Hi guys, well I ran the command
esxcli system settings kernel set --setting=iovDisableIR -v TRUE
Which worked for about a week but now I have a system hanging again that I can't connect to. I will be looking at the logs later today but does anyone know if running the above command can lead to other issues?
I had a very similiar issue to what that you encountered but I'm using a GA-990FXA-UD3 Board w/ FX8130 Processor. I had 4 Nic cards (2-e1000e,1-e1000, and the on board Realtek) I'd experience lockups just like you are. I was able to induce the lockups by having multiple VMs transfer large files 2+ GB at the same time. I was able to trace the problem to the PCI slot on my board, as soon as I removed the Intel e1000 card the system has been rock solid for over a month. I reintroduced the e1000 last night and I had 2 lock ups and once the system just powered off completely on me. I removed the card again and its running like a champ. I hope you're able to isolate you issue, I know how fustrating it is to have these types of lockups so I wanted to lend my perspective.
In my case its either a bad PCI slot, IRQ sharing issue, Esxi configuration issues running both e1000 and e1000e in the same machine, or bad card (Don't think its a card I tried 2 different intel cards and the system crashed with either in the PCI slot.
Were you receiving any specific errors in the log files that pointed you to the NIC cards?