Can I use the same TEP Pool in both sites when using Multi-POD?

[a.azambuja]

Can I use the same TEP Pool in both sites when using Multi-POD since my network is too small with 2 spines and 4 leaves in site A and 1 spine and 2 leaves at site B and I want to configure all APICs in same subnet with contiguous address like 10.0.0.1, 10.0.0.2, 10.0.0.3...?

[micgarc2]

No you need unique TEP pools for each POD. Each Pod should be assigned a separate and non-overlapping TEP pool.

Thank you for participating in the Cisco Support Forum for ACI! If you have other questions related to this post, please let us know. If this response answers your questions, please mark this post "answered" and assign a rating to the response(s) provided. This will help notify other viewers that your question(s) is answered and this helps us provide better responses for this and future questions.

Regards,

Michael G.

[richmond · Accepted solution]

While you use different subnets for each pod's TEP pool, the APICs will take their infra IP from the pod 1 TEP pool and be numbered like your example.

E.g. if pod 1 has TEP pool 10.1.0.0/16 and pod 2 has 10.2.0.0/16 then the APICs will be addressed as 10.1.0.1, 10.1.0.2 and 10.1.0.3, even when the APIC resides physically in pod 2.

[micgarc2 · Accepted solution]

Yes good point and sorry if that is confusing. Essentially with multipod this is all still one "cluster". Therefore all APICs will use the TEP pool of POD1. Then you would have a separate TEP pool for the actual allocation of nodes in POD2.

[a.azambuja]

Many thanks for the answer!

I was in doubt since the doc bellow shows the same TEP Pool in both PODs.

https://www.cisco.com/c/en/us/solutions/collateral/data-center-virtualization/application-centric-infrastructure/white-paper-c11-739714.html

Tks!

[RedNectar · Accepted solution]

Hi @a.azambuja,

The document is correct.  The point at which you see the same TEP pool IPs is in Table 1 of that document (https://www.cisco.com/c/en/us/solutions/collateral/data-center-virtualization/application-centric-infrastructure/white-paper-c11-739714.html).

However, if you read the following text, the explanation is quote clear:

The same TEP address pool (10.111.0.0/16 in the specific example in Table 1) must be used as the initialization configuration parameter for all of the APIC nodes that are connected to the multipod fabric, independently from the specific pod to which the nodes are connected. This is because all of the APIC nodes get assigned an IP address from the TEP pool that is associated to the first pod that was brought up as part of the multipod fabric (also known as the “seed” pod).

The diagrams in the document show (you may need a magnifying glass) that Pod2 is using TEP addresses of 10.112.0.0/16

I hope this helps

---

---

[a.azambuja]

You´re absolutely right!
Sorry about my mistake!
Tks!!!