[CGN VSM] Logical Partition + Throughput

[Carlos A. Silva]

Hi, 

While reading through the CGN VSM deployment guide:

https://community.cisco.com/t5/service-providers-documents/asr9000-xr-cgv6-on-vsm-cgn-nat44-deployment-guide/tac-p/3741638#M5431

I saw the following:

• Logical Partitioning inside VSM - Internally, VSM is divided into 2 Logic Partitions. Hence, in order to achive full throughput and scale limit (per VSM), you should have minimal of 2 Inside VRFs, 4 ServiceApp interfaces (2 in each partition), 2 Public IP address pool, etc. Traffic / subscribers which needs to avail NAT44 service also needs to be divided into 2 groups (each one associated to each Inside VRF). Following table shows the partitioning of ServiceApp interfaces between these 2 groups (which needs to be kept in mind during configuration).

It is my understanding from reading this section that for me to use full VSM throughput, I would have to configure, at least, two inside serviceapp interfaces (plus 2 corresponding outside serviceapps). Is this correct? What would be the expected behavior if I just configure a single pair SA1(outside)/SA2(inside)? Would only half the throughput be available to me?

The reason I ask is that I currently have a setup with single in/out serviceapps and yet the VSM is processing around 30G in peak hours (2G in2out, 28G out2in), which is about what I expected full throughput to be around.

Can anyone please clarify this (with an example if possible)?

Thanks!

c.

[ebenav11]

Hi @Carlos A. Silva I was deployed the CGN on VSM 500 by many years.

From your post "It is my understanding from reading this section that for me to use full VSM throughput, I would have to configure, at least, two inside serviceapp interfaces (plus 2 corresponding outside serviceapps). Is this correct?"

For to use the VSM full you need almost 4 ServiceApp interface, One pair in/out in the logical partition 1, the second pair in the logicali partition 2 

What would be the expected behavior if I just configure a single pair SA1(outside)/SA2(inside)?

You to use one partinion on the VSM-500

Would only half the throughput be available to me?

Yes

Internally, VSM is divided into 2 Logic Partitions. Hence, in order to achive full throughput and scale limit (perVSM), you should have minimal of 2 Inside VRFs, 4 ServiceAppinterfaces (2ineachpartition), 2 Public IP address pool, etc.

Traffic / subscribers which needs to avail NAT44 service also needs to bedivided into 2 groups (each one associated to each Inside VRF). Following table shows the partitioning of Service Appinterfaces between these 2 groups (which needs to be kept in mind during configuration).

[ebenav11]

Use the ABF ACL´s for the traffic redirection

KR!