Hello - Recently my unraid server crashes every 3-4 days. I turned on syslog logging and this is what it recorded when it crashed.

I've attached the syslog output. Any ideas as to what this all means. Looking at it, it may have something to do with KVM and CPU 29?

 

Running unraid 6.12.10

syslog-172.19.101.9 (1).txt

Edited July 7, 20241 yr by KnightOwl1000
Attached syslog instead of having it in the body

Does it still crash if you leave the VM service disabled?

Hi @JorgeB, I have not tried doing that. Went to try doing that and stopped my VMs gracefully, and right after doing that, unraid crashed again. Attached is the syslog from today. It looks like there was another stack trace 10 hours ago as well that did not result in a crash.

 

Since it just happened, I decided to check a few things. UI is not responding, ssh is responding but takes over 2 minutes to login, accept any command. File share still works, containers appear to still be running/reachable (containers running still is interesting - they usually stop which is what gives me an indication that it's crashed - could be timing). Direct console access to the server is just as slow as SSH. 

 

Parity check was running right before the crash as well. 

 

syslog-jul-7.txt

Edited July 7, 20241 yr by KnightOwl1000
adding more information

Most call traces are still kvm related, also are you using ipvlan or macvlan?

I have docker setup to use macvlan. I did downgrade after the crash to 6.12.9 - trying to see if it's an issue with 6.12.10. I think it started after the upgrade. 

It's also worth changing to ipvlan to rule that out, one of the call traces may have been macvlan related.

It's been nearly a week with no crashes since I downgraded to 6.12.9 - so it seems there's a bug 6.12.10 that was causing the crashes...

6 hours ago, KnightOwl1000 said:

so it seems there's a bug 6.12.10

Not aware of any other reports, so I think that's unlikely, but try 6.12.11 once it's out, it should be released this week.