So I hit a bug the other day with the Cisco Nexus 1000v and thought I’d spread the word, just in case…. My setup was an ESX farm using Generation 1 Emulex CNAs. Since you can’t do a standard vPC with Gen 1 cards (known limitation…not a problem with Gen 2) I was using vPC-HM with the Nexus 1000v. Like a good little engineer I was running redundant VSMs (Virtual Supervisor Modules). The problem was that when I would take down both VSMs and reboot them to simulate a maintenance event the VEMs (Virtual Ethernet Modules) wouldn’t reconnect and the VSM console would get flooded with:
2009 Aug 14 02:33:31 N1KvSW %KERN-1-SYSTEM_MSG: Dropping received frames from duplicate VSM – kernel
Every time a VEM would come online and talk to the VEM it would get flooded with these and drop. I’ve found out that this is a bug and will be fixed. In the mean time if you hit this my work around was to log in to the Nexus 5K that one set of CNAs was plugged in to and down those ports leaving only one CNA per hot connected. Once everything stabilizes and the VEMs are talking (and CDP info is being passed!) you can “no shut” those ports again and the vPC-HMs come up no problem.
Just something to keep in mind.
If this was all non-sense to you or you aren’t really familiar with the Nexus 5K and 1Kv yet check these out:
[...] Nash brings to light a bug in Cisco Nexus 1000V when used in conjunction with CNAs. Be sure to have a look if this has any similarity to your [...]