I keep getting Out of memory errors, and have had them intermittently since installing Unraid a year ago.

This results in the server being unresponsive, and not able to shutdown cleanly.

I have limited the memory on all of the containers, and it still happens.

It's getting to the stage where I'm considering moving back to Truenas, or even Windows.

Am I being ambitious with the number of containers I'm running?

Do I need more than 64GB of RAM for the task?

Please help me fix this once and for all.

TIA

mickey-diagnostics-20250823-1133.zip

What VMs are you running? Any media servers set to transcode to RAM?

There's a python process that seems to keep running away and get killed, you can tuse something like Beszel to track container RAM usage and figure out where that comes from.

50 minutes ago, AlanC said:

It's getting to the stage where I'm considering moving back to Truenas, or even Windows.

Different OS isn't gonna help you if an app has a problem.

A windows VM that only runs Veeam backup and a Home Assistant VM.

I thought it was the frigate container causing the problems using tmpfs, so I stopped using tmpfs, and began writing to an SSD instead.

I have run into the python issue before, but I can't track down where it's coming from.

I've installed Beszel, thanks for the heads up

On 8/23/2025 at 12:38 PM, Kilrah said:

There's a python process that seems to keep running away and get killed, you can use something like Beszel to track container RAM usage and figure out where that comes from.

Different OS isn't gonna help you if an app has a problem.

That's the annoying thing, why is it so difficult for a user to find the cause of a problem?

To say there's a python process causing a problem isn't much use when so many containers use python.

Hence why I mentioned a tool that will record CPU/RAM usage of each container and that you can go check after it happens.

Kinda comes with the job of server admin you're taking on when you build a server with a bunch of services instead of using a commercial service where you pay someone else to do it for you...

On 8/23/2025 at 7:47 AM, AlanC said:

I have limited the memory on all of the containers, and it still happens.

Your container is hitting the memory limit and being killed, and the system is reporting that

15 hours ago, AlanC said:

To say there's a python process causing a problem isn't much use when so many containers use python.

Runps -auxf > ps.txt to pipe it to a file (run within a share path so it will create the file there) that will show you what is launching the python process that's causing the issue.

On 8/26/2025 at 10:06 AM, Michael_P said: Your container is hitting the memory limit and being killed, and the system is reporting that Runps -auxf > ps.txt to pipe it to a file (run within a share path so it will create the file there) that will show you what is launching the python process that's causing the issue. Expand to Edit

I haven't had the problem for 2 weeks now, so haven't needed to run this command. The only thing different, is that I haven't run the ollama and open webui containers so I'm guessing the problem lies somewhere there. Thanks for the help

Edited September 6, 2025Sep 6 by AlanC