UCS c220 ESXi5.5 purple screen

[Todd MacQueen]

Hi all,

We've had a few crashes on this specific machine that we are hoping to get to the bottom of. It is running ESXi 5.5U3-3123032. We had the crashes when it was running 1331820 and hoped the update would resolve the problem.

The ESXi log had this: 

2017-03-09T09:07:25.966Z cpu0:12566047)VMware ESXi 5.5.0 [Releasebuild-3116895 x86_64]
LINT1/NMI (motherboard nonmaskable interrupt), undiagnosed. This may be a hardware problem; please contact your hardware vendor.2017-03-09T09:07:25.967Z cpu0:12566047)cr0=0x80050031 cr2=0x6407a9bd1a cr3=0x207c68e000 cr4=0x42668

The UCS log had these at the same time as the purple screen:

System Software event: Critical Event sensor, PCI SERR Bus: 4, Device: 4, Function: 0  was asserted

System Software event: Critical Event sensor, PCI SERR Bus: 4, Device: 2, Function: 0  was asserted

System Software event: Critical Event sensor, PCI SERR Bus: 3, Device: 0, Function: 0  was asserted

How do I trace the UCS events to the correct device?

[Kirk J]

Greetings.

Do you happen to have a Broadcom 5709 based PCI-E adapter installed?

Check the UCS interop matrix for your HUU/firmware level, and the esxi version, and make sure your BNX2 drivers are up to date.  Once you determine the BNX2 driver version from the interop matrix, you will want to search on the VMware website for that driver version, that applies to your esxi version.

Thanks,

Kirk...

[Todd MacQueen]

Yes. That adapter is installed. I will follow your suggestion. Thanks!