Allen-Bradley Stratix 5700™ Network Address Translation (NAT)
Allen-Bradley Stratix 5700™ Network Address Translation (NAT)
Allen-Bradley Stratix 5700™ Network Address Translation (NAT)
Synopsis
Machine integration onto a plant’s network architecture can be difficult as OEM
IP-address assignments rarely match those of the end-user network and network
IP addresses are generally unknown until the machine is being installed – adding
cost and time to the commissioning of the equipment, and delays moving that
equipment into production.
The Allen-Bradley Stratix 5700 with Network Address Translation (NAT) is a hardware
Layer 2 implementation that provides “wire speed” 1:1 translations ideal for
automation applications where performance is critical.
NAT allows for:
• High performance and simplified integration of IP-address mapping from a set of local, machine-level IP addresses to the
end user’s broader plant network
• OEMs to deliver standard machines to end users without programming unique IP addresses
• End users to more simply integrate the machines into the larger network
• Easier machine maintenance because machine configuration remains standard
The Stratix 5700 switch with NAT technology also allows users to have the flexibility to segment or isolate network traffic by
determining which devices are exposed to the larger network. By limiting access to certain devices, they can be isolated from
unneeded network traffic, which can help optimize network performance at the local level.
Line Controller
Stratix 8300
0 1 2 3 4 5 6 7 A0 B0 Z0 0 1 2 3 4 5 6 7 A0 B0 Z0
Module Module
SOURCE
SOURCE
IN
24VDC
HIGH SPEED
IN
24VDC
HIGH SPEED
SINK\
INPUT
SINK\
INPUT
DC
Status
DC
Status
COUNTER
COUNTER
8 9 10 11 12 13 14 15 A1 B1 Z1 8 9 10 11 12 13 14 15 A1 B1 Z1
Network C Network C
OUT
OUT
SOURCE
OUTPUT
SOURCE
OUTPUT
24VDC
0 2 FUSE
24VDC
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 2 FUSE
DC
DC
Activity Activity
1 3 OK 1 3 OK
8 9 10 11 12 13 14 15 A D 8 9 10 11 12 13 14 15 A D
Network B E Network B E
Status Status
ETHERNET
A=ENABLE A=ENABLE
System System
05 13 Z1+ Z1- Power
B= REGEN
C=DATA ENTRY 05 13 Z1+ Z1- Power
B= REGEN
C=DATA ENTRY
D=FAULT D=FAULT
06 14 +V -V Field E=COM ACTIVITY 06 14 +V -V Field E=COM ACTIVITY
Power Power
07 15 OUT OUT 07 15 OUT OUT
0 2 0 2
COM COM OUT OUT COM COM OUT OUT
0 1 1 3 0 1 1 3
NC NC COM COM NC NC COM COM
V V V V
+V +V OUT OUT
0+ 0+ +V +V OUT OUT
0+ 0+
V V V V
00 08 OUT OUT
0+ 0+
24VDC 00 08 OUT OUT
0+ 0+
24VDC
V V
OUT OUT
INPUT V V
OUT OUT
INPUT
01 09 0+ 0+ 01 09 0+ 0+
V V V V
02 10 OUT OUT
0+ 0+ 02 10 OUT OUT
0+ 0+
V V
03 11 COM OUT
0+ 03 11 COM OUT
0+
DANGER V DANGER V
04 12 COM OUT
0+ 04 12 COM OUT
0+
00:00:BC:66:0F:C7 00:00:BC:66:0F:C7
MORTOR FEEDBACK
MORTOR FEEDBACK
V V V V
05 13 OUT OUT
0+ 0+ 05 13 OUT OUT
0+ 0+
V V V V
06 14 OUT OUT
0+ 0+ 06 14 OUT OUT
0+ 0+
I I I I
07 15 OUT OUT
0+ 0+ 07 15 OUT OUT
0+ 0+
COM COM COM COM BRAKE/ COM COM COM COM BRAKE/
0 1 DC BUS 0 1 DC BUS
What Is NAT?
Network Address Translation is a service that can translate a packet from one IP address
to another IP address. NAT can be found either on a Layer 2 device or on a Layer 3 device.
NAT can be understood easiest with the introduction of the concept of a private network
and a public network (Figure 2)*. These two networks are separated by a boundary;
a device that implements NAT is this boundary. NAT can take on multiple forms including
one-to-many NAT and one-to-one NAT (our implementation).
One-to-many NAT is also known as Port Address Translation and allows one public IP
address to be shared by many private IP addresses. This function is commonly found in
consumer grade routers. A one-to-many NAT device contains a table that allows unique
private host ports to be exposed on the single public IP address (Figure 3).
Public Subnet
(Example: 10.0.0.X)
Private Subnet
(Example: 192.168.1.X)
• Note that we use the terms private and public to differentiate the two networks on either side of the NAT device.
This does not infer that the public network must be Internet routable
Stratix 5700 NAT Whitepaper | 3
1:1 NAT allows a manufacturer to keep duplicate machines identical while providing a
unique identity (alias) to the larger industrial network. The feature also gives a granular
method of granting or restricting access to an end device (I/O blocks, drives, etc.) on the
machine in one place.
1:1 NAT works by replacing the IP header on a packet and recalculating the packet
checksums as it finds the appropriate match in the NAT table when it passes through
the NAT device (Figure 5).
Downlink Ports
The NAT feature is integrated in hardware between the uplink ports and the rest of the
switch. It supports one or two uplinks which can be used in star, redundant star and ring
topologies. One uplink would be used for a standard star topology. Two uplinks could be
used for either a redundant star using spanning tree, or a ring topology using REP (Figure 7).
Line Controller
REP
0 1 2 3 4 5 6 7 A0 B0 Z0 0 1 2 3 4 5 6 7 A0 B0 Z0
Module Module
SOURCE
SOURCE
24VDC
HIGH SPEED
IN
HIGH SPEED
IN
24VDC
SINK\
INPUT
SINK\
INPUT
DC
Status
DC
Status
COUNTER
COUNTER
8 9 10 11 12 13 14 15 A1 B1 Z1 8 9 10 11 12 13 14 15 A1 B1 Z1
Network C Network C
OUT
OUT
SOURCE
OUTPUT
SOURCE
OUTPUT
24VDC
0 2 FUSE
24VDC
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 2 FUSE
DC
DC
Activity Activity
1 3 OK 1 3 OK
8 9 10 11 12 13 14 15 A D 8 9 10 11 12 13 14 15 A D
Network B E Network B E
Status Status
ETHERNET
A=ENABLE A=ENABLE
System System
05 13 Z1+ Z1- Power
B= REGEN
C=DATA ENTRY 05 13 Z1+ Z1- Power
B= REGEN
C=DATA ENTRY
D=FAULT D=FAULT
06 14 +V -V Field E=COM ACTIVITY 06 14 +V -V Field E=COM ACTIVITY
Power Power
07 15 OUT OUT 07 15 OUT OUT
0 2 0 2
COM COM OUT OUT COM COM OUT OUT
0 1 1 3 0 1 1 3
NC NC COM COM NC NC COM COM
V V V V
+V +V OUT OUT
0+ 0+ +V +V OUT OUT
0+ 0+
V V V V
00 08 OUT OUT
0+ 0+
24VDC 00 08 OUT OUT
0+ 0+
24VDC
V V
OUT OUT
INPUT V V
OUT OUT
INPUT
01 09 0+ 0+ 01 09 0+ 0+
V V V V
02 10 OUT OUT
0+ 0+ 02 10 OUT OUT
0+ 0+
V V
03 11 COM OUT
0+ 03 11 COM OUT
0+
DANGER V DANGER V
04 12 COM OUT
0+ 04 12 COM OUT
0+
00:00:BC:66:0F:C7 00:00:BC:66:0F:C7
MORTOR FEEDBACK
MORTOR FEEDBACK
V V V V
05 13 OUT OUT
0+ 0+ 05 13 OUT OUT
0+ 0+
V V V V
06 14 OUT OUT
0+ 0+ 06 14 OUT OUT
0+ 0+
I I I I
07 15 OUT OUT
0+ 0+ 07 15 OUT OUT
0+ 0+
COM COM COM COM BRAKE/ COM COM COM COM BRAKE/
0 1 DC BUS 0 1 DC BUS
NAT Instances
The software interface for NAT has been implemented using the concept of instances
(Figure 8). Each instance contains a name, “Private to Public” NAT table, “Public to Private”
NAT table, VLAN and interface association, specific packet fix-ups, and specific types of
traffic that can be blocked or passed through. Typically only one instance will be used;
however, multiple instances can be supported to differentiate between different VLAN
configurations. NAT can be attached to one or many VLANs, but will not translate traffic
across VLANS (i.e. change a VLAN tag) or break other existing VLAN rules.
NAT Tables
An important concept of the Layer 2 NAT implementation is how NAT interacts with
private and public subnets (Figure 9). Each NAT instance has two tables, a “Private to
Public” table, and a “Public to Private” table. “Private” devices must be assigned a unique IP
address in the “Private to Public” NAT table on the “public” subnet. Likewise “public” devices
must be assigned unique IP addresses in the “Public to Private” NAT table on the “private”
subnet. The implementer is responsible for defining these addresses. The addresses must
be unique and unused on other attached devices and throughout the switch.
Public Subnet
Downlink Ports
Private Subnet
Each private subnet device that is expected to talk on the public subnet must have
a “Private to Public” translation. However, not all private subnet devices must have
translations. They can be kept behind the NAT barrier to increase security, decrease traffic
on the uplink port, and conserve public address space.
If the uplinks are connected to a Layer 3 switch or router, only one “Public to Private”
translation must be used – the default gateway (Figure 10). If the uplinks are connected via
a Layer 2 switch to other devices on the public subnet, each public subnet device must
have a unique IP address in the “Public to Private” table.
Types Of Translations
There are three types of translations that can be defined: single, range and subnet.
A single translation will have one private address and one public address. A range will have
a starting private address, a starting public address and a number of entries (Figure 11).
A subnet translation allows the definition of a Class “B” subnet (mask: 255.255.0.0), Class
“C” subnet (mask: 255.255.255.0) or a fraction of a Class “C” subnet (Figure 12). A maximum
number of 128 NAT entries can be created per switch. These entries can be defined in one
instance or up to 128 separate instances. These entries can be of any type and are defined
by the rules below (Table 1). Subnet translations will have a starting private address and
public address that must be aligned on proper subnet boundaries (Table 2).
The use of a subnet translation will allow for many more than 128 IP addresses to be
translated. For example the table shown in Figure 11 uses 12 NAT entries but provides
75 actual translations.
8 | Stratix 5700 NAT Whitepaper
VLANs
When configuring NAT, you can assign one or more VLANs to a NAT instance (Figure 13).
When you assign a VLAN to a NAT instance, the traffic associated with that VLAN is subject
to the configuration parameters of the NAT instance. Configuration parameters include
whether traffic is translated, fixed up, blocked or passed through.
Certain types of traffic may have IP addresses embedded within the packet and may
need to be “fixed-up” for them to work properly (Figure 14). Two types of traffic can be
fixed-up – ARP and ICMP. Fix-ups can be assigned on a per-instance basis. These are
typically enabled in all configurations.
Unsupported Traffic
The following is a list of traffic that is not supported across the NAT boundary due
to its use of embedded IP addresses that are not fixed-up, encrypted IP addresses, or
reliance on multicast traffic. This traffic is supported on either side of the NAT boundary.
These limitations are typical for all NAT devices.
• Traffic encryption and integrity checking protocols generally incompatible with
NAT (e.g. IPSec transport mode*)
• Applications that use dynamic session initiations, such as Netmeeting*
• FTP*
• Rockwell Automation 1791-ES safety module (IP address is in the safety signature and is
not fixed-up) This is planned to be changed in V22.
• Microsoft DCOM (used in OPC communications)
• Multicast traffic and applications which use multicast including CIP Sync™ (IEEE-1588)
and ControlLogix® redundancy
*
Source: www.tcpipguide.com
RSLinx Support
As of RSLinx® 3.51, IP addresses that are changed with NAT will be shown using the
Ethernet Devices driver (Figure 15). You can tell your device is NAT’ed because the IP
Address on the “Port Configuration” screen does not match the header and address you
used to browse to the device. The EtherNet/IP driver will show the NAT’ed address of the
device, but you will not be able to connect.
10 | Stratix 5700 NAT Whitepaper
Statistics
Statistics for NAT on the Stratix 5700 provide the ability to “drill down” into the configuration
(Figure 16). This allows the user to see a global view for both operation and loading, then
drill down into specific instances to see a detailed analysis of traffic for troubleshooting
purposes (Figure 17).
Use Cases
Example 1: Using NAT With A Layer 3 Uplink
This scenario shows communications between the Line Controller (LC) and
Controller 1 (C1) and the LC to Controller 2 (C2) with a Layer 3 switch, such as the
Stratix 8300™, or router in between.
The LC and HMI are on the same VLAN, but a separate VLAN from Machine 1 (M1) and
Machine 2 (M2). M1 and M2 are on the same VLAN and subnet. M1 is a duplicate of M2,
so each share exactly the same IP Address configuration.
HMI 10.200.1.2
Mach
Machin
inee 1 (M1)
Machine ( 1))
(M Machine 2 (M2)
Stratix 5700 NAT Stratix 5700 NAT
(NAT1) (NAT2)
192.168.1.2 192.168.1.2
VLAN 10 VLAN 10 VLAN 10 VLAN 10 VLAN 10 VLAN 10
Figure 18 – Example 1
The circles show devices that fall within the private subnet of each Stratix 5700 NAT
enabled switch. Communications between devices within the private subnet use private IP
addressing schemes, for instance the I/O device (in M1) would not need a translation to talk
to C1 and vice versa.
In this example NAT translations are done through port Gi1/1 of both switch NAT1
and switch NAT2.
12 | Stratix 5700 NAT Whitepaper
C1 to LC Setup
This setup includes a translation for C1, giving it a public address of 10.10.1.10 and a
translation for the default gateway. 10.10.1.10 is an address that could be any unused address
on the 10.10.1.x subnet.
C1 will have a default gateway selected to be 192.168.1.1, which is an alias to 10.10.1.1.
Once again, 192.168.1.1 could be any unused address in the 192.168.1.x subnet. Each device
on the 192.168.1.x subnet will need to be configured to have a default gateway of 192.168.1.1.
With this setup, C1 will be accessible to the LC, HMI and any other routed device on a
different subnet (Figure 19).
C2 to LC Setup
This setup includes a translation for C2, giving it a public address of 10.10.1.11 and a
translation for the default gateway. 10.10.1.11 is an address that could be any unused address
on the 10.10.1.x subnet. C2 will have a default gateway of 192.168.1.1, which is an alias to
10.10.1.1. Once again, 192.168.1.1 could be any unused address in the 192.168.1.x subnet.
Each device on the 192.168.1.x subnet will need to be configured to have a default gateway
of 192.168.1.1. With this setup C2 will be accessible to the LC, HMI and any other routed
device on a different subnet (Figure 20).
The NAT instances on each switch will be attached to VLAN 10 of Interface Gi1/1.
Line Controller
10.200.1.100
Stratix 8300
VLAN 10: 10.10.1.1
192.168.1.1 (NAT GW)
VLAN 200 GW: 10.200.1.1
chine 1 (M1)
Machine ( 1))
(M Mach
Machin
chine 2 (M
in
Machine (M2)
Stratix 5700 NAT Stratix 5700 NAT
(NAT1) (NAT2)
192.168.1.2 192.168.1.2
VLAN 10 VLAN 10 VLAN 10 VLAN 10
VLAN 10 VLAN 10
Machine
Ma
Mach
chin ((M3)
inee 3 (M 3)) Mach
Ma chin
chinee 4 (M
in
Machine (M4)
4)
Stratix 5700 NAT Stratix 5700 NAT
(NAT3) (NAT4)
192.168.1.2 192.168.1.2
VLAN 10 VVLAN
LAN 10 VLAN 10 VLAN 10 VLAN 10 VLAN 10
Figure 21 – Example 2
This scenario shows communications both between the Line Controller (LC) and Controller 1
(C1) and the Line Controller (LC) to Controller 2 (C2) in a ring configuration. Communications
flow through a Layer 3 switch or router (such as the Stratix 8300) in between.
The LC is on a separate VLAN from Machine 1 (M1) and Machine 2 (M2). M1 and M2 are on
the same VLAN and subnet. M1 is a duplicate of M2, so each share exactly the same IP
address configuration.
The circles show devices that fall within the private subnet of each Stratix 5700 NAT
enabled switch. Communications between devices within the private subnet use private
IP addressing schemes, for instance the I/O device (in M1) would not need a translation to
talk to C1 and vice versa.
In this example, NAT translations are done through both ports Gi1/1 and Gi1/2 of each of
the NAT enabled switches.
14 | Stratix 5700 NAT Whitepaper
C1 to LC Setup
This setup includes a translation for C1, giving it a public address of 10.10.1.10 and a
translation for the default gateway. 10.10.1.10 is an address that could be any unused address
on the 10.10.1.10 subnet.
C1 will have a default gateway that has been selected to be 192.168.1.1, which is an alias to
10.10.1.1. Once again, 192.168.1.1 could be any unused address in the 192.168.1.x subnet.
Each device on the 192.168.1.x subnet will need to be configured to have a default gateway
of 192.168.1.1. With this setup C1 will be accessible to the LC and any other routed device
on a different subnet (Figure 22). In this scenario the translation will be applied to the same
VLAN (10) on both ports Gi1/1 and Gi1/2. This will allow ring topologies to converge.
C2 to LC Setup
This setup includes a translation for C2, giving it a public address of 10.10.1.11 and a
translation for the default gateway. 10.10.1.11 is an address that could be any unused address
on the 10.10.1.x subnet. C2 will have a default gateway of 192.168.1.1, which is an alias to
10.10.1.1. Once again, 192.168.1.1 could be any unused address in the 192.168.1.x subnet.
Each device on the 192.168.1.x subnet will need to be configured to have a default gateway
of 192.168.1.1. With this setup C2 will be accessible to the LC, and any other routed device on
a different subnet (Figure 23).
In this scenario the translation will be applied to the same VLAN (10) on both ports Gi1/1
and Gi1/2. This will allow ring topologies to converge.
C2 to LC
VLAN 10 VLAN 10
Mach
Ma chin
inee 1 (M1)
Machine ( 1))
(M Machine 2 (M2)
Stratix 5700 NAT Stratix 5700 NAT
(NAT1) (NAT2)
192.168.1.2 192.168.1.2
VLAN 10 VLAN 10 VLAN 10 VLAN 10 VLAN 10 VLAN 10
Figure 24 – Example 3
This scenario shows communications both between the Line Controller (LC) and
Controller 1 (C1) and Line Controller (LC) to Controller 2 (C2) with a Layer 2 Switch such as the
Stratix 8000™ in between.
In this example, everything is on the same VLAN but there are three separate subnets.
The circles show devices that fall within the private subnet of each Stratix 5700 NAT enabled
switch. Communications between devices within the private subnet use private IP
addressing schemes, for instance the I/O device (in M1) would not need a translation to
talk to C1 and vice versa.
In this example NAT translations are done through port Gi1/1 of both switch NAT1 and
switch NAT2.
16 | Stratix 5700 NAT Whitepaper
C1 to LC Setup
This setup includes a translation for C1, giving it a public address of 10.10.1.10 and a
translation for the LC. 10.10.1.10 is an address that could be any unused address on the
10.10.1.x subnet (Figure 25).
The LC has an alias of 192.168.1.100, and device C1 does not need a gateway defined to talk
to the LC. 192.168.1.100 is an address that could be any unused address on the 192.168.1.x
subnet. With this setup, C1 will be accessible to the LC and any device on its private subnet
(I/O1, Drive1) (Figure 26).
Stratix 5700 NAT Whitepaper | 17
C2 to LC Setup
This setup includes a translation for C2, giving it a public address of 10.10.1.11 and a
translation for the LC. 10.10.1.11 is an address that could be any unused address on the
10.10.1.x subnet (Figure 27).
The LC has an alias of 192.168.1.100, and device C2 does not need a gateway defined to talk
to the LC. 192.168.1.100 is an address that could be any unused address on the 192.168.1.x
subnet. With this setup, C2 will be accessible to the LC and any device on its private subnet
(I/O2, Drive2) (Figure 28).
The NAT instances on each switch will be attached to VLAN 10 of Interface Gi1/1.
In this example, if C1 or C2 wants to send a message to the LC, the destination address
specified in C1 or C2 would be 192.168.1.100.
18 | Stratix 5700 NAT Whitepaper
Mach
Machin
inee 1 (M1)
Machine ( 1))
(M Machine 2 (M2)
Stratix 5700 NAT Stratix 5700 NAT
(NAT1) (NAT2)
192.168.1.1 192.168.1.1
VLANN 10 VLAN
10 VLAN 10 VL
VLAAN 10
1 VLAN
LANN 10
10 VLAN 10 VLAN
VLAAN
VL AN 10
VLAN 10
Figure 29 – Example 4
This scenario shows communications between Controller 1 (C1) and Controller 2 (C2) with
two NAT enabled Stratix 5700 switches communicating directly with each other. In this
example, everything is on the same VLAN but there are three separate subnets.
The circles show devices that fall within the private subnet of each Stratix 5700 NAT
enabled switch. Communications between devices within the private subnet use private IP
addressing schemes, for instance the I/O device (in M1) would not need a translation to talk
to C1 and vice versa.
C1 to C2 Setup
This setup includes a translation for C1, giving it a public address of 10.10.1.10 and a
translation for C2. C2 has a public alias of 192.168.1.20, and device C1 does not need a
gateway defined to talk to C2 (Figure 30, Figure 31).
The NAT instances on each switch will be attached to VLAN 10 of Interface Gi1/1.
Summary
Whether you’re a machine and equipment builder or end user, Network Address Translation
(NAT) can provide “wire speed” 1:1 IP address translations ideal for automation applications where
performance is critical.
Allen-Bradley, LISTEN. THINK. SOLVE. and Rockwell Software are trademarks of Rockwell Automation, Inc.
Trademarks not belonging to Rockwell Automation are property of their respective companies.
Publication ENET-WP032A-EN-E – August 2013 Copyright © 2013 Rockwell Automation, Inc. All Rights Reserved. Printed in USA