NSP Network Services Platform

NSP
Network Services Platform

Network Functions Manager - Packet (NFM-P)
Release 18.3
NFV Solutions Guide
3HE-14080-AAAA-TQZZA
Issue 1
March 2018
NFM-P
Legal notice
Nokia is a registered trademark of Nokia Corporation. Other products and company names mentioned herein may be trademarks or
tradenames of their respective owners.
The information presented is subject to change without notice. No responsibility is assumed for inaccuracies contained herein.
© 2018 Nokia.
Release 18.3
March 2018
2 3HE-14080-AAAA-TQZZA Issue 1
Contents NFM-P
Contents
About this document............................................................................................................................................5
Part I: Getting started ...........................................................................................................................................7
1 Safety information..........................................................................................................................................9
1.1 Structure of safety statements ............................................................................................................9
2 What’s new?..................................................................................................................................................11
2.1 What’s new in NFM-P Release 18.x?................................................................................................11
3 NFV solution overview.................................................................................................................................13

3.1 Network function virtualization overview ...........................................................................................13
Part II: CBAM as VNF manager solution ..........................................................................................................15
4 CBAM as VNF manager ...............................................................................................................................17

4.1 CBAM NFV solution ..........................................................................................................................17
4.2 CBAM VNFD requirements ...............................................................................................................20
5 Network Supervision web application .......................................................................................................23

5.1 Network Supervision web application ...............................................................................................23
6 VNF threshold policy templates .................................................................................................................33

6.1 Policy template design ......................................................................................................................33
6.2 Policy template sample .....................................................................................................................38
7 CBAM GNE driver information....................................................................................................................41

7.1 CloudBand Application Manager driver version 1.0.0 .......................................................................41
7.2 Discovery and management .............................................................................................................42
Part III: NFM-P as VNF manager solution .........................................................................................................43
8 NFM-P as VNF manager...............................................................................................................................45

8.1 NFM-P NFV solution .........................................................................................................................45
8.2 Supported VNFs................................................................................................................................47
9 VNF Manager web application ....................................................................................................................51

9.1 VNF Manager web application ..........................................................................................................51
10 NFV use cases..............................................................................................................................................57

10.1 Automatic scale-out and healing .......................................................................................................57
10.2 To configure automatic scale-out or automatic healing .....................................................................58
Release 18.3
March 2018
Issue 1 3HE-14080-AAAA-TQZZA 3
Contents NFM-P
11 NFM-P VNF descriptor .................................................................................................................................61

VNF descriptor .............................................................................................................................................61
11.1 NFM-P NFV requirements.................................................................................................................61
OpenStack Heat templates..........................................................................................................................62
11.2 Heat templates ..................................................................................................................................62
11.3 Deployment template design ............................................................................................................63
11.4 Scaling template design ....................................................................................................................66
Meta file configuration and requirements..................................................................................................71
11.5 Meta file configuration .......................................................................................................................71
11.6 Other requirements ...........................................................................................................................75
11.7 To enable OpenStack message logging............................................................................................75
Release 18.3
March 2018
About this document NFM-P
About this document

Purpose
The NSP NFM-P NFV Solutions Guide describes the network function virtualization (NFV) solution
and provides information and workflows for managing and monitoring virtualized network functions
with the NFM-P.
Safety information
For your safety, this document contains safety statements. Safety statements are given at points
where risks of damage to personnel, equipment, and operation may exist. Failure to follow the
directions in a safety statement may result in serious consequences.
Document support
Customer documentation and product support URLs:
• Customer Documentation Welcome Page
• Technical support
How to comment
Documentation feedback
• Documentation feedback
Release 18.3
March 2018
About this document NFM-P
Release 18.3
March 2018
Getting started NFM-P
Part I: Getting started
Overview
Purpose
This part provides information on the NFV features added in the most recent releases of the NFM-P
and provides an overview of the NFV solutions.
Contents
Chapter 1, Safety information 9

Chapter 2, What’s new? 11
Chapter 3, NFV solution overview 13
Release 18.3
March 2018
Getting started NFM-P
Release 18.3
March 2018
Safety information NFM-P
Structure of safety statements
1 Safety information
1.1 Structure of safety statements

Overview
This topic describes the components of safety statements that appear in this document.
General structure
Safety statements include the following structural elements:
L
CAUTION
MP E
Lifting hazard
SA
Lifting this equipment by yourself can result in injury
due to the size and weight of the equipment.
Always use three people or a lifting device to transport
and position this equipment. [ABC123]
Item Structure element Purpose

1 Safety alert symbol Indicates the potential for personal injury
(optional)
2 Safety symbol Indicates hazard type (optional)
3 Signal word Indicates the severity of the hazard
4 Hazard type Describes the source of the risk of damage or
injury
5 Safety message Consequences if protective measures fail
6 Avoidance message Protective measures to take to avoid the hazard
7 Identifier The reference ID of the safety statement
(optional)
Release 18.3
March 2018
Safety information NFM-P
Structure of safety statements
Signal words
The signal words identify the hazard severity levels as follows:
Signal word Meaning

DANGER Indicates an extremely hazardous situation which, if not avoided, will
result in death or serious injury.
WARNING Indicates a hazardous situation which, if not avoided, could result in
death or serious injury.
CAUTION Indicates a hazardous situation which, if not avoided, could result in
minor or moderate injury.
NOTICE Indicates a hazardous situation not related to personal injury.
Release 18.3
March 2018
What’s new? NFM-P
What’s new in NFM-P Release 18.x?
2 What’s new?
2.1 What’s new in NFM-P Release 18.x?

Overview
This section highlights new NFV management features for NFM-P Release 18.x and provides
pointers into the documentation for information about using the features. Feature lists and high-
level feature descriptions are also available in the NSP NFM-P Release Description.
Maintenance releases
Some releases may not be listed in this section, either because no new features are introduced or
the features introduced do not require documentation.
What’s new in NFM-P 18.3

Table 2-1 NFM-P 18.3 features
Feature Description and Reference
Discovering CBAM as a GNE Starting with NFM-P Release 18.3, the GNE drivers are installed with the NFM-P
software. It is no longer necessary to obtain these drivers separately from the NSP
software delivery site. See Chapter 7, “CBAM GNE driver information”for information
about driver capabilities and driver-specific discovery and management information.
Condensed VNFC count KPI A count of components in a VNF is now displayed as a single KPI on the matrix view
tile. The # Components (Complete/VNFC Only/Card Only) KPI displays component
counts separated by / characters. See Chapter 5, “Network Supervision web
application”.
Operational and administrative state KPIs The operational and administrative states of a VNF are displayed on a tile as a KPI.
VNF submenu in the Network Supervision You can perform VNF-specific actions from the VNF submenu of the View More menu
application on a matrix view tile in the Network Supervision application. The following actions have
been moved to the VNF submenu:
• Add/Delete Policy to NE
• Scale
• Refresh
• All Application KPIs
See Chapter 5, “Network Supervision web application”
NSPF-85197 | VIM location parameter The location of a VNFC is displayed on the tile in the Network Supervision application,
and you can search for objects by location using the search bar. See Chapter 5,
“Network Supervision web application”.
NSPF-137936 | CBAM IPv6 support You can specify an IPv6 address when configuring a CBAM Access Point. If the CBAM
access point uses an IPv6 address, the CBAM must be able to resolve the NFM-P
hostname. See Chapter 4, “CBAM as VNF manager”.
Release 18.3
March 2018
What’s new? NFM-P
What’s new in NFM-P Release 18.x?
Release 18.3
March 2018
NFV solution overview NFM-P
Network function virtualization overview
3 NFV solution overview
3.1 Network function virtualization overview

General information
Network function virtualization (NFV) allows network administrators to uncouple network functions
from underlay hardware NEs so that the functions can run as software images. These network
functions include load balancers, firewalls, and NAT. The purpose of NFV is to provide a simpler
way to deliver and manage the network components required for a virtualized infrastructure.
Network administrators are able to dynamically deploy network elements and services without
needing to physically provision the underlying routers. The virtualized network element that
represents the physical node is called a virtualized network function (VNF).
The NFM-P currently provides two separate solutions for NFV. The solutions exist independently
and do not need to be used in conjunction with one another.
The two solutions are differentiated by where the VNF manager component resides:
• CBAM as VNF manager — the VNF manager component resides within the external application
CBAM to which the NFM-P provides an interface using the Network Supervision application
• NFM-P as VNF manager — the VNF manager component resides within the NFM-P architecture
and can be accessed using the VNF Manager application
In future releases of the NFM-P, CBAM will be used as the primary VNF manager application for the
NFM-P solution.
Release 18.3
March 2018
NFV solution overview NFM-P
Network function virtualization overview
Release 18.3
March 2018
CBAM as VNF manager solution NFM-P
Part II: CBAM as VNF manager solution
Overview
Purpose
This part provides information on the CBAM as VNF manager solution
Contents
Chapter 4, CBAM as VNF manager 17

Chapter 5, Network Supervision web application 23
Chapter 6, VNF threshold policy templates 33
Chapter 7, CBAM GNE driver information 41
Release 18.3
March 2018
CBAM as VNF manager solution NFM-P
Release 18.3
March 2018
CBAM as VNF manager NFM-P
CBAM NFV solution
4 CBAM as VNF manager
4.1 CBAM NFV solution

Overview
The Nokia CloudBand Application Manager (CBAM) is a VNF manager that automates VNF
lifecycle management and cloud resource management. CBAM has standards-based APIs to allow
it to work with any vendor VNF, EMS, VIM, or NFV Orchestrator (NFVO).
The NFM-P provides an interface with CBAM, which acts as the VNF manager in this solution. The
NFM-P uses the Ve-Vnfm-Em reference point to exchange notifications on VNF lifecycle changes
and monitor virtual resources with CBAM. The Network Supervision application allows you to
configure a CBAM access point to monitor managed VNFs and execute lifecycle management
actions.
The NFM-P provides the following functions when interfacing with CBAM:
• element management for VNFs
• VNF assurance, alarm monitoring, and status tracking
• VNF KPI monitoring
• lifecycle management proxy actions
• policy-based lifecycle changes
Note: When using CBAM as the VNF manager, the NFM-P VNF Manager application is not
required. Only the Network Supervision application is needed for the solution.
Release 18.3
March 2018
CBAM NFV solution
XML API NFV Management

and Orchestration
Service, VNF and

infrastructure
description
NFV-O
NFM-P VE-VNFM-EM
VNFs (CMM, CBAM

CMG SWG,
CMG PWG)
NFV Infrastructure
Virtual Virtual Virtual Virtualized

Computing Storage Network Infrastructure
Manager
Virtualization Layer (OpenStack,
CBIS)
Computing Storage Network

Hardware Hardware Hardware
26388
Supported VNFs
The CBAM as VNF manager solution currently supports the following VNFs:
• CMG
• CMM
• VSR-I
CBAM access point

The NFM-P supports CBAM integration through a CBAM access point created in the Network
Supervision application. You must input CBAM and NFM-P login credentials in order to create the
access point. You must also specify NFM-P discovery rules to enable VNF automatic discovery.
This procedure requires administrator or nfvMgmt access privileges.
Once the access point is successfully created, you can view a list of CBAM access points
associated with the NFM-P. From the list of CBAM access points, you can cross-launch to the
CBAM GUI, rescan CBAM VNFs, or open the Details tab for an access point. The Details tab allows
you to view the associated discovery rule and the connection status for the specified CBAM access
point. The connection status is verified by the NFM-P once every two minutes.
The following are prerequisites before creating a CBAM access point:

• Ensure the specified CBAM login credentials have access to the CBAM APIs via ReST. See the
Installing CloudBand Application Manager Guide for more information.
• Ensure the NFM-P SSL certificates are installed on CBAM. See the CloudBand Application
Manager Administrator Guide for more information.
Release 18.3
March 2018
CBAM NFV solution
• If the CBAM access point uses an IPv6 address, the CBAM must be able to resolve the address
or hostname of the NFM-P server.
See Chapter 5, “Network Supervision web application” for more information about CBAM access
point creation.
VNF discovery
You cannot instantiate VNFs using a CBAM interface. You can only discover VNFs from CBAM
using NFM-P discovery rules specified during access point creation. The NFM-P requires that the
VNFD template for discovered VNFs have post-instantiation scripts to enable SNMP and configure
other protocols necessary for automatic node discovery. When the NFM-P discovers VNFs from a
CBAM access point, it adds them as a rule element for the associated discovery rule.
The Unmanaged VNFs tab lists the VNFs managed by the CBAM instance that were not
discovered by the specified NFM-P discovery rule. The list is updated once every two minutes.
Lifecycle change notifications

Lifecycle change notifications (LCNs) are messages sent from CBAM to the NFM-P with details on
VNF lifecycle updates. When the CBAM access point is created, the NFM-P requests two different
LCN subscriptions for the CMG and CMM. LCNs are used to inform the NFM-P of changes related
to VNF instantiation, termination, scaling, healing, or variable modifications. When the NFM-P
receives an LCN, it scans the VNF information from the CBAM access point and updates its VNF
database accordingly.
Regardless of LCNs, the NFM-P automatically polls the CBAM access point for VNF updates once
every hour.
Note: In a redundant deployment scenario, the CBAM LCN subscription fails after main server
switchover takes place. To resubscribe, you must restart the Network Supervision application
and open the CBAM access point.
VNF lifecycle management

Certain VNF lifecycle changes can be initiated from CBAM or the Network Supervision application.
Whenever a lifecycle change is triggered in CBAM, it informs the NFM-P via an LCN.
VNFs can be instantiated in CBAM and advertised to the NFM-P via an LCN. When the NFM-P
receives information on a newly instantiated VNF, it creates an associated VNF object and attaches
a discovery rule to that object automatically. When the NFM-P discovers a VNF, it retrieves
information related to supported operations, scaling and healing templates, extensions, and
compute resources.
VNFs can be terminated in CBAM and advertised to the NFM-P via an LCN. When the NFM-P
receives information on a terminated VNF, it unmanages the VNF object and removes all
associated VNFCs.
VNFs can be deleted in CBAM and advertised to the NFM-P via an LCN. When the NFM-P receives
information on a deleted VNF, it removes the VNF from its database and unmanages the
associated network element.
Release 18.3
March 2018
CBAM VNFD requirements
VNFs can be healed to trigger a reboot in CBAM or the Network Supervision application. Healing
must be enabled in the CBAM VNFD before this operation can be performed in the GUI. If the
VNFD requires additional parameters for VNF healing, the parameters are visible in the Network
Supervision application.
VNFs can be scaled in or scaled out in CBAM or the Network Supervision application. Scaling must
be enabled in the CBAM VNFD before this operation can be performed in the GUI. When
performing a scaling operation, you must specify a scaling aspect and a new level. The scaling level
cannot exceed the maximum scaling level specified in the CBAM VNFD. If the VNFD requires
additional parameters for VNF scaling, the parameters are visible in the Network Supervision
application.
Note: For the CMM, only CPPS modules can be scaled.
VNF threshold policies

You can assign a threshold policy to a VNF to allow the NFM-P to trigger automatic lifecycle
management operations based on defined KPIs or alarms. A threshold policy allows you to monitor
a set of pre-defined KPIs and create rules to define when the application indicates an overload,
underload, or healing condition. The policy also allows you define an automatic triggered action to
be performed when any of these conditions is met. These corrective actions include performing a
scaling operation, performing a healing operation, or raising an alarm. When a lifecycle
management action is triggered, the NFM-P automatically sends a lifecycle change notification to
CBAM.
You can create a template to define a list of conditions and specify an action to be automatically
performed when those conditions are met. The template can be used to create a VNF threshold
policy, but you can modify the default conditions and actions imported from the template each time
you create a new policy.
See Chapter 5, “Network Supervision web application” for more information about configuring and
using VNF threshold policies in the Network Supervision application. See Chapter 6, “VNF
threshold policy templates” for more information about templates.
Network Supervision application

The primary NFM-P interface for using CBAM as a VNF manager is the Network Supervision
application. See Chapter 5, “Network Supervision web application” for more information.
4.2 CBAM VNFD requirements

Overview
The VNF Descriptor is a package that describes the configuration of the VNF network. It consists of
OpenStack Heat templates which define VNF specifications. This section describes the
configuration requirements for the CBAM VNFD to allow the NFM-P to discover and manage CBAM
VNFs. The VNFD package must be compatible with the release of the CBAM being used to
manage the VNFs.
For more information on CBAM VNFD template creation, see the CBAM documentation suite.
Release 18.3
March 2018
Template requirements
Ensure the CBAM VNFD templates meet the following requirements:
• The template must populate the vnfProductName parameter. This parameter allows the NFM-P
to determine which VNFs it is not currently managing. The value of the parameter must use one
of the following valid product names:
− Cloud VMG
− Cloud MG
− Virtualized Service Router - Integrated
− CMM
− C-SGN
• The VNF instantiation workflow should include the application startup. This ensures that CBAM
sends LCNs to the NFM-P only when the VNF application is up. If the application startup is not
included in the VNF instantiation workflow, the discovery of the VNF into the NFM-P will be
delayed.
• The ansible workflow should push initial configuration on the VNF. This configuration entails the
prerequisite configuration requirements for a network element to be discovered by the NFM-P.
The system interface should be configured based on the value defined in the systemIpAddr
extension. This configuration minimizes the error situation where the actual system interface
IPAddress is different from its definition in the extension.
• The VNFC healing workflow, if implemented, should include the additionalParam vnfcToHeal
configured with a resourceId. The template must not us a UUID as the parameter value, as is
already in use as an OpenStack term. The VNFD should be VIM agnostic and should use only
CBAM-specific information.
• The template must include required parameters that should be pushed to the VNF during
deployment. The following parameters should be available to the NFM-P as VNF extensions or
VNFC resource metadata:
− The systemIpAddr parameter must be available as a VNF extension.
− Each VNFC Resource must contain slotId information that uniquely identifies the card object.
For VNFs that do no support cards, there must be information to uniquely identify the object
that the NFM-P creates.
− The OAM/CPM VNFC must include the mgmtIP. The key for this metadata should be
nokia_vnf_ipAddr. For a CMG, the IP from that vnfcResource will be used for the mgmtIP
with a nokia_vnf_slotId of A.
− Each CMG aspect defined in the template must include an extension that defines the type of
card to which is corresponds. This extension helps the domain make decisions during
preScale.
Release 18.3
March 2018
Release 18.3
March 2018
Network Supervision web application NFM-P
Network Supervision web application
5 Network Supervision web application
5.1 Network Supervision web application

General information
The Network Supervision application is the interface for the CBAM as VNF manager solution. The
application provides a dashboard for assessing the overall health of your network. You can use the
Network Supervision application to monitor groups of VNFs, which are automatically updated as
VNFs are added or removed. Each group can be assessed based on KPIs and affecting alarms.
You can also use the application to assign VNF objects to a special watch list and to perform alarm
management tasks. If you create a CBAM access point, you can use the application to perform
manual and automatic lifecycle management operations,
Click on the application menu and view Start Tour and How To? for more information about features
and workflows in the Network Supervision application. This chapter discusses only the Network
Supervision features that are specific to the NFV solution integrating with CBAM. Refer to the
documentation in the application help menu for more general information on using the application.
Product help tours

You can click on the menu in the application toolbar or on some panel toolbars to view a list of help
tours. These tours are designed to explain the application features and provide workflows for
completing management and monitoring tasks.
Application interface
The Network Supervision application interface includes a watch view, summary view, and group
matrix. A user with administrator privileges must create supervision groups and summary views
using the NFM-P Java UI.
The following figure shows the Network Supervision application with a supervision group selected.
Release 18.3
March 2018
Figure 5-1 Network Supervision UI
You can use the group matrix to drill down to VNF and VNFC lists and associated alarm lists. The
following figure shows the Network Supervision application with an alarm list displayed.
Figure 5-2 Network Supervision alarm list
Release 18.3
March 2018
CBAM access point creation

Before you can perform manual and automatic lifecycle management tasks through the Network
Supervision application, you must configure a CBAM access point. You must input CBAM and
NFM-P login credentials in order to create the access point. You must also specify NFM-P discovery
rules to enable VNF automatic discovery. You can specify one discovery rule for each supported
VNF: CMM, CMG, and VSR-I. This procedure requires administrator or nfvMgmt access privileges.
If the CBAM access point uses an IPv6 address, the CBAM must be able to resolve the address or
hostname of the NFM-P server.
Once the access point is successfully created, you can view a list of CBAM access points
associated with the NFM-P. From the list of CBAM access points, you can cross-launch to the
CBAM GUI, rescan CBAM NFVs, or open the Details tab for an access point. The Details tab allows
you to view the associated discovery rules and the connection status for the specified CBAM
access point. The connection status is verified by the NFM-P once every two minutes.
Figure 5-3 CBAM access point details tab
Release 18.3
March 2018
VNF monitoring and management

The matrix view provides a graphical representation of all the VNFs in a supervision group, or
VNFCs if you have drilled down to the VNF components from a VNF. Each tile represents a VNF or
VNFC. The tile color of a VNF provides an indication of how many VNFCs are currently in an
affected state. Provisioned cards without a VNFC and VNFCs without cards are displayed as grey
tiles.
Note: After a VNF is discovered, VNFCs are temporarily displayed as separate from their
parent VNF. After two minutes, the links between the VNFs are discovered, and the VNFCs
are moved into the appropriate VNF.
You can expand a tile in the matrix view to see more information about the VNF. Depending on the
type of VNF, different KPIs are displayed. For example, the CMM tile displays subscriber count and
capacity utilizations statistics. When there are more KPIs than a tile can contain, expanding a tile
displays a panel instead. These statistics can be used to indicate an overload or underload
condition. These conditions can trigger automatic lifecycle management operations using a VNF
threshold policy.
Figure 5-4 CMM matrix view tile
Release 18.3
March 2018
You can perform a manual scaling operation from the matrix view. Perform a scale-In operation to
reduce the processing capacity of the VNF. Perform a scale-Out operation to increase it. Configure
the scaling aspect and level to specify the type and number of VMs to be scaled. The scaling level
cannot exceed the maximum scaling level specified in the CBAM VNFD. If the templates in the VNF
descriptor specify additional custom properties, you can configure them before executing the
scaling operation.
On the CMM, only the CPPS VM is supported for scaling operations.
Figure 5-5 Manual scaling operation
You can perform a manual healing operation from the drilled-down VNFC view in the matrix view.
Healing is supported for CMG VMs. A healing operation attempts to reboot a failed VNFC by
attempting a soft reboot followed by a hard reboot. Healing must be enabled in the CBAM VNFD
before this operation can be performed. If the templates in the VNF descriptor specify additional
custom properties, you can configure them before executing the scaling operation.
Release 18.3
March 2018
CSGN
A CSGN appears in the Network Supervision application as a supervision group that contains the
VNFs that compose the CSGN. You can discover a CSGN by specifying a discovery rule for each
VNF when you create the CBAM access point associated with the CSGN
VNF threshold policies

You can assign a threshold policy to a VNF to allow the NFM-P to trigger automatic lifecycle
management operations based on defined KPIs or alarms. A threshold policy allows you to monitor
a set of pre-defined KPIs and create rules to define when the application indicates an overload,
underload, or healing condition. The policy also allows you define an automatic triggered action to
be performed when any of these conditions is met. These corrective actions include performing a
scaling operation, performing a healing operation, or raising an alarm. When a lifecycle
management action is triggered, the NFM-P automatically sends a lifecycle change notification to
CBAM. You can assign only one VNF threshold policy to a VNF.
You can assign a CMM or CMG template to define a list of conditions and specify an action to be
automatically performed when those conditions are met. The template can be used to create a VNF
threshold policy, but you can modify the default conditions and actions imported from the template
each time you create a new policy.
Use the policy agent in the Network Supervision application to create a VNF threshold policy. The
policy agent includes default policy templates for the CMM and CMG. After you select a policy
template, you can review the overload, underload, and healing settings to select the actions for
each condition. You can also configure the hold time to specify how long the NFM-P should wait
before performing the specified action.
See Chapter 6, “VNF threshold policy templates” for more information about templates.
Note: Average KPI values are using to trigger scaling operations on the CMG. The average
KPI value is calculated by dividing the monitored KPI value by the number of MG VMs.
Note: The sample templates use the VNFCHealingRequired and LSS_hostReset alarms to
trigger automatic healing. However, any implicitly clearable alarm can be used to trigger an
automatic healing operation as long as it references the affected VNFC. For example, the
alarm could reference a VM number for the CMG or a UUID for the CMM.
Release 18.3
March 2018
Figure 5-6 VNF threshold policy creation
KPI monitoring window

When you select a template, you should define a monitoring window and sampling frequency.
These parameters ensure the quality of the KPI data. The monitoring window specifies the window
of time that the NFM-P monitors the specified KPIs. A monitoring window is a sliding window of
time. The sampling frequency specifies how frequently the KPI data is retrieved within the specified
monitoring window.
Sampling frequency is applicable to the CMG only. The CMM polls the KPI data every two minutes.
Maximum retries
When you create a VNF threshold policy using the Network Supervision application, you can
specify a maximum number of consecutive retries for rules in the policy using the Maximum Retries
parameter. When a LCM request for a rule fails, the Retry Count for that rule is increased; when an
LCM request succeeds, the Retry Count is reset. When the Retry Count reaches the specified
maximum number of retries, the rule is suspended and a major alarm is raised.
Perform the following to resolve the alarm:

1. Investigate the alarm using the NFM-P client or the alarm list in the Network Supervision
application.
2. Investigate the cause of the alarm using the CBAM interface.
3. Reset the Retry Count and clear the alarm by clicking the Show More button on the policy and
selecting Reset Retry Counts. Performing this action will reset the Retry Count for all rules in
the policy and implicitly clear any related alarms.
Note: Before using the Reset Retry Counts action, ensure that the NFM-P has an active LCN
subscription for the associated CBAM access point.
Release 18.3
March 2018
Automatic healing wait time

You can specify a wait time for an automatic healing rule. When an alarm would trigger a VNFC
healing operation, the NFM-P delays the operation for the specified length of time. If the alarm is no
longer present after the wait time has elapsed, the healing operation is not performed. This delay
allows the VNF to execute any recovery mechanisms that would cause the healing action to be
unnecessary.
The wait time is defined under the healing condition rules in the policy template. You can configure
the wait time after choosing the template.
Note: If the NFM-P is not generating VNFCHealingRequired alarms for the CMG, you should
ensure automatic healing is enabled for the VMG in the Java GUI. See 10.2 “To configure
automatic scale-out or automatic healing” (p. 58) for more information.
Applying a VNF threshold policy

Once you have created a VNF threshold policy, you can select Add/Delete Policy to NE from the
VNF submenu of the matrix view tile menu to apply the policy to the VNF. If any of the conditions
defined in the policy have scaling or healing operations specified, you are presented with a step
form when applying the policy to the VNF. You must configure the Aspect Name parameter for each
overload and underload threshold rule. The Aspect Name parameter specifies aspect type to be
scaled when the scaling operation is triggered. You can click the + button to add custom properties
for overload, underload, and healing threshold rules.
When applying a VNF threshold policy to a CMG, you can choose to add a custom MIB statistics
policy to the VNF. This option is enabled by default and ensures that statistics collection for the
statistics associated with the KPIs in the VNF threshold policy is enabled. If you choose not to add
a custom MIB statistics policy, you should use the NFM-P Java GUI to ensure there is a MIB
statistics policy associated with the VNF. The policy must enable statistics collection for the
statistics associated with the KPIs in the VNF threshold policy. See “Condition syntax” (p. 35) for a
table that shows the mapping between VNF KPIs and NFM-P statistics.
Release 18.3
March 2018
Figure 5-7 VNF threshold policy application
Note: When applying a VNF threshold policy with only alarm actions (as opposed to healing
or scaling actions) to a CMM, no policy application configuration steps are required.
When applying a VNF threshold policy with only alarm actions (as opposed to healing or
scaling actions) to a CMG with no associated MIB statistics policy, no policy application
configuration steps are required.
Monitoring a VNF threshold policy

When a policy has been applied to a VNF, the policy monitoring icon appears on the VNF tile in the
matrix view. You can click on the policy monitoring icon to view the status of each of the overload,
underload, and healing threshold rules specified in the policy. The policy info view shows the
conditions, thresholds, and actions for each rule. It also shows the current value of the condition,
allowing you to assess how close the condition is to reaching the specified threshold.
Release 18.3
March 2018
Figure 5-8 VNF threshold policy status
When a threshold is crossed and the VNF threshold policy triggers a scaling or healing operation,
the policy monitoring icon changes to red and the operation is shown in-progress on the VNF tile.
While the operation is being performed, you can disable the policy to cancel the action.
Figure 5-9 VNF threshold policy tiles
Release 18.3
March 2018
VNF threshold policy templates NFM-P
Policy template design
6 VNF threshold policy templates
6.1 Policy template design

Overview
VNF threshold policies specify the KPIs used to indicate whether a VNF or VNFC is in an overload,
underload, or healing condition. The policy can be used to automatically trigger lifecycle
management operations such as scaling or healing. To create a VNF threshold policy, you must
specify a template that details the default KPIs and actions for each condition. The NFM-P includes
sample templates for the CMM and CMG, but you can also create a new template. This chapter
describes the format and syntax requirements for creating a new VNF threshold policy template.
YAML
OpenStack templates are written using the YAML markup language. This chapter describes only the
formatting and syntax requirements to create a .yaml file for VNF threshold policy templates. For
more information about YAML, see the CBAM or OpenStack documentation suites.
Workflow to create a VNF threshold policy template

The following workflow describes the steps to create a VNF threshold policy template. See
“Template format” (p. 34) for information about the overall template format and keyword syntax.
1
Define the template name, NE type, monitoring window, maximum retries, and sampling
frequency at the top of the .yaml file. See “Template format” (p. 34) and “Template keywords”
(p. 34).
Note: At least one of the following three steps is required for template creation.
If required, create an Overload_Condition_Criteria section and define rules elements. See
“Rules” (p. 35) and “Rule keywords” (p. 35).
3
If required, create an Underload_Condition_Criteria section and define rules elements. See
“Rules” (p. 35) and “Rule keywords” (p. 35).
4
If required, create a Healing_Condition_Criteria section and define rules elements. See “Rules”
(p. 35) and “Rule keywords” (p. 35).
Release 18.3
March 2018
5
Review the template for any syntax errors.
6
Format the template to ensure the Network Supervision application user can read and
understand contents clearly.
7
Save the .yaml file with a filename that indicates the purpose of the template.
8
Upload the template to the template directory in the NSP NFM-P file system. See “Template
directory” (p. 38).
Template format
The policy template must follow the format below. The template can exclude any of the sections for
overload, underload, or healing condition criteria, but it must include at least one of them. The
instances of <string>, <integer>, and <rule #> should be replaced with values to be
determined by the user.
-
Name: <string>
NE_Type: <string>
Monitoring_Window: <integer>
Sampling_Frequency: <integer>
Overload_Condition_Criteria:
Rules:
- <rule 1>
- <rule 2>
- <rule n>
Underload_Condition_Criteria:
Rules:
- <rule 1>
- <rule 2>
- <rule n>
Healing_Condition_Criteria:
Rules:
- <rule 1>
- <rule 2>
- <rule n>
Template keywords
You must specify values for the following keywords at the top of the template:
• Name — identifies the policy as it will appear in the Network Supervision application policy agent
Release 18.3
March 2018
• NE_Type — specifies whether the template is for the CMM or CMG; specify “cmm” or “cmg”
• Monitoring_Window — specifies the window of time that the NFM-P monitors the specified
KPIs; specify a time between 1 and 120 minutes
• Sampling_Frequency — specifies how frequently the KPI data is retrieved within the specified
monitoring window; specify 1, 5, 10, 15, 30, 45, or 60 minutes
Note: Sampling_Frequency is not required for the CMM. By default, the sampling window
for the CMM is 2.
Rules
If the policy template includes a section for overload, underload, or healing criteria, it must include
at least one rule in the section. Rules must follow the format below. The instances of variables such
as <string> should be replaced with values to be determined by the user.
—
- Name : <string>
Condition : (<KPI_Value> <operator> $<Threshold_Value>)
Action : <action>
Hold_Time: <integer>
Values : {$<Threshold_Value>: '<integer>'}
—
Note: $ is a prefix used to identify variables.
Rule keywords
You must specify values for the following keywords within a template rule:
• Name — identifies the rule name as it will appear in the Network Supervision application
• Condition — specifies a logical statement including a KPI value, operator, and threshold value
• Action — specifies the action to perform if the condition is satisfied; specify “scaleOut”,
“scaleIn”, “heal”, or “alarm”
• Hold_Time — specifies the time, in minutes, before waiting to trigger the action again if the
condition is still satisfied
• Values — defines values for variables used in the logical conditions
Condition syntax
The Condition line is a logical statement that needs to be satisfied to perform the specified action. It
includes at least one KPI value, logical operator, and threshold value. Multiple conditions can be
grouped using logical operators. You must include a space between each KPI, operator, and
variable.
The threshold value is a user-defined variable used to compare with the specified KPI value. The
threshold value name has no syntax requirements, but it is recommended that you choose a name
that indicates the KPI and type of threshold. For example, a threshold value that defines the
maximum value for the KPI value CMG_NE_KPI_Bearer_Count should be defined as CMG_NE_
Release 18.3
March 2018
KPI_Bearer_Count_Max. Following this convention allows you to more easily compare the
configured threshold value with the current value during VNF policy creation.
The following table lists the KPIs that can be used as part of a rule condition.
Table 6-1 VNF condition KPIs
KPI value NFM-P class name Property MIB name
CMG NE level KPIs
CMG_NE_KPI_Subscriber_Count isa.PdnGwCardStats ues TIMETRA-MOBILE-PDN-MIB.

tmnxMobPdnStatEntry
CMG_NE_KPI_Session_Count isa.PdnGwCardStats sessions TIMETRA-MOBILE-PDN-MIB.

tmnxMobPdnStatEntry
CMG_NE_KPI_Bearer_Count isa.PdnGwCardStats bearers TIMETRA-MOBILE-PDN-MIB.

tmnxMobPdnStatEntry
CMG_NE_KPI_Combined_Session_Count isa.PdnGwCardStats combinedP- TIMETRA-MOBILE-PDN-MIB.

dnSessions tmnxMobPdnStatEntry
CMG_NE_KPI_Combined_Bearer_Count isa.PdnGwCardStats combinedBear- TIMETRA-MOBILE-PDN-MIB.

ers tmnxMobPdnStatEntry
CMG alarm KPI 1
CMG_NE_ALARM_Number_Of_Occurrences_ — — —
<alarm name>
For example, CMG_NE_ALARM_Number_Of_
Occurrences_VNFCHealingRequired
CMG_NE_ALARM_Last_Time_Detected_ — — —
<alarm name>
CMG VM level KPIs
MG_VM_KPI_CPU_Utilization equipment. busyCoreUtil TIMETRA-SYSTEM-MIB.

CpuUtilizationStats tmnxCardCpuResMonitorEntry
MG_VM_KPI_Subscriber_Count isa.PdnGwCardStats ues TIMETRA-MOBILE-PDN-MIB.

tmnxMobPdnStatEntry
MG_VM_KPI_Session_Count isa.PdnGwCardStats sessions TIMETRA-MOBILE-PDN-MIB.

tmnxMobPdnStatEntry
MG_VM_KPI_Bearer_Count isa.PdnGwCardStats bearers TIMETRA-MOBILE-PDN-MIB.

tmnxMobPdnStatEntry
MG_VM_KPI_Combined_Session_Count isa.PdnGwCardStats combinedP- TIMETRA-MOBILE-PDN-MIB.

dnSessions tmnxMobPdnStatEntry
MG_VM_KPI_Combined_Bearer_Count isa.PdnGwCardStats combinedBear- TIMETRA-MOBILE-PDN-MIB.

ers tmnxMobPdnStatEntry
CMG LB VM level KPI
LB_VM_KPI_CPU_Utilization equipment. busyCoreUtil TIMETRA-SYSTEM-MIB.

CpuUtilizationStats tmnxCardCpuResMonitorEntry
CMG card count KPIs
Release 18.3
March 2018
Table 6-1 VNF condition KPIs (continued)
KPI value NFM-P class name Property MIB name
MG_VM_Count — — —
Card type: card_iom_mg_vsr
LB_VM_Count — — —
Card type: card_iom_vsr
CMM NE level KPIs
CMM_NE_KPI_UE_Capacity ltecmm. totalNumUes —

CMMInstanceCapac-
ity
CMM_NE_KPI_Percent_Utilization ltecmm. percentUtiliza- —

CMMInstanceCapac- tion
ity
CMM alarm KPI 1
CMM_NE_ALARM_Number_Of_Occurrences_ — — —
<alarm name>
For example, CMM_NE_ALARM_Number_Of_
Occurrences_CmmLSS_hostReset
CMM_NE_ALARM_Last_Time_Detected_ — — —
<alarm name>
CMM CPPS VM level KPIs
CPPS_VM_KPI_UE_Capacity ltecmm. uesInCpps —

CppsVNFCMember-
Capacity
CPPS_VM_KPI_Percent_Utilization ltecmm. percentUtiliza- —

CppsVNFCMember- tion
Capacity
Notes:
1. In rule conditions, alarms must always be checked against the number of occurrences. For example,
CMM_NE_ALARM_CmmLSS_hostReset > $No_of_Occurrence.
The following table lists the operators that can be used in rule conditions.
Operator Usage Operator Usage

$ Prefix used to identify % modulus
variables
&& and == equal to
|| or != not equal to
! not > greater than
+ addition < less than
Release 18.3
March 2018
Policy template sample
Operator Usage Operator Usage

- subtraction >= greater than or equal
to
* multiplication <= less than or equal to
/ division
Note: The Network Supervision threshold policy framework may trigger multiple concurrent
actions for the same policy. You should ensure that rule definitions in the threshold policy
result in logical actions for each condition.
Template directory
Templates must be saved in the /opt/nsp/nfmp/KPITCA/TemplateLibrary directory. The template

directory includes the following sample templates:
• CMM_Rule_Template_Example_n.yaml
• CMG_Rule_Template_Example_n.yaml
• Sample_CMM_Rules.yaml
• Sample_CMG_Rules.yaml
Note: The Reference_Sample files are for reference only. Only the Template_Example files
can be used.
The NFM-P supports redundancy for template files. Whenever an rsync operation is triggered, files
in the /opt/nsp/nfmp/KPITCA/tca_rsync directory with a yaml, .sh, .txt., and .json. extension are
duplicated into a standby file server. Changes in this directory on the active server are duplicated
every 30 minutes by default.
6.2 Policy template sample

Overview
The policy template sample in section demonstrates the format and syntax required to create a
valid policy template.
Sample
Name: Sample CMG Rules
NE_Type: cmg
Monitoring_Window: 15
Sampling_Frequency: 5
Overload_Condition_Criteria:
Rules:
- Name : VNF Bearer Overload
Condition : ( CMG_NE_KPI_Bearer_Count > $CMG_NE_KPI_Bearer_Count_Max )
Release 18.3
March 2018
Action : scaleOut
Hold_Time: 15
Values : {$CMG_NE_KPI_Bearer_Count_Max: '200'}
- Name : VNFC(MG) Sesion Overload

Condition : ( MG_VM_KPI_Combined_Session_Count > $MG_VM_KPI_Combined_Session_
Count_Max )
Action : scaleOut
Hold_Time: 15
Values : {$MG_VM_KPI_Combined_Session_Count_Max: '200'}
- Name : VNF Subscriber + VNFC(LB) CPU Utilization Overload

Condition : (( CMG_NE_KPI_Subscriber_Count > $CMG_NE_KPI_Subscriber_Count) ||
(LB_VM_KPI_CPU_Utilization > $LB_VM_KPI_CPU_Utilization))

Action : scaleOut
Hold_Time: 15
Values : {$CMG_NE_KPI_Subscriber_Count: '300',$LB_VM_KPI_CPU_Utilization:
'80'}
Underload_Condition_Criteria:
Rules:
- Name : VNF Bearer Underload
Condition : ( CMG_NE_KPI_Bearer_Count > 0 && CMG_NE_KPI_Bearer_Count <
$CMG_NE_KPI_Bearer_Count_Min )
Action : scaleIn
Hold_Time: 15
Values : {$CMG_NE_KPI_Bearer_Count_Min: '100'}
- Name : VNFC(MG) Sesion Underload

Condition : ( MG_VM_KPI_Combined_Session_Count > 0 &&
MG_VM_KPI_Combined_Session_Count < $MG_VM_KPI_Combined_Session_
Count_Min )
Action : scaleIn
Hold_Time: 15
Values : {$MG_VM_KPI_Combined_Session_Count_Min: '50'}
- Name : VNF Subscriber + VNFC(LB) CPU Utilization Underload

Condition : ((CMG_NE_KPI_Subscriber_Count >0 && CMG_NE_KPI_Subscriber_Count <
$CMG_NE_KPI_Subscriber_Count) || (LB_VM_KPI_CPU_Utilization > 0 && LB_VM_KPI_CPU_
Utilization < $LB_VM_KPI_CPU_Utilization))
Action : scaleIn
Hold_Time: 15
Values : {$CMG_NE_KPI_Subscriber_Count: '100',$LB_VM_KPI_CPU_Utilization:
'20'}
Healing_Condition_Criteria:
Rules:
- Name : VNFCDown Heal Alarm
Release 18.3
March 2018
Condition : ((CMG_NE_ALARM_Number_Of_Occurrences_VNFCHealingRequired >

$No_Of_Occurrence) &&
(CMG_NE_ALARM_Last_Time_Detected_VNFCHealingRequired +
$Wait_Time < CURRENT_TIME))
Action : heal
Hold_Time: 15
Values : {$No_Of_Occurrence: '1',$Wait_Time: '5'}
Release 18.3
March 2018
CBAM GNE driver information NFM-P
CloudBand Application Manager driver version 1.0.0
7 CBAM GNE driver information
7.1 CloudBand Application Manager driver version 1.0.0

Driver capabilities
The following table lists CloudBand Application Manager driver version 1.0.0 capabilities. The driver
extends the management capabilities of the NFM-P for CBAM in the following applicable areas.
Configuration management
GNE profile automation: The GNE profile for the CloudBand Application Manager is
automatically created when the driver is installed.
Alarm catalog integration:
Not Applicable
Service management
Not applicable
Tunnel management
Not applicable
Network assurance
Not applicable
Service assurance
Not applicable
Fault management
Not applicable.
Closed issues
There are no closed issues to report.
Outstanding issues
There are no outstanding issues to report.
Release 18.3
March 2018
CBAM GNE driver information NFM-P
Discovery and management
7.2 Discovery and management

CloudBand Application Manager post-installation configuration instructions
The following instructions provide information specific to CBAM server configuration and discovery.
The NSP NFM-P User Guide chapter "Device commissioning and management" should be
consulted for full procedural details.
Create mediation policies and configure a discovery rule

1. Use the NFM-P client to create an SNMP v2 mediation policy for CBAM management. See the
NSP NFM-P User Guide for information about creating mediation policies.
2. Use the NFM-P to configure a discovery rule for the CloudBand Application Manager that
includes the SNMP v2 mediation policy; see the NSP NFM-P User Guide for information about
creating discovery rules.
Release 18.3
March 2018
NFM-P as VNF manager solution NFM-P
Part III: NFM-P as VNF manager solution
Overview
Purpose
This part provides information on the NFM-P as VNF manager solution.
Contents
Chapter 8, NFM-P as VNF manager 45

Chapter 9, VNF Manager web application 51
Chapter 10, NFV use cases 57
Chapter 11, NFM-P VNF descriptor 61
Release 18.3
March 2018
NFM-P as VNF manager solution NFM-P
Release 18.3
March 2018
NFM-P as VNF manager NFM-P
NFM-P NFV solution
8 NFM-P as VNF manager
8.1 NFM-P NFV solution

Overview
The NFM-P can be used as the VNF manager for the NFV solution by using the VNF Manager
application. The NFM-P provides an interface with OpenStack to enhance VNF management with
more monitoring, assurance, and management functionality. This VNF-aware network management
system (NMS) extends NFM-P existing network element management and assurance to VNFs and
VNF lifecycle functions. The NFM-P NFV feature set is composed of two applications (VNF
Manager and Network Supervision) as well as support for network function element management
system (EMS) in the NFM-P Java GUI.
Figure 8-1 NFM-P as VNF manager solution
XML VNF
API Configuration
NFM-P NFV Solution

VNF
catalogue
REST
API NBI NFV-O
Network NFM-P VNF Manager

Supervision EMS Application
Application
API
VNF (CMG,
CMM, VSR)
VM
OpenStack
25305
OpenStack
OpenStack is an open-source cloud management system (CMS) that can be used for NFV
management. OpenStack provides the NFV infrastructure and orchestration components that can
be used to perform lifecycle management tasks on managed VNFs and other virtualized network
elements.
Release 18.3
March 2018
NFM-P NFV solution
The NFM-P provides an interface with OpenStack through the VNF Manager application.
Orchestration
OpenStack orchestration allows the user to manage the lifecycle of VNFs and network
infrastructure within the OpenStack cloud. The module used for orchestration is called Heat. The
Heat orchestration engine is designed to start cloud applications based on template files that are
specialized for different types of VNFs. These Heat orchestration template (HOT) files describe a
list of OpenStack resources called a stack. Heat maps these resources to virtual machines (VMs)
based on VNF requirements.
HOT files can be used to define the following VNF specifications:

• resource allocation
• lifecycle tasks, including deployment, rebooting, and scaling
• auto-scaling instructions
Cloud Management
In addition to Heat, OpenStack uses several other modules for VNF management. OpenStack
cloud management programs include the following:
• Nova—used for compute resources
• Glance—discovery and provisioning for disk and server images
• Neutron—OpenStack networking
• Cinder—block-level storage
• Ceilometer—billing and telemetry
• Keystone—authentication
VNF application cross-launch

The NFM-P Java GUI supports cross-launching the VNF Manager and Network Supervision
applications from the VNF or VNFC in the equipment tree. Cross-launching to the VNF Manager
application allows you to perform manual lifecycle management tasks such as VNF instantiation or
scaling. Cross-launching the Network Supervision application allows you to view the selected VNF
or VNFC in the context of its supervision group. To cross-launch a VNF application from the
equipment tree, right-click a network element or card slot and choose VNF Management to select
an application to launch.
Licensing
The NFM-P allows you to view NFM-P NFV license statuses and counters in the Java GUI. The
license status is listed as either valid, invalid, no license, or locked. You can view license counters
to see how many NFM-P VNF and VNFC licenses of each type have been consumed. The NFM-P
raises a node license alarm when the license is invalid and includes a UUID.
Note: After the NFM-P discovers the number of CMG VNFs allowed by the license, the
NFM-P can no longer discover the CMG.
Release 18.3
March 2018
Supported VNFs
VNF assurance and monitoring

The NFM-P improves visibility of faults in the VNF network with alarm management. You can use
the Network Supervision application to view all alarms in a VNF supervision group, and have the
ability to drill down to a specific VNF or VNFC. The NFM-P reduces alarm volume and quickens
fault resolution by correlating alarms between the VNF application and the virtual infrastructure. You
can also view alarms for VNFs in the NFM-P Java GUI, which gives you access to more alarm
management functionality, such as threshold-crossing alarms (TCAs).
The Network Supervision application provides a simplified dashboard view of VNFs in the network.
The application allows you to track key performance indicators (KPIs) such as down VNFCs, to
determine the overall health trend of your network. You can quickly assess network performance,
outages, and VNF availability using the Watch List or Group Matrix. You can use this information to
determine if a capacity change (scale-in or scale-out) is required.
You can configure the NFM-P to perform an automatic scale-out and automatic healing based on
capacity thresholds or network reachability events.
VNF lifecycle management
The NFM-P provides an interface with OpenStack Heat to allow you to perform VNF lifecycle tasks
from the VNF Manager application. You can perform the following lifecycle tasks directly from within
the VNF Manager application:
• Instantiation—create a VNF instance from a specified cloud access point and a VNF catalog
• Deletion—delete a VNF instance
• Deployment—deploy a VNF instance to the cloud network
• Scaling—reduce or expand processing capacity by adding VMs to a VNF
• Healing—reboot a failing VNF component
• Sync—synchronize a VNF with the OpenStack tenant
Note: Automatic scale-in is not supported.
VNF Manager application

The primary NFM-P interface for this solution is the VNF Manager application. See 9.1 “VNF
Manager web application” (p. 51) for more information.
8.2 Supported VNFs

Overview
The NFM-P extends basic EMS support to virtual network elements. Discovery, provisioning, and
management workflows are the same as for physical nodes.
The NFM-P supports the following VNFs:

• CMG
• VMM
Release 18.3
March 2018
Supported VNFs
• VSR
See the NSP NFM-P User Guide for more information about NE discovery and management.
CMG
The CMG supports the following virtualized applications:

• Packet Data Network Gateway (VMG-PGW)
• Gateway GPRS Support Node (VMG-GGSN)
• Serving Gateway (VMG-SGW)
• Evolved Packet Data Gateway (VMG-ePDG)
These virtualized applications can be deployed as separate network functions or in combination (for
example, VMG-SGW/PGW/GGSN or VMG-SGW/ePDG).
The CMG supports multiple mobile gateway functions including PGW, GGSN, SGW, and the
combined SGW/PGW/GGSN. It also supports LTE profiles and policies, golden configuration, LTE
reference points, EPS peers and paths, and statistics. These functions operate the same as they do
on a physical 7750 MG.
The CMG consists of VMs operating as VNFCs. Each VM is dedicated to a specific set of functions
that are replicated across other VMs. A group of VMs are represented as a single instance of an
application. The VMs in the group operate in synchronization to support a network functionality that
can scale horizontally as required.
The CMG is composed of the following VNFCs:

• operations, administration, and management VM (OAM-VM)—performs control plane functions
including routing protocols, management interface functions such as CLI configuration, and VNF/
VNFC management
• load balancer VM (LB-VM)—provides network connectivity to mobile gateway functions, load
distribution across the MG-VMs, and forwarding of GTP-C/GTP-U and UE-addressed packets to
the MG-VM
• mobile gateway VM (MG-VM)—manages services including 3GPP call processing (control and
data plane), PCEF, and application assurance (PCEF-enhanced with ADC for application
detection and control, and L7 service classification for policy charging control
You can configure automatic scale-out or automatic healing on the CMG. If automatic healing is
enabled, the NFM-P will attempt to reboot a VNFC when it goes down. The NFM-P attempts a soft
reboot first before attempting a hard reboot, if necessary.
See Chapter 10, “NFV use cases” for more information about automatic scale-out and automatic
healing.
VMM
You can enable automatic scale-out on the VMM and configure an automatic scale-out threshold.
The automatic scale-out threshold defines the point at which an automatic scale-out operation can
be triggered, where the threshold is the total UE capacity of the VMM multiplied by the scale-out
factor. When the threshold is reached and/or one or more MAFs is affected by a resource overload
Release 18.3
March 2018
Supported VNFs
node alarm, the NFM-P automatically increases the processing capacity of the VMM by creating an
additional virtual MAF on the NE.
The Perform Scale-out (VMM) parameter defines the conditions under which the automatic scale-
out is triggered. You can specify AND or OR for the following conditions:
• The number of UEs on the WMM is over capacity.
• At least one MAF cards has a threshold crossing alarm against the number of UEs.
The Auto Scale-out Timer parameter prevents the NFM-P from triggering additional scale-out
operations while a previous scale-out is still in progress. You should specify a length of time that
exceeds the expected time for a virtual MAF to become operational.
You can also enable automatic healing on the VMM. If automatic healing is enabled, the NFM-P will
attempt to reboot a VNFC when it goes down. The NFM-P attempts a soft reboot first before
attempting a hard reboot, if necessary.
See Chapter 10, “NFV use cases” for more information about automatic scale-out and healing.
VSR
The Virtual Service Router (VSR) is a software-only version of the 7750 SR. VSR and VSR-I
chassis types are supported.
The VSR consists of VMs operating as VNFCs. Each VM is dedicated to a specific set of functions
that are replicated across other VMs. A group of VMs are represented as a single instance of an
application. The VMs in the group operate in synchronization to support a network functionality that
can scale horizontally as required.
The VSR is composed of the following VNFCs:

• IOM-V
• CPM-V
On the VSR-I, the IOM-V and CPM-V are deployed on a single VM.
Release 18.3
March 2018
Supported VNFs
Release 18.3
March 2018
VNF Manager web application NFM-P
VNF Manager web application
9 VNF Manager web application
9.1 VNF Manager web application

Overview
The VNF Manager application is the interface used for the NFM-P as VNF manager solution. The
application allows you to instantiate, maintain, and terminate VNFs managed by the NFM-P. The
NFM-P provides an interface with the cloud management entity that provisions cloud resources.
The NFM-P validates these resources and provides assurance and monitoring through the VNF
Manager application and other features in the NFV solution.
Product help tours

You can click on the menu in the application toolbar or on some panel toolbars to view a list of help
tours. These tours are designed to explain the application features and provide workflows for
completing management and monitoring tasks.
Application interface
The VNF Manager application interface is divided into three panels for managing catalogs, cloud
access points, and VNFs.
Release 18.3
March 2018
Figure 9-1 VNF Manager application GUI
VNF catalogs
You can use the VNF Manager application to create and manage VNF catalogs. A VNF catalog is a
collection of VNFs with a template that determines the type of VNF managed. The catalog includes
deployment specifications, KPIs, and recipes for VNF management functions that are applicable to
the VNF type. These become the default VNF settings that are defined when you instantiate a VNF
using the catalog. You are able to configure these settings for specific VNFs during VNF
instantiation.
You can create the following types of VNF catalogs:

• Generic
• VMG
• VMM
• VSR
• VSR-I
In addition to specifying the catalog type, you must also define a directory name. The directory
name specifies the VNFD directory where the HOT files are located. The VNF Manager
automatically determines the catalog type and version based on the HOT files in the specified
directory. You must configure a VNF catalog before you create a VNF object. See Chapter 11,
“NFM-P VNF descriptor” for more information about template creation.
Release 18.3
March 2018
Generic VNF catalogs

While the VMG, VMM, and VSR catalog types include deployment settings and recipes specific to
each NE type, the generic VNF catalog can be used for other NE types without node-specific
settings. The onboarding, instantiation, and lifecycle management functions of generic VNF
catalogs are the same as for the NE-specific catalogs. You must specify a VNFD directory and
configure HOT template files for a generic VNF catalog.
Cloud access point

The VNF Manager application provides an interface with OpenStack for VNF management. You
must define a cloud access point before you create a VNF object. You can use multiple cloud
access points for VNFs under different tenants. For each cloud access point, you must provide login
credentials for the tenant group and specify a Keystone URL. Specify an access URL, port number,
and Keystone version in the format http://127.227.135.61:5000/v2.0.
The cloud access point defines parameters such as VNFC image/flavor and network/subnet ID,
which are configurable during VNF instantiation.
VNF management
The VNF Manager application provides an at-a-glance view of VNFs and VNFCs in the network.
VNF onboarding allows the application to archive, upload, and validate VNF software images with
OpenStack Heat.
Figure 9-2 VNF Management
Release 18.3
March 2018
The following table describes the lifecycle management tasks that can be executed from the VNF
Manager application.
Table 9-1 VNF lifecycle management tasks
Task Description
Add Instantiates a new VNF. You must choose a catalog and cloud
access point when you create a new VNF.
Delete Deletes a VNF.
Deploy Deploys the VNF to OpenStack Heat. Newly instantiated VNFs
are not automatically deployed. The Deployment State
parameter shows whether the VNF has been deployed.
Scale-out Increases processing capacity by adding a VM. You must
specify the scaling template from the drop-down menu. The
scale-out template defines the type and number of VNFCs to be
added.
Scale-in Decreases processing capacity by removing a VM. You must
specify the scaling template from the drop-down menu. The
scale-in template defines the type and number of VNFCs to be
removed.
Sync Synchronizes the VNF with OpenStack Heat.
Rescan Discovery Rule Manually scan VNF discovery rules for new rule elements.
Reboot Reboots the VNF component. The Reboot button is available
from the VNF component list that opens when you select an
VNF.
Manual scaling
CAUTION
Service Disruption
The manual scale-in function removes a VM without checking the node for subscribers. Nokia
recommends you check the VM for subscribers and move them to another VM before using the
scale-in function.
You can use the VNF Manager application to manually scale-in or scale-out to adjust the
processing capacity of a VNF. When you use a scaling operation, you must select a scaling
template to use. Each scaling template specifies a type and number of VNFCs to be scaled in or
out. The scale-in and scale-out operations use the same scaling templates.
These scaling templates are defined by the VNF_Type.userdef_chars.grow.hot.yaml files in the
grow sub-directory of the VNF catalog directory. The name of the scaling templates must specify
the applicable VNF. A VMG scaling template must lead with VMG in the template name and a VMM
Release 18.3
March 2018
scaling template must lead with VMM. For example, a scaling template designed to scale one load
balancing VNF on a CMG might be named CMG.1LB.grow.hot.yaml. Nokia recommends that the
template name specify the number and type of VNFCs to be scaled.
Note: The userdef_chars part of the template filename cannot contain a period.
For automatic scaling, the NFM-P looks for a file with the name VNF_Type.default.grow.hot.yaml.
You can include a grow_meta_data entry in the meta file to specify which parameters are user-
configurable during a VNF scaling operation in the VNF Manager application. For example, you can
include gw_subnet in the grow_meta_data entry to allow the user to specify a subnet during VNF
scale-out. See “VNF scaling parameterization” (p. 72).
VNF instantiation
The VNF Manager application uses VNF catalogs to instantiate VNFs on a cloud management
system such as OpenStack. When you use the application to instantiate a VNF, you must specify a
cloud access point which includes a cloud management system URL to which the NFM-P deploys
the VNF. The VNF catalog selected during VNF instantiation defines the default settings that are
required by the VNFD for lifecycle management operations. These settings are read from the
catalog .env.yaml file. You can customize these settings for a specific VNF during instantiation, as
required.
The cloud access point provides the list of available VNFC images and flavors. You can select the
required images and flavors from drop-down menus during VNF instantiation.
You can also configure the System Address during VNF instantiation. These parameters uniquely
identify the VNF. Alarms affecting the VNF or associated VNFCs list these parameters as System
Name and Site ID.
Note: The instantiated VNF is not deployed until you click Deploy.
Note: The system address provided for the VNF should match the system address configured
in the NE.
Initial configuration for CMG and VSR

During VNF instantiation, you can specify an initial configuration file for the CMG and VSR. Specify
an FTP server address with the configuration file using the primary_config parameter. You must
also specify the login credentials, system name, and SNMPv2 or SNMPv3 community string. The
configuration file is uploaded to the file storage specified for the VNF instance. When the VNF is
instantiated, it is automatically configured with the specifications in the initial configuration file.
Topology discovery rules

During VNF instantiation, you can specify a topology discovery rule for VSR, CMG, or VMM. If no
discovery rules are defined, you can cross-launch the NFM-P Java GUI to create one. The NFM-P
automatically creates a discovery rule element for the VNF once the VNF is successfully deployed.
You can also manually trigger a discovery rule scan by clicking the Rescan Discovery Rule button.
The ready-only Managed State identifies whether the VNF is managed by a discovery rule.
When the VNF is deleted, the NFM-P automatically removes the discovery rule element.
Release 18.3
March 2018
VNF component reboot

You can use the Reboot function to manually reboot a VNF component from the application GUI. To
view a list of VNF components associated with a VNF, double-click the VNF. When you click the
Reboot button, you can select a soft or hard reboot.
The read-only Task State parameter on the VNF component shows whether the component has any
NFM-P component operations in progress. VNF component can be triggered only if the Task State
parameter is Deployed. The read-only VM State parameter displays the OpenStack VM state. VNF
component operations — such as reboot — can be performed only if the VM State parameter is
Active, ShutOff, or Rescued.
VNF component evacuation

You can perform VNF component evacuation to move a VNF component to a new compute host.
This functionality is useful if a VNF component goes down due to a compute host failure. This
action does not reboot the VNF.
VNF component evacuation is available from the VNF Manager REST API only.
VNFD redundancy
The NFM-P supports redundancy for the HOT files that define the lifecycle management tasks
performed by the VNF Manager application. These files are kept in the <NFMP_INSTALL_DIR>/os
directory and require NFM-P administrator read/write privileges. Whenever an rsync operation is
triggered, files in the directory with a yaml, .txt, .sh, and .json extension are duplicated into a
standby file server. Changes in the <NFMP_INSTALL_DIR>/os directory on the active server are
duplicated every 30 minutes by default.
REST API
The VNF Manager application publishes a set of URLs which point to resources, or web services,
managed by them. The URLs that are available to users are documented. These URLs can be
accessed through a browser by any authorized user, including OSSs which can use them to cross
launch from their own application. To view the published URLs of a given application:
http(s)://<host>/VNFManager/api-docs
Where host is the hostname or IP address which hosts the application.
Release 18.3
March 2018
NFV use cases NFM-P
Automatic scale-out and healing
10 NFV use cases
10.1 Automatic scale-out and healing

Overview
You can configure the NFM-P to perform automatic scale-out and healing operations based on
system-defined and user-defined processing thresholds and network events.
A scale-out operation increases the call processing capacity of a VNF. Processing capacity is
increased by automatically creating and provisioning a VNFC. A healing operation reboots a VNFC.
Automatic healing is triggered by the NFM-P when an alarm is raised against a VNFC. The NFM-P
performs additional tasks based on the results of the attempted healing operation. You can enable
automatic scale-out and automatic healing for the VMM or CMG. This configuration must be
performed by a system administrator.
VMM automatic scale-out

You can enable automatic scale-out on the VMM and configure an automatic scale-out threshold.
The automatic scale-out threshold defines the point at which an automatic scale-out operation can
be triggered, where the threshold is the total UE capacity of the VMM multiplied by the scale-out
factor. When the threshold is reached and/or one or more MAFs is affected by a resource overload
node alarm, the NFM-P automatically increases the processing capacity of the VMM by creating an
additional virtual MAF on the NE.
The Perform Scale-out (VMM) parameter defines the conditions under which the automatic scale-
out is triggered. You can specify AND or OR for the following conditions:
• The number of UEs on the VMM is over capacity.
The Auto Scale-out Timer parameter prevents the NFM-P from triggering additional scale-out
operations while a previous scale-out is still in progress. You should specify a length of time that
exceeds the expected time for a virtual MAF to become operational.
CMG automatic scale-out

Automatic scale-out on the CMG is triggered by user-defined thresholds. You can specify low and
high threshold bearer limits by configuring a threshold group on the PDN gateway. When the bearer
management threshold limit is reached, a threshold-crossing alarm is raised. This alarm triggers an
automatic scale-out. The NFM-P automatically increases the processing capacity of the CMG by
creating an additional auto-provisioned ISM MG Card on the NE.
CMG automatic healing

Automatic healing is triggered when an LSS_hostReset or VNFC Down alarm is raised on a VNFC
and the VM state is either Active, Shut Off, or Rescued. The NFM-P first attempts a soft reboot. If
the alarm is not cleared, the NFM-P then attempts a hard reboot. If the alarm is still not cleared, the
NFM-P raises a VNFCAutoHealFailed alarm on the VNFC.
Release 18.3
March 2018
NFV use cases NFM-P
To configure automatic scale-out or automatic healing
The Auto Heal Timer specifies the length of time the NFM-P waits between reboot attempts before
proceeding with the next action.
VMM automatic healing

Automatic healing is triggered on the VMM when an LSS_hostReset alarm is raised on a VNFC, the
VM state is Active, and the deployed state is Deployed. The NFM-P first attempts a soft reboot. If
the alarm is not cleared, the NFM-P then attempts a hard reboot. If the alarm is still not cleared, the
NFM-P raises a VNFCAutoHealFailed alarm on the VNFC.
The Auto Heal Timer specifies the length of time the NFM-P waits between reboot attempts before
proceeding with the next action.
Note: Automatic healing is not triggered if the LSS_hostReset alarm is raised on the active
OAM VNFC.
Automatic healing is not triggered if the VMM is in MaintState,
Reboot status
You can see the status of the VNFC reboot from the VNF Manager application by viewing the Auto
Heal State. The Auto Heal State is a real-time indicator that describes the current state of the
automatic healing process. The Auto Heal State displays one of the following statuses:
• No Attempt — no automatic healing operation was attempted
• Soft Reboot — a soft reboot has resolved the alarm that triggered the automatic reboot
• Hard Reboot — a hard reboot has resolved the alarm that triggered the automatic reboot
• Reboot failed — the soft and hard reboot have failed to resolve the alarm that triggered the
automatic reboot
10.2 To configure automatic scale-out or automatic healing

Purpose
Perform this procedure to configure automatic scale-out or automatic healing for the VMM, CMG, or
VSR. This procedure requires system administrator access rights.
Steps
Enable automatic scaling or automatic healing
1
Choose Administration→System Preferences from the NFM-P main menu. The System
Preferences form opens.
2
Click on the NFV tab and configure the required parameters for automatic scale-out or
automatic healing.
Release 18.3
March 2018
NFV use cases NFM-P
3
If you are configuring automatic scale-out for the VMM, configure the scale-out Factor (VMM)
(%), Auto scale-out Timer (VMM) (hours), and Perform Scale-out (VMM) parameters.
The scale-out Factor (VMM) (%) parameter defines the threshold that triggers the automatic
scale-out operation, where the threshold is the UE capacity of the VMM multiplied by the scale-
out factor.
The Auto scale-out Timer (VMM) (hours) parameter defines the minimum time that the NFM-P
will wait before triggering an automatic scale-out operation while a previous scale-out operation
is still in progress. You should specify a value that exceeds the expected time for a virtual MAF
to become operational.
The Perform Scale-out (VMM) parameter defines the conditions under which the automatic
scale-out is triggered. You can specify AND or OR for the following conditions:
• The number of UEs on the WMM is over capacity.
4
If you are configuring automatic healing, configure the Auto Heal Timer (VMM) (min) and Auto
Heal Timer (VMG) (min) parameters.
The Auto Heal Timer parameters specify the time that the NFM-P waits between reboot
attempts before proceeding with the next action.
5
Save your changes and close the form.
For CMG automatic scaling, configure a threshold group
6
On the equipment tree, right-click on a CMG and choose Properties. The Network Element
(Edit) form opens.
7
Choose a PGW and click Properties. The PDN Gateway (Edit) form opens.
8
Click on the Threshold Groups tab and click Create. The Threshold Group (Create) form opens.
9
Configure the required parameters.
Select Bearer Mgmt Limits for the Threshold Groups parameter.
Set the Administrative State parameter to Down before you create a threshold group counter.
Release 18.3
March 2018
NFV use cases NFM-P
10
Click on the Threshold Group Counters tab and click Create. The Threshold Group Counter
(Create) form opens.
11
Configure the required parameters.
Select Number of Bearers for the Threshold Counter parameter.
12
Save your changes and close the forms.
END OF STEPS
Release 18.3
March 2018
NFM-P VNF descriptor NFM-P
NFM-P NFV requirements
11 NFM-P VNF descriptor

VNF descriptor
11.1 NFM-P NFV requirements

Overview
The VNF Descriptor is a package that describes the configuration of the VNF network. It consists of
OpenStack Heat templates and a meta file. The Heat templates define VNF specifications, and the
meta file creates custom parameters required by NFM-P NFV. This chapter describes the
configuration requirements and recommendations for the Heat templates and meta file.
VNF Descriptor sample

You can view sample VNF Descriptor packages in the NFM-P server installation directory. The files
are located in /opt/nsp/nfmp/server/nms/sample/NFV. There are sample VNFD sample packages
for the CMG and VMM. Each sample package includes sample Heat templates and a sample meta
file. You can use these samples to help build the templates and meta file with configurations
required for your NFV network setup.
Release 18.3
March 2018
OpenStack Heat templates
Heat templates
11.2 Heat templates

Overview
The Heat orchestration engine is designed to start cloud applications based on template files that
are specialized for different types of VNFs. These Heat orchestration template (HOT) files describe
a list of OpenStack resources called a stack. OpenStack maps these resources to VMs based on
VNF requirements. HOT files list the stack resources, and specify their type and configuration
parameters. HOT files must update dynamically as the stack resources change due to lifecycle
management operations.
The NFM-P uses HOT files to interact with OpenStack Heat. Whenever a stack resource is
changed through the VNF Manager application, the NFM-P dynamically recreates the HOT file. The
lifecycle management tasks that can be performed from the VNF manager application require
specifications from pre-configured HOT files. Before you can use this functionality in the application,
you must configure these HOT files with a format specific to NFM-P functionality. The default HOT
file for a VNF is created from a generic node template. You can customize the deployment
parameters using a Heat environment file.
This chapter describes the HOT file requirements for the VNF functions supported by the NFM-P.
Note: Basic Heat environment file requirements are not described in this chapter. This chapter
describes only the additional requirements for NFM-P functionality. This chapter should be
used in conjunction with OpenStack Heat template documentation.
NFM-P NFV template components
The NFM-P requires several HOT files to manage VNFs and communicate with OpenStack Heat.
One of each of the following static template files are required per VNF type and release:
• base template
• custom resources templates
• sample environment
• growth unit template — required only for scaling operations
One of each of the following dynamic template files are required per VNF instance and lifecycle
management event:
• deployment-specific base template
• custom resource template
• deployment-specific environment file
• meta file
Release 18.3
March 2018
Deployment template design
Template archive
The NFM-P requires an archive for each VNF type (VMM, CMG, and VSR). This archive contains
templates for initial deployment and, optionally, scaling.
For example, the <NFMP_INSTALL_DIR>/os/VMG archive for a VNF contains the following files:
• VMG.hot.yaml
• VMG.env.yaml
• VMG.env.meta.yaml
• VMG_LB.template.yaml
• VMG_OAMA.template.yaml
• VMG_MG.template.yaml
• VMG_OAMB.template.yaml
• grow/
− VMG.1_LB_card + 1_VMG_card.grow.hot.yaml
− VMG.2_VMG_card.grow.hot.yaml
− VMG.default.grow.hot.yaml
The VNF folders in the <NFMP_INSTALL_DIR>/os must be created manually. The .yaml files must
be in the directory that is specified in the VNF catalog.
Note: The <NFMP_INSTALL_DIR>/os directory and all template files must have nsp
permissions.
11.3 Deployment template design

Overview
Each VNF must have one base template and a base environment file for each node release. These
template files define VNF parameters and resources. If custom resources are defined, there must
also be custom template files for each custom resource.
Base template — x.hot.yaml

The base template defines the Heat template version and description. It defines parameters that the
user can configure during VNF instantiation. The base template lists basic OpenStack Heat
resources and can optionally list custom resources.
Environment file — x.env.yaml

The environment file corresponds with the base template. It must contain all deployment-specific
parameters that are defined in the base template. The environment file contains a resource registry
for custom resources, and includes file locations for the custom templates that define these
resources. It must not contain any image or flavor names.
Release 18.3
March 2018
Custom resource template — x.template.yaml

The custom resource templates define custom resources. They must be located in the same file
directory as the environment file that lists the custom resources under the resource registry. Custom
resource templates should contain only native OpenStack Heat resource types.
Heat template version

The Heat template version must be specified at the top of the base and custom template files in the
following format:
heat_template_version: YYYY-MM-DD
where YYYY-MM-DD is a Heat template version such as 2014-10-16
Custom resources
Custom resources can be listed in the base template file. The resource types and file location must
be included under the resource_registry in the environment file and then defined in separate
template files.
Nokia recommends that all custom resource types follow the same naming convention, starting with
NOK::. Custom resource types must be derived from basic OpenStack resources with no more than
five levels of nested stacks (for example, OS::Nova::Server).
Consider the following when creating custom resources:

• Resource IDs must be unique. If two resources have the same name, OpenStack Heat will
create only one of them. It will not produce any validation errors.
• Resource names for resources of type OS::Nova::Server must contain the Heat stack name as a
prefix. This convention is required for some NFM-P monitoring functions.
• Resource names for resources of type OS::Neutron::Port and OS::Cinder::Volume should
contain the Heat stack name as a prefix. This convention improves debugging and allows for
quick association between resources and the Heat stack.
Parameters
The NFM-P parses the top level environment file that contains the deployment-specific parameters.
Each parameter entry is translated into an input parameter that can be configured by an operator
during VNF instantiation.
Consider the following when creating parameters:

• Parameters must be contained in a file with a .env.yaml extension.
• Parameter labels and comments appear in the VNF Manager application GUI during VNF
instantiation. These appear to the operator as the parameter name and description. Use clear
and descriptive labels and comments to explain the parameter functionality.
• Strings are preferred for operator input. Avoid using maps or lists, unless they are as simple as
possible.
• Nokia recommends not defining image or flavor names as parameters in the environment file.
Doing this will result in operators being required to have knowledge of image and flavor names.
Image and flavor names should be defined in the template resources for the corresponding
Release 18.3
March 2018
OS::Nova::Server resource. You can still achieve template reusability across VNF software
releases by using the software version as a parameter in the environment file. You can then use
image naming conventions that contain the prefix and the software version. See the following
example.
—
image:
str_replace
template: LCP_MAF_$sw_version
params:
$sw_version: {get_param : sw_version}
Pre-defined parameter keywords

The NFM-P supports pre-defined keywords that can be used in parameter names. When these
keywords are used, the NFM-P performs additional actions. These keywords can be used by any
VNF type when defining parameter names in the environment file. The following table describes the
pre-defined parameter keywords.
Keyword Input type NFM-P actions

auth_url String The NFM-P defines a read-only parameter with the
authentication endpoint of the OpenStack tenant into
which the VNF is being deployed.
tenant_id String The NFM-P defines a read-only parameter with the
tenant ID of the OpenStack tenant into which the
VNF is being deployed.
tenant_name String The NFM-P defines a read-only parameter with the
tenant name of the OpenStack tenant into which the
VNF is being deployed.
username String The NFM-P defines a read-only parameter with the
tenant username of the OpenStack tenant into which
the VNF is being deployed.
password String The NFM-P defines a read-only parameter with the
tenant password of the OpenStack tenant into which
the VNF is being deployed.
VNF component slot ID

When the OS::Nova::Server is defined, you must include a metadata property. The metadata
property tags the server with a slot and shelf ID for VNF components. This property stores arbitrary
key/value server metadata and helps correlate domain-specific card alarms to VNF component
alarms. See the following example.
—
type:
OS::Nova::Serverproperties:
Release 18.3
March 2018
Scaling template design
metadata: {"nokia_vnf_slotId":{get_param:slot}}
11.4 Scaling template design

Overview
Scaling templates define the resources that are added or removed when a scaling operation is
performed. For each VNF that supports scaling, there must be a scaling template. These templates
define the number and type of VNF components to be scaled. They also describe all resources that
are required for the scaling operation.
The same scaling template is used for multiple scaling operations. For example, a scale-out
operation of three VNF components is derived from the scaling template for a single VNF
component. The same scaling template is used for scale-in and scale-out operations.
The NFM-P manipulates the base deployment template to dynamically add or remove resources
defined in the scaling template. There are some scaling template configuration requirements to
facilitate this functionality.
Template naming
Scaling templates are defined by the VNF_Type.userdef_chars.grow.hot.yaml files in the grow sub-
directory. The name of the scaling templates must specify the applicable VNF. A CMG scaling
template must lead with VMG in the template name and a VMM scaling template must lead with
VMM. Nokia also recommends that the template name specify the number and type of VNF
components to be scaled. For example, a scaling template designed to scale one load balancing
VNF on a CMG might be named VMG.1LB.grow.hot.yaml.
Note: The userdef_chars part of the template filename cannot contain a period.
For automatic scaling, the NFM-P looks for a file with the name VNF_Type.default.grow.hot.yaml.
Template design
Consider the following when creating a scaling template:
• The template must include a resources section, as it would appear in the base deployment
template.
• The template must not include any new parameter definitions. Any new parameters required
must be defined in the environment file.
• Parameter and output sections are not required. The NFM-P is not able to dynamically update
these sections, so they are not parsed.
• All resources that should be created together during a scale-in or healing operation (for example,
server, ports, volume attachments, software deployments) must be grouped together in a single
custom resource or in a OS::Heat::ResourceGroup resource.
• The NFM-P does not manipulate the outputs section of a scaling template.
• The Heat template version must be specified at the top of the base and custom template files in
the following format:
heat_template_version: !!str YYYY-MM-DD
Release 18.3
March 2018
where YYYY-MM-DD is a Heat template version such as 2014-10-16

The NFM-P supports some pre-defined properties that can be used as part of a custom resource in
a scaling template. When these properties are used, the NFM-P performs additional actions.
resource_id
The resource_id property is a common prefix used for resources of the same type that are subject
to horizontal growth. Each consecutive scaling operation adds a new instance of this resource. The
NFM-P dynamically calculates the resource_id and name to be used when updating the Heat stack.
group_index
The group_index property is a list of properties defined in the scaling resource. The group_index is
used to permit dynamic increments of a value of a common property for multiple stack resources.
For example, group_index can be used for a VNF that requires a property called “Slot number” that
uses an new increment with each scale-out operation.
When a scale-out operation is performed on a template where the group_index property is used,
the NFM-P calculates the new incremental value of each property in the list based on the maximum
value of the property with the same name among the resources already configured in the stack.
Strings in the group_index property must be listed in the following format:
[optional string][optional leading 0][number]
For example: Card slot: 04
group_index example
A stack contains resourceA of custom type resA. The custom type resA has a property called
myCardNumber with a value of “05” in the initial deployment.
—
resources:
resourceA:
type: resA
properties:
deployment_prefix: {get_param: "OS::stack_name"}
myCardNumber: "05"
—
If an operator requires a scale-out operation that adds one more resource of type resA with a new
value for myCardNumber, the scaling template specifies the following:
—
resources:
resourceA:
type: resA
properties:
group_index: ["myCardNumber"]
Release 18.3
March 2018
myCardNumber: "anything"
—
When this scale-out operation is performed, the NFM-P dynamically creates the following HOT file.
The resource name and myCardNumber value are incremented.
—
resources:
resourceA:
type: resA
properties:
myCardNumber: "05"
resourceA1:
type: resA
properties:
group_index: ["myCardNumber"]
myCardNumber: "06"
—
dependency_group
The dependency_group property is a map of strings that defines related resources in the scaling
template. A scaling template contains only the basic list of objects that need to be scaled, but the
actual names of the resources being added depend on how many previous scaling operations there
were in the lifecycle of the stack. The resource names are dynamically calculated during the scaling
operation using the group_index property. The dependency_group property address the issue of a
resource that refers to another resource during a scaling operating when the name of the resource
is not yet known, having been updated due to previous scaling operations. When the dependency
of a type is identified, the dependency_group property can be used to dynamically resolve resource
names. See the following example:
—
dependency_group: {{propertyX: resourceY}}
—
For each entry in the dependency_group map, the NFM-P replaces the entry of resourceY with the
actual calculated name of the resource with the prefix resourceY in the same scaling template.
dependency_group example
A stack has resourceA of custom type resA and resourceB of custom type resB. One of the
properties of resourceA refers to resourceB.
—
resources:
resourceA:
Release 18.3
March 2018
type: resA
properties:
propertyXyz: {get_resources: resourceB}
resourceB:
type: resB
properties:
...
—
After two scale-out operations, the NFM-P dynamically updates the resource names to
accommodate for names that have been changed during scaling.
—
resources:
resourceA1:
type: resA
properties:
propertyXyz:
{get_resources: resourceB1}
resourceB1:
type: resB
properties:
...
—
resources:
resourceA2:
type: resA
properties:
propertyXyz: {get_resources: resourceB2}
resourceB2:
type: resB
properties:
...
—
ResourceGroup example
All resources that should be created together during a scale-in or healing operation (for example,
server, ports, volume attachments, software deployments) must be grouped together in a single
custom resource or in a OS::Heat::ResourceGroup resource.
—
ALU-LCP-Pair%card%:
type: OS::Heat::ResourceGroup
dependency_group: {{property3: resource_idX}}
properties:
Release 18.3
March 2018
count: <<count>>
resource_def:
type: <customized resource type NOK::>
properties:
group_index: ["property1", "property2"]
...
property1: somevalue1
property3: {get_resource: resource_idX}
...
<resource_idY>
type: <customized resource type NOK::>
properties:
group_index: ["property1", "property2"]
...
property3: {get_resource: resource_idX}
Release 18.3
March 2018
Meta file configuration and requirements
Meta file configuration
11.5 Meta file configuration

Overview
In order to improve usability and ensure OpenStack compatibility with certain NFM-P NFV
functions, you should perform the following preconfiguration tasks while creating design templates.
• Create meta files for each VNF type.
• Define VNF management IDs.
• Define VNF component slot IDs.
Meta file design

A meta file is useful for the NFM-P to properly render the environment file and display the correct
information to the user. The meta file improves usability and readability for VNF settings in the VNF
Manager UI. For example, it allows you to display drop-down menus for properties such as flavor_
map and subnets rather than plain text boxes.
You should create one meta file for each VNF type (VSR, VMM, and CMG). A meta file uses the
extension env.meta.yaml.
The NFM-P retrieves the following resources from OpenStack resourcesduring VNF instantiation.
Each of these resources should be described in a JSON format in the meta file. The nok_os_type
uses one of these resources to retrieve the appropriate resources for the environment parameters.
• networks
• subnets
• images
• flavors
• volumes
• availabilityZoneInfo
Note: If a meta file is not present, parameter groupings and OpenStack resource retrieval are
not available during VNF instantiation. The NFM-P displays the parameter keys and values as
a simple label. Properties such as flavor and image maps appear as text boxes rather than
drop-down menus.
Meta file keywords
The following keywords are specific to NFM-P NFV meta files:

• nok_readwrite — defines read/write attributes
• nok_readonly — defines read-only attributes
• nok_os_type — defines the key type in the environment file
• nok_os_display_field — specifies which OpenStack property is displayed to the NFM-P user
Release 18.3
March 2018
• nok_os_ret_field — specifies which OpenStack property is returned to the environment file

• nok_hot_env_name — defines the parameter name in the environment file
VNF scaling parameterization

You can include a grow_meta_data entry in the meta file to specify which OpenStack parameters
can be configured by the user during a VNF scaling operation. For example, you can include image
in the grow_meta_data entry to allow the user to select from a list of images during a VNF scale-
out.
The following example shows a grow_meta_data entry including the image variable.
—
grow_meta_data:{
image:{
nok_os_type:images,
nok_readwrite:{
nok_os_display_field:name,
nok_os_ret_field:id
}
}
}
—
Parameterization is supported for the following scale-out variables:

• flavor
• image
• network
• subnet
• availability zone
For each parameterized variable in the grow_meta_data entry, you must also include the variable in
the resource template .yaml using the $MKS_variable$ format. The following example shows the
entry for an LB resource with parameterization for the image and flavor variables.
—
LB-%slot%:
type: ALU::VMG::LB
properties:
stack_name: { get_param: "OS::stack_name" }
instance: 1
slot: 1
group_index: [ "slot", "instance" ]
gw_network: { get_param: [EXTnet_info, gw_network, id] }
gw_subnet: { get_param: [EXTnet_info, gw_network, gw_subnet, id] }
internal_network: { get_param: [EXTnet_info, internal_network, id]
Release 18.3
March 2018
}
mmeIf_network: { get_param: [EXTnet_info, mmeIf_network, id] }
gwIf_network: { get_param: [EXTnet_info, gwIf_network, id] }
security_group: { get_param: security_group }
image: $MKS_image$
flavor: $MKS_flavor$
—
Meta file samples

The following example shows how to map a simple JSON object such as flavor_map where all of
the keys present represent the same data type.
—
Environment file:
—
flavor_map: {
"oam-flavor": m1.medium,
"lb-flavor": m1.medium,
"mg-flavor": m1.medium,
}
imageA: " VMG_70S201“
—
Meta file:
—
flavor_map: {
nok_os_type: flavors,
nok_readwrite: {
nok_os_display_field: name,
nok_os_ret_field: id
}
}
imageA: {
nok_os_type: images,
nok_readwrite: {
}
}
—
The following example shows how to map networks and subnetworks into a meta file.
—
Environment file:
Release 18.3
March 2018
EXTnet_info: {
pnf2net: {
id: 52e6e28d-a7bf-4c57-a741-d42051011991,
V4subnet1: {
id: 99cf155f-6fb5-48a2-a4b4-552d0776c39c,
cidr: 135.111.51.128/26,
default_gateway: 135.111.51.190,
},
},
snf3net: {
id: 4e09c5bc-2290-4980-8d36-eec507f33e0e,
V4subnet1: {
id: 943fc247-dd9b-40ea-a869-de7d375a8f80,
cidr: 172.16.0.0/16,
},
V4subnet2: {
id: a94b3443-9c81-4137-9127-6714d673aa28,
cidr: 172.17.0.0/16,
},
},
}
—
Meta file:
—
meta_data :
env_meta_data: {
EXTnet_info: {
nok_os_type: networks,
nok_readwrite: [
{
nok_hot_env_name: id,
}
],
nok_subnet: {
nok_os_type: subnets,
nok_readonly: [
{
nok_hot_env_name: cidr,
nok_os_display_field: cidr
},
Release 18.3
March 2018
Other requirements
{
nok_hot_env_name: default_gateway,
nok_os_display_field: gateway_ip
}
],
nok_readwrite: [
{
nok_hot_env_name: id,
}
],
}
}
}
11.6 Other requirements

VNF management IP
For NFV functions, EMS functions, and alarm correlation, the NFM-P requires the management IP
of the VNF when it is instantiated. The management IP must be provided in a key/value format in
the parameters section of the environment file. See the following example.
—
parameters:
oama_ip: 192.169.60.60
oamb_ip: 192.169.60.61
—
In the above example, oama_ip and oamb_ip must be defined in the base template.
Note: If the VNF management IP is not defined, the VNF cannot be instantiated.
11.7 To enable OpenStack message logging

Purpose
You may need to view OpenStack request/response message logs to troubleshoot possible
communication issues. To enable OpenStack message logging on the NFM-P, you must modify the
nms-server.xml file.
After performing this procedure, OpenStack message logs can be found in the following folder:
/opt/nsp/nfmp/server/nms/log/NFV_debug
Other NFV messages are logged in the NFV.log file:
/opt/nsp/nfmp/server/nms/log/server/NFV
Release 18.3
March 2018
To enable OpenStack message logging
Steps
1
Log in to the main server station as the nsp user.
2
Open the /opt/nsp/nfmp/server/nms/config/nms-server.xml file using a plain-text
Note: Contact your Nokia technical support representative before you attempt to
modify the nms-server.xml file. Modifying the nms-server.xml file can have serious
consequences that can include service disruption.
3
Add the following line:
<systemNFVApiLog enabled="yes" path="../log/NFV_debug"/>
4
Run /opt/nsp/nfmp/server/nms/bin/nmsserver.bash read_config .
5
View message logs in the following folder:
/opt/nsp/nfmp/server/nms/log/NFV_debug
END OF STEPS
Release 18.3
March 2018

NSP Network Services Platform

Uploaded by

Copyright:

Available Formats

NSP Network Services Platform

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NSP Network Services Platform

Uploaded by

Copyright:

Available Formats

NSP

Network Services Platform

NFV Solutions Guide

About this document............................................................................................................................................5

Part I: Getting started ...........................................................................................................................................7

3 NFV solution overview.................................................................................................................................13

Part II: CBAM as VNF manager solution ..........................................................................................................15

4 CBAM as VNF manager ...............................................................................................................................17

5 Network Supervision web application .......................................................................................................23

6 VNF threshold policy templates .................................................................................................................33

7 CBAM GNE driver information....................................................................................................................41

Part III: NFM-P as VNF manager solution .........................................................................................................43

8 NFM-P as VNF manager...............................................................................................................................45

9 VNF Manager web application ....................................................................................................................51

10 NFV use cases..............................................................................................................................................57

11 NFM-P VNF descriptor .................................................................................................................................61

About this document

Part I: Getting started

Chapter 1, Safety information 9

1.1 Structure of safety statements

Item Structure element Purpose

Signal word Meaning

2.1 What’s new in NFM-P Release 18.x?

What’s new in NFM-P 18.3

Feature Description and Reference

3 NFV solution overview

3.1 Network function virtualization overview

Part II: CBAM as VNF manager solution

Chapter 4, CBAM as VNF manager 17

4 CBAM as VNF manager

4.1 CBAM NFV solution

XML API NFV Management

Service, VNF and

VNFs (CMM, CBAM

Virtual Virtual Virtual Virtualized

Computing Storage Network

CBAM access point

The following are prerequisites before creating a CBAM access point:

Lifecycle change notifications

VNF lifecycle management

Note: For the CMM, only CPPS modules can be scaled.

VNF threshold policies

Network Supervision application

4.2 CBAM VNFD requirements

5 Network Supervision web application

5.1 Network Supervision web application

Product help tours

Figure 5-1 Network Supervision UI

Figure 5-2 Network Supervision alarm list

CBAM access point creation

Figure 5-3 CBAM access point details tab

VNF monitoring and management

Figure 5-4 CMM matrix view tile

Figure 5-5 Manual scaling operation

VNF threshold policies

Figure 5-6 VNF threshold policy creation

KPI monitoring window

Perform the following to resolve the alarm:

Automatic healing wait time

Applying a VNF threshold policy