Need help interpreting error

TrueNAS General
1

Hi All,

Ive been having recurring issues. Ive recently started using TrueNas and currently am running immich and have an smb storage for file sharing.

I kept getting quota exceeded error and looking at the logs i see:

Truenas kernal: python (165364)" segfault at 0 ip 00007fac4a106564 sp 00007fac4a1065064 error4 in libonnxruntime_providor_cuda.so[333564.7fac49e16000+ac2000] likely on cpu 0 (core 0. Socket 0)

I had assumed this was something to do with my nvidia gpu that was bieng used by immich but after removing the nvidia driver and reinstalling immich so it won’t use nvidia gpu im still getting the error

Just looking to understand this error and any tips on troubleshooting.

My config is;

Amd 1300x cpu

Nvidia gtx1070 gpu

8gb ram

2

For the “cores” dataset?

You probably have too many core dumps. You can delete them and see if they start to populate the “cores” dataset again.

The location of these core dumps are under /var/db/system/cores, unless SCALE/CE uses a different path.

You can find out with:

mount | grep cores
3

8GB of RAM is the bare min for TrueNAS. I don’t know if I would even attempt Immich. Requirements | Immich

4

RAM: Minimum 4GB, recommended 6GB. I thought 8gb would be enough.

5

It might work okay but the System Requirements for TrueNAS with ZFS are 8GB min, 16Gb recommended. The Immich requirements may be just for the container or VM alone. See the Memory Sizing section in the Hardware Guide on sizing suggestions.

1 Like
6

You probably have too many core dumps. You can delete them and see if they start to populate the “cores” dataset again.

I navigated there through shell ls showed the folder was empty.

7

What is the exact error message about the exceeded quota?

What is the following output?

zfs list -r -t fs -o space,quota

I would run multiple memtest passes to rule out the RAM.

Here are eight examples of people whose issues arose from bad RAM.