We had a major power loss in our server room two days ago and several Cisco Nexus 2232PP FEX devices went down. We were able to get everything back online, but since that event we've been experiencing some crazy network issues, high latency, RDP sessions disconnecting and reconnecting, slow access to shared folders. When I connect to the Hyper-V host and use the console to view individual VMs, they show "connected" with internet access, but they cannot be pinged from anything other than the VM host. I have to manually go into the settings for the specific VM in the hyper-v manager, disable and re-enable the network device, and they immediately pop back to life for a while (sometimes 12 hours, sometimes just 2 hours). Both of our VM hosts are plugged into the FEX devices. Our network admin is on vacation and unreachable (i'm the sysadmin). Has anyone ever experienced anything like this, or a second question, does anyone know how to connect to a FEX and see if that could be the problem?
Asked
Active
Viewed 694 times
1
-
What kinf of network adapters on the host? If broadcomm verify drivers & firmware are updated. maybe disable VMQ. – Zoredache Oct 23 '14 at 15:53
-
some of the network adapters are broadcom. I knew about the VMQ issue and disabled it on every VM last night, but still having problems. I just flushed all of the roles from one host to the other, so while i have an empty VM host i'll try updating drivers and firmware. – user44650 Oct 23 '14 at 16:01
-
What are you doing for link. aggregation? If using a mode that requires switch config, then check the switch. – Zoredache Oct 23 '14 at 16:14