Machine Check Events detected on your server - how to get deatails

[matty2k]

Dear team,

 

last night I encountered a sudden server reboot/crash.

The log and also fix common problems show "Machine Check Events detected on your server".

Also some warning about: mcelog: failed to prefill DIMM database from DMI data.

I have attached the diagnostics. Is there some serious hardware error I need to worry about? How to get more details?

Server was running nearly 2 weeks on 6.10RC2.

 

kind regards

nasa-diagnostics-20211119-0754.zip

[matty2k]

Got one unexpected reboot/crash again.

3 days later on 23rd of November.

Strangely nearly the same time 02:07.

Last time Nov 19 02:03.

How to figure out?

regards

[matty2k]

nasa-diagnostics-20211126-1050.zipAgain, today 26th of November I have a new entry in the log:

02:00:14 NASA kernel: mce: [Hardware Error]: Machine check events logged

very similar timestamp. how to figure out details? already installed mcelog and activated syslogserver. but still no clue.

 

Edited November 26, 2021 by matty2k

[Squid]

Nov 24 15:44:09 NASA mcelog: failed to prefill DIMM database from DMI data
Nov 24 15:44:09 NASA mcelog: Kernel does not support page offline interface

Never seen it before, but it looks to me like a FYI rather than anything else.  (Although you might look for BIOS updates) Doubt it has to do with your crashing.  Run memtest from the boot menu for a minimum of a pass or 2

[matty2k]

OK. Did run memtest86 latest version 9.3 two times with 4 passes for all 13 different RAM tests with finally 0 failures.

I also checked for new bios upgrade but I am on the latest.

Again I would like to mention that the problem occurred with 6.10RC2 on RC1 the system was stable more than 14 days.

 

best regards.

[matty2k]

Meanwhile I tried to investigate a little more.

Regarding the time (see above) is when Jellyfin does the library scan and extracting covers (both scheduled at 02:00). Perhaps the crash / mce error is linked to this activity.

 

Furthermore, mce seems to be outdated. I now get the note to update to newer version. I am on RC2 actually. Since 14 days the server is running without crash/reboot now.

 

kind regards