Hello everybody,

 

I encountered some random kernel panic on my unraid 6.10.3 (after the update from 6.9.2)

Here some syslog :

 

Quote

Aug 3 12:19:05 192.168.5.5 Aug 3 14:19:05 OCTOPUS kernel: smartctl[20116]: segfault at 18 ip 000015263e808cbf sp 00007ffe4264ea90 error 4 in libc-2.31.so[15263e793000+178000]

Anybody knows this syslog error?  FYI, I'm on a ryzen 3700x base (C-state disabled and dual chan RAM speed to 2666 mhz (following the documentation)

I will try a memtest tonight during 24h to 48h... but I'm lost

 

Here my diag in this post.

 

If anybody has any informations, I don't know how to debug this issue.

 

best regards

 

octopus-diagnostics-20220803-2044.zip

No segfaults in the diags posted, run memtest and see if you get anymore.

Hello, thank you for your answer, memtest in progress

Hello, memtest during 24 hours, memories are clean

 

I also flash my BIOS to the latest version (Asrock X570 Taichi from 3.6 to 4.8)

Unraid uptime 8hours but I still have kernel logs panic (not kernel panic yet)

You can see stacktrace in the diag report...

 

Quote

Aug  5 02:00:44 OCTOPUS kernel: BUG: unable to handle page fault for address: ffffffff811c7763
Aug  5 02:00:44 OCTOPUS kernel: #PF: supervisor write access in kernel mode
Aug  5 02:00:44 OCTOPUS kernel: #PF: error_code(0x0003) - permissions violation
Aug  5 02:00:44 OCTOPUS kernel: BUG: unable to handle page fault for address: ffffffff811c7763
Aug  5 02:00:44 OCTOPUS kernel: #PF: supervisor write access in kernel mode
Aug  5 02:00:44 OCTOPUS kernel: #PF: error_code(0x0003) - permissions violation
Aug  5 02:00:44 OCTOPUS kernel: PGD 220e067 P4D 220e067 PUD 220f063 PMD 10001e1 
Aug  5 02:00:44 OCTOPUS kernel: Oops: 0003 [#1] SMP NOPTI

 

 

octopus-diagnostics-20220805-0752.zip

Try updating to v6.11.0-rc2, the much newer kernel might help if it's some compatibility issue.

I Still have kernel panic every 2 days... I update to 6.11.0-rc2 and 6.11.0-rc3, I change memories too, nothing change.

I will change the LSI card for a new one...

 

If I will still have Kernel Panic, I will abandon Unraid.. too much instability

Edited August 14, 20223 yr by softdrinker

Enable the syslog server and post that after a kernel panic.

After enable syslog server, 2 differents kernel panic with unraid 6.11.0-rc3.

 Do I have to rollback to 6.10 unraid version?

 

 

1syslog_kp.log 2syslog_kp.log

That looks more hardware related, especially if it happens with both v6.10 and v6.11, but try going back to v6.9 to confirm if it's still stable there.

Hello,

 

Just disabled AMD fTPM in BIOS of my MB Asrock X570 taichi !

Looks like I don't have kernel stack trace anymore in syslog. 24hours without kernel panic.

 

I keep in touch if I have something new !