GPU Disconnects Over Time

mikewaggs

New Member
Sep 25, 2024
1
0
1
Hello All,

I'm running into a strange issue between my Proxmox and Ubuntu VM setup. I'm not sure if the issue lies with Proxmox or Ubuntu.

After about a week of usage my passed through NVIDIA card (3060) will stop working within the Ubuntu VM.

A few things I've noticed:
  • dmesg does not populate with anything while it has stopped working.
  • nvidia-smi still shows the card as normal.
I did think that this was just a Docker/GPU issue at first, but with dmesg not populating any results it has pushed me more towards Proxmox/Ubuntu. The journal has no error messages for the time that it disconnected.

The issue can be remedied by restarting the VM. It's an easy fix once I notice, but I'd rather not have to go and restart all of my services once a week. Outside of this, the GPU works with absolutely no problem.

It's a little tricky to troubleshoot since I have to wait about a week at a time to see if something has worked.

Thank you,
Mike