Hi all! First time using the forum but I have been running into an issue with my new Unraid server over the past month or so.

At random times the web GUI will become unresponsive. I am no longer able to access it from any device. If I plug into the server directly I am greeted by a black screen with a static cursor in the top left corner that cannot be interacted with. I am also unable to SSH into the server when this happens. Tailscale still shows the server as active but I am unable to connect through that either. I am able to ping the server and get a response. The physical power button also will not power off the server normally when this occurs. I can press and hold the physical button for 10 seconds to shutdown the server in an unclean manner.

Curiously some of my docker containers continue to work completely normally. For example, Jellyfin can still be accessed and used as if nothing has occurred.

Some actions I have taken so far in no particular order:

1. memtest86 - passed

2. BIOS update

3. set memory limits on all docker containers

4. pinned all processes to not utilize cpu 0 besides the server

5. increased docker image size

6. turned off c states in BIOS

7. Tried new USB slot for USB drive

8. Uninstalled Uconnect

My system is an old HP Z2 G4 tower with an I5 - 8500 and 16 GB of RAM. I have one cache 500GB drive and a couple HDDs. I did turn on sys logging to the cache. Attached are my diagnostics files and sys log.

nelson-diagnostics-20251001-1134.zip nelson-syslog-192.168.0.48-20251001-1113.zip

Nothing relevant was logged in the last day, there is an OOM error caused by a Docker container. It was 2 or 3 days ago, so could be unrelated, but it should still be resolved.

2 hours ago, JorgeB said:

Nothing relevant was logged in the last day, there is an OOM error caused by a Docker container. It was 2 or 3 days ago, so could be unrelated, but it should still be resolved.

How do I resolve the OOM error? I can see it but am unable to determine which container caused it.

The latest crash happened between 3-4am on 1 Oct based on the mover executing at 3am but not at 4am.

I can't tell which container, failing a better option, reboot and only start half the containers and retest, if no issues, try the other half, then keep drilling down.