PowerEdge R820 - Trying to identify hardware failure

Question

I have a Dell PE R820 that's been running VMware ESX for years without issue.  Over the last two days it has purple screened twice.  I ran the onboard disgnostics from the Lifecycle controller but they came back clean aside from commenting on the entrys in the IPMI log about the crashes, which were: 'Non-recoverable.  CPU Machine Chk: Processor sensor, transition to non-recoverable was asserted.' What kind of hardware failure does this typically point to?

DELL-Joey C · Answer

Hi,

Since if there is no hardware issues via the diagnostics, we might need to dig into logs to check for clues.

Try going into vmkernel logs to check if there is any traces of which it went into PSOD.

Usually VMWare go into PSOD with machine chk errors, it usually ran into low memory conditions.

Also, if you have a chance, disable C-State if you haven't. It's at the BIOS power profile section.

Setis_Dagre · Answer

I ended up updating all the firmware via the machine ISO (was too old to do via Lifecycle Controller) and so far it hasn't crashed since. When I looked at the vmkernel logs after reboot, nothing stood out as specifying what caused the purple screen.

PowerEdge Hardware General

Was this post helpful?