructure
voicemail
VoIP
WAN
3-51
11-3
4-1
sizing considerations
VXME
8-42
system migration
VXML
21-47, 21-48
trunks
25-2
26-2
6-2
Unified CM applications
voice messaging
18-2
19-2
IN-31
Index
white list
20-58
3-75
14-26, 14-27
3-26
8-6
3-26
8-35
wireless
access points
endpoints
3-62
3-64, 8-34
IP Phone 7921G
8-34
IP Phone 7925G
8-34
IP Phone 7925G-EX
IP Phone 7926G
IP phones
LAN
8-34
8-34
8-34
3-61
3-63, 3-72
3-61, 8-40
3-61, 8-40
3-63, 3-72
3-75
WMM TSPEC
3-76
X
XMPP clients
20-59, 25-21
IN-32