N5K vPC double failure behaviour

[joan.pijpker]

Hi guys,

I have a setup with two Nexus 5548UP switches in a vPC configuration. There is a NetApp filer connected to the two switches with a LACP port-channel.

While testing we found that when we shut down the vPC peer-link and the keep-alive both switches become primary, that's the way it should be. However our connected NetApp filer becomes unreachable. Is this normal behaviour or is there something wrong in our configuration? Documentation on dual-active scenarios by Cisco states that attached devices should remain reachable, at least that is what I understand.

TIA!

Joan.

[Jerry Ye]

Taken the PL and PKL together will create a dual-active scenario and this is normal for traffic flow, since it will behave abnormally. We saw the same behavior in the lab with IP TV stream.

Can you provide the link of the link of the documentation?

Regards,

jerry

[joan.pijpker]

Hi Jerry,

The link is:

http://www.cisco.com/en/US/prod/collateral/switches/ps9441/ps9670/C07-572835-00_NX-OS_vPC_DG.pdf

See page 19/20.

Regards, Joan.

[Jerry Ye]

Hi Joan,

If you read to again, it said it might work or might not work:

"While not intuitive, if both the vPC peer link and Peer Keepalive link fail, it is still possible and desirable to continue forwarding traffic from the access layer to the aggregation layer without drops for existing traffic flows, provided that the Address Resolution Protocol (ARP) tables are already populated on both Cisco Nexus 7000 Series peers for all needed hosts."

The key word is "possible" and "existing traffic flow". I am not sure what is your test scenario and I only tell my customer is dual-active = bad and don't try it in production.

HTH,

jerry