Hi,

After upgrading to 7.3.0 last week I've been experiencing daily complete freezes of the server, in which I can't SSH in, access the web UI, or even log in directly on the box. I have to hard reset and the server works okay again for a few hours.

It seems to be some kind of corruption that occurs in docker.img each time. I never experienced any of these issues before upgrading to 7.3.0 and switching to internal boot.

```~# btrfs device stats -z /var/lib/docker

[/dev/loop2].write_io_errs 0

[/dev/loop2].read_io_errs 0

[/dev/loop2].flush_io_errs 0

[/dev/loop2].corruption_errs 11

[/dev/loop2].generation_errs 0

```

Things I've tried to no avail:

• memtest (no errors after 6 passes)

• updated BIOS

• fully deleted docker.img and recreated containers from scratch

ckmedianode-diagnostics-20260523-0829.zip

Edited May 22May 22 by gaffes

Enable the syslog server and post that fater the next crash.

I swapped out the RAM with sticks from another PC to fully rule that out. After 12 hours or so the errors started happening again, I see a log of errors like "kernel: python3[934714]: segfault" in the syslog. I've attached an excerpt.

When this starts happening I start to get reports that my plex server is unavailable. In this latest case I was able to log in and attempt a clean reboot, but it seemed like it was never able to fully stop the services, and I had to hard reset again.

syslog.txt

If it's only python that keeps segfaulting, it could be an issue with a specific container or plugin. Try booting in safe mode, leave Docker disabled, and retest, if OK, start enabling the container one by one, or a few at a time.

I've done the following test: stopped all docker containers and ran stress-ng for 2 hours to see if I could trigger any issues. No issues seemed to occur.

I then re-enabled only the Plex docker container. Within 24 minutes, the Unraid GUI became unreachable, and I saw a bunch of kernel errors in syslog. I've attached that log.

syslog2.txt

That suggests an issue with Plex. Recommend looking for its support thread; there can be some config or known issue.

I managed to track this down - it was a bad nvme drive that I installed to move to internal boot. I switched back to booting off the USB and the server has been stable with no errors in the syslog.

When I get a chance I'll try doing internal boot off a partition on the existing nvme drive I have that I know is good.

Weird how these issues manifested.