Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. If you'd prefer to explore, try our test area to get started. And see here for current known issues.

New Member

Nexus 5000 vPC suspended during reload delay period

Hi ,

after reloading on vPC-Peer-Switch be box comes up and all vPC-Member-Ports on the box are in suspended state until the reload delay time expired.

Unfortunately the link of the vPC-Member Ports are already up. This behaviour leads us in some problems if we connect a Cisco-UCS-FI with a LACP-Portchannel to a vPC on N5K.

Because the link of the suspended Port is up the FI detects the port also as up and running and set it to individual state, because of missing LACP-BPDUs, So at this time the FI hast two uplinks, one Port-Channel and one individual Ports. After 30 seconds the FI starts to repinning the servers over these two uplinks. Because the individual Port is not in forwarding state an the reloaded N5K until reload delay timer expired.

So during this period all the servers which are pinned to the individual Port are blackholed.

Possible Workarrounds

1. Creating a Pin-Group for the Port-Channel and pinning all Servers to this Pin-Group to avoid in case on channel-Member goes to individual state, any server is pinned to this individual Port . This could be a solution

2.Configuring the Port-Channel on FI for "suspend individual". Unfortunately I could not find a way to achive this. This would avoid that the individual Port is considered as possible uplink-port, so no pinning to the individual Port would happen.

3. Find a way that during the delay restore time on the suspended vPC-Member-Ports also the link is down. (In my opinion this would be the best way)

I am not sure if configuration of "individual suspend" on the vPC on the N5K would help.

any other ideas?


New Member

Re: Nexus 5000 vPC suspended during reload delay period

What I really want is a command I can use to prevent VPC from turning off ports at all.  I'd much rather have an active-active situation than have my entire network go down just because the primary VPC peer rebooted. VPC is not designed correctly to deal with that situation.  And yes, it has happened.  Multiple times with different VPC keepalive setups.