[6.9.1/6.9.2] Kernel panics - not syncing: fatal exception in interrupt

[sempervirens]

A few weeks ago I had a power outage, my UPS turned off my server cleanly and when I came back to turn it on I decided to upgrade to 6.9.2. It was up for a day or two, then crashed with a "Kernel panic - not syncing: Fatal exception in interrupt". I rebooted, and a few days later it crashed again. So I downgraded back to 6.9.1 which I had had multiple multi-month uptimes on, but the kernel panics are still coming (always the same, I've had 4 crashes total now).

 

I did 2 passes of Memtest after the 3rd crash and it found no errors.

 

I can still reach the server via IPMI, so I've attached a screenshot from the iKVM screen showing the panic; 'mirror syslog to flash' has been on so after rebooting I zipped the syslog and attached that. Diagnostics also attached in case it's helpful, but the log there is just of a new boot.

 

 

Edited September 23, 2021 by sempervirens
don't like leaving my logs out forever

[JorgeB]

Nothing much of relevance logged, there's a network related crash, so if you changed anything regarding dockers with custom IP address this might be worth checking:

 

See if this applies to you:

https://forums.unraid.net/topic/70529-650-call-traces-when-assigning-ip-address-to-docker-containers/

See also here:

https://forums.unraid.net/bug-reports/stable-releases/690691-kernel-panic-due-to-netfilter-nf_nat_setup_info-docker-static-ip-macvlan-r1356/

[sempervirens]

So I turned off PiHole-Template on the 9th, as it was my only docker using br0 as the network (everything else was bridge or host), and I've been up and stable since then. Appears to have been the culprit somehow (since it's been running for months, I actually hadn't changed anything with it at all).

I'll update this if something changes, but that seems to be the cause of the issue.