Hardware: MSI MAG Z790 Tomahawk WiFi, 13900K, 64GB RAM

1. Symptoms:

   • GUI becomes unresponsive, SSH still works

   • rc.nginx restart hangs on "Stopping nchan processes"

   • pkill -9 nginx hangs — processes in D-state

   • Only recoverable via SysRq reboot

   • Deleting local-kv.db temporarily fixes but corruption returns

2. Kernel bug seen in dmesg: BUG: scheduling while atomic: swapper/2/0/0x00000000

Have been trouble shooting with the help of Claude, should have came here first but here I am now. I managed to get Diagnostics zip if that helps?

tower-diagnostics-20260619-2000.zip

Edited June 19Jun 19 by display
added zip

11 hours ago, display said:

13900K

This would be my main suspect, since it's one of the more affected models by the Intel 13/14 gen issues, lot of users with stability issues with this and similar CPUs, look for a BIOS updatye, it can sometimes help if the CPU is not too far gone.

1 hour ago, JorgeB said:

This would be my main suspect, since it's one of the more affected models by the Intel 13/14 gen issues, lot of users with stability issues with this and similar CPUs, look for a BIOS updatye, it can sometimes help if the CPU is not too far gone.

Hadn't even considered that, before having this server as headless I was keeping it updated, current version is BIOS dated: Thu 26 Sep 2024 12:00 AM but there is an update available. IK'm going to have to gran Keyboard Mouse and monitor and update as there is a new version available.

Can you confirm then you can't see any software cause for the issue?

I see multipal apps segfaulting, this typically is a hardware issue, RAM or CPU, in this case I would lean more to the CPU, there are some users on their 3rd and 4th CPUs.

Damn, my worst case scenario

For info. That firmware update I just did, dropped CPU temps from 60-70 down to 35-40. vCore from 1.4 to 1, power draw I screenshot because otherwise you wouldn’t believe me!!!!!!!!

needless to say I’m going to try and do a warranty claim!