Kernel Panics on Host Machines running 2.6.32-16

[msmart]

My host machines are hard locking on VM start (and intermittently elsewhere) with a kernel:Kernel panic - not syncing: Fatal exception in interrupt message.

The same nodes appear to function without issue on 2.6.32-11.
All of the nodes that are exhibiting this behavior are dell blade servers with AMD Opteron 2378 procs.
I have a couple nodes that are not dell blades and they behave fine running 2.6.32-16

I am not sure where to look for debug info to see about narrowing this down. Nothing jumped out at me in dmesg or syslog. Let me know what info you would like to see and I will post it subsequently.

Thanks in advance,
Matt

 

[e100]

I do not know of any specific issues with your hardware but have you checked for BiOS updates?

 

[msmart]

I have verified that I am running the latest bios. Thanks for the idea though.

I should mention that this hardware has been running proxmox in production for a couple years now. I have been running 2.1 on the nodes being affected by this issue for some time and it has been rock solid.