VMware Cloud Community
mortetnma
Contributor
Contributor
Jump to solution

esxi 3.5 purple screens / crashes

I have an esxi 3.5 that keeps "purple screening" / crashes. The server runs only a single VM and the "crash pattern" is: server runs for a few days and then crashes

The hardware is a HP proliant box.

This happened on both 3.5 U2 and 3.5 U3 (with all latest patched/upgrades).

The error patterns in the log (attached) does not seem to match those listed in KB: 1008130 (http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1008130)

Reply
0 Kudos
1 Solution

Accepted Solutions
Lightbulb
Virtuoso
Virtuoso
Jump to solution

Well you might want to swap up the memory on the system. Or at least run Memtest. HP also has a bootable Diagnostic CD that you could run against the system.

View solution in original post

Reply
0 Kudos
8 Replies
nick_couchman
Immortal
Immortal
Jump to solution

What's the message on the PSOD? The archive you attached is pretty big to be digging through to try to figure out where the error message is...if you could just post it, that would be great...

Reply
0 Kudos
mortetnma
Contributor
Contributor
Jump to solution

Hi Nick

I dug through the archives/logfiles but was unsure exactly where to look.

The exported diagnostic files has a "viclient-support" folder and a file. Extracted, the "[esxsupport-2009-01-29@12-24-33.tgz]" has a quite elaborate folder structure.

Can you point me en the right direction? I could then dig out the right info (as I have the times to a few of the crashes)

/Morten

Reply
0 Kudos
mortetnma
Contributor
Contributor
Jump to solution

Hi Nick

last night I again lost connection to the esxi server.Loggin on via iLO or directly on the console I have a freezed console (I'm not getting the PSOD).

The last logged error:

2009-01-31 15:08:41.781 System.AccessViolationException: Attempted to read or write protected memory. This is often an indication that other memory is corrupt.

at QuickMksAxLib.IQuickMksAxCtl.Disconnect()

at AxQuickMksAxLib.AxQuickMksAxCtl.Disconnect()

at VpxClient.MKSview.DetachMKS()

It's kind of strange. The last logged error indicates that I should have lost connection to the server on 31st Jan, however, my monitor did not notify me until 03-02-2009 02:20:09 via a "No answer ping test" on the ip address of the esxi server.

Reply
0 Kudos
Lightbulb
Virtuoso
Virtuoso
Jump to solution

Well you might want to swap up the memory on the system. Or at least run Memtest. HP also has a bootable Diagnostic CD that you could run against the system.

Reply
0 Kudos
bburtonpa
Contributor
Contributor
Jump to solution

Did you install the "HP" version or the vanilla version of esxi? I realize this doesn't identify your issue, but I found that using the vanilla version does not result in the psod, for me anyway...

Reply
0 Kudos
mortetnma
Contributor
Contributor
Jump to solution

Good question. I wasn't aware of different versions of esxi.

I installed the exsi server available from vmware's download page.

Reply
0 Kudos
Lightbulb
Virtuoso
Virtuoso
Jump to solution

That is the standard unmodified. Good point though as tyhe HP agents can cause alot of trouble.

Reply
0 Kudos
mortetnma
Contributor
Contributor
Jump to solution

Indeed it turned out to be faulty memory. Thank for your help

Reply
0 Kudos