How to troubleshoot ESXi purple screen of death

[vandreytrindade]

Hi!

Today an ESXi server of our company crashed again on the same week.

The server is a HP ProLiant DL380 Gen9.

We don't have VMWare support anymore.

My question is: What logs I need to check to find out what was the root problem for that PSD?

I have got some files from the server (sfcb-intelcim-zdump.001 and vmkernel-zdump.2), but I'm not able to read them.

[daphnissov]

You need VMware support. We don't analyze PSODs or crash dumps here. Start by patching your host to the latest 6.7 build.

[vandreytrindade]

Sorry, but where I asked someone to analyze the PSOD for me?

Again, my question is: What logs I need to check...

[continuum · Accepted solution]

When you want to troubleshoot a PSOD you read the screen and write down the line that has "exception in World ..." and enter that line into google.

Result in your case - see

VMware Knowledge Base

   ESXi 6.5 and 6.7 host fails with PSOD when IPV6 is globally disabled (2150794)

In ESXi 6.5 this problem is fixed - not so in 6.7 but the KB has a workaround.

---------------------------------------------------------------------------------------------------------

Was it helpful? Let us know by completing this short survey here.

[vandreytrindade]

Thanks continuum​ !

Indeed the IPv6 are disabled in all of our ESXi servers.

Funny because we have two servers with the same exact version of ESXi and hardware and only this server crashed twice.

Anyways. Thanks for your time and attention.

[continuum]

ESXi hosts are sneaky. The second one may just pretend to be ok but next time you look away may follow his brother.

Disabled IPv6 is the root cause of the PSOD but very likely other conditions  must match as well to cause the PSOD.

So I would recommend to apply the workaround to all of your hosts.