Cisco UCS And 1000v Weirdness v1.3(1c) and VEM v121-esx_184.108.40.206.3.1.0-1.20.2 – Isolated Issue
Note: This is not a bug, just an isolated issue with one of our blades
So we had one of our ESX host that did not want to funtion correctly once 1000v and vSphere was patched. The virtual switch worked fine, but the deistributed Cisco 1000v switch did not. When a VM was migrated over, the pings would drop. VLAN settings and NIC assignments were all verified. I tried to the following to fix the issue
1. Rebooted VM – Still no network
2. Remove vSphere host from distributed switch and re-added – Still no network
3. Removed 1000v VEM from the host manually, rebooted, and VEM added – Still no network
4. Removed 1000v VEM, shutdown the host, disassociated service profile, re-associated, booted, and added back to the distributed virtual switch (1000v) – worked!
It was the 4th attempt that worked. For some reason 1000v seemed to not be picking up the VLANs. I know that there is a bug out there for this, but it was for an older version of 1000v. Either way, just wanted to say if all else fails, disassocate the service profile, then re-associate it back to the same blade. By doing this, it re-pushes the service profile configuration and basically re-burns the config to the blade.