Unraid crashes every couple of days

[vincentvriends]

I made a new unraid server with an MSI B460 Torpedo and i3 10th gen CPU running unraid 6.10.0-rc2... but the server crashes every couple of days. Due to this issue I logged the syslog to flash but I don't see anything wrong with it.

What can cause the freezing of the server every couple of days?

Attached the diagnostics and syslog file. The last crash occured at 12 february at 1h00.

 

Thanks

nas-diagnostics-20220212-0123.zip syslog.txt

[dalben]

I was having the same problem with my server.  The advice I got was to remove the CoreFreq plugin.  After removing CoreFreq the server hasn't crashed for about 2 months now.

 

plugin: installing: /boot/config/plugins/corefreq.plg

 

 

Edited February 12, 2022 by dalben

[JorgeB]

Could also be this issue:

 

https://forums.unraid.net/bug-reports/prereleases/69x-610x-intel-i915-module-causing-system-hangs-with-no-report-in-syslog-r1674/?do=getNewComment&d=2&id=1674

 

[vincentvriends]

10 hours ago, JorgeB said:

Could also be this issue:

 

https://forums.unraid.net/bug-reports/prereleases/69x-610x-intel-i915-module-causing-system-hangs-with-no-report-in-syslog-r1674/?do=getNewComment&d=2&id=1674

 

Is there a fix available for that issue? Can I just block the i915 driver?

[vincentvriends]

17 hours ago, dalben said:

I was having the same problem with my server.  The advice I got was to remove the CoreFreq plugin.  After removing CoreFreq the server hasn't crashed for about 2 months now.

 

plugin: installing: /boot/config/plugins/corefreq.plg

 

 

Thanks, I will remove that plugin and try again.

[JorgeB]

17 hours ago, vincentvriends said:

Is there a fix available for that issue? Can I just block the i915 driver?

I don't use that but all the info should be in the above linked thread, or ask there if there are any doubts.

[CyrIng]

@vincentvriends Hello,

 

Was it a hard crash, CPU frozen, or you were still able to reach a shell ?

 

Because in your kernel log, beside this original successful startup message:

 

CoreFreq(2:6): Processor [ 06_A5] Architecture [Comet Lake] SMT [8/8]

 

I don't find the evidence of a Processor crash (usually, displayed with registers trapped and execution callflow)

 

Version 1.88 is by the way outdated, I recommend you switch to the latest 1.89.3 version. See with the unRAID plugin author for availability.

 

If CoreFreq is indeed the root cause, I will appreciate you create an issue with the  necessary detail on my GitHub project at:

 

github.com/cyring/CoreFreq/issues

 

Best regards

CyrIng

 

[vincentvriends]

1 hour ago, CyrIng said:

@vincentvriends Hello,

 

Was it a hard crash, CPU frozen, or you were still able to reach a shell ?

 

Because in your kernel log, beside this original successful startup message:

 

CoreFreq(2:6): Processor [ 06_A5] Architecture [Comet Lake] SMT [8/8]

 

I don't find the evidence of a Processor crash (usually, displayed with registers trapped and execution callflow)

 

Version 1.88 is by the way outdated, I recommend you switch to the latest 1.89.3 version. See with the unRAID plugin author for availability.

 

If CoreFreq is indeed the root cause, I will appreciate you create an issue with the  necessary detail on my GitHub project at:

 

github.com/cyring/CoreFreq/issues

 

Best regards

CyrIng

 

I think it was a hard crash, the GUI was unreachable and when I connected a screen to the server there was no output at all.

 

I now uninstalled the CoreFreq plugin to see if that fixes the issue, it hasn’t crashed so far but sometimes it took a couple of days after the reboot to crash again, so we will see…

[CyrIng]

1 hour ago, vincentvriends said:

I think it was a hard crash, the GUI was unreachable and when I connected a screen to the server there was no output at all.

 

I now uninstalled the CoreFreq plugin to see if that fixes the issue, it hasn’t crashed so far but sometimes it took a couple of days after the reboot to crash again, so we will see…

Thanks a lot for your reply.

 

If feasible, you can boot my CoreFreq live ISO image at: cyring.fr

 

Release is 1 89.3

Development is 1.90.0

 

Those images are ArchLinux based with ideal kernel boot options (see project Readme.md).

 

CoreFreq build and startup are fully automated.

Those ISOs are huge to download, about 1 GB. 

 

if your Processor is stable with those ISO then your Processor is supported by CoreFreq.

Edited February 13, 2022 by CyrIng