NVIDIA Tesla T4 GPU on Electric Eel 24.10.2 + Jellyfin

I have read so many articles on the NVIDIA GPU issues since the upgrade to Electric Eel. I still cannot get mine to work. Here is what I can tell you:

Dell R7425
2x AMD EPYC 7601
256GB of RAM
RAIDZ2 - 8 wide - 51.5TB Usable Space
Tesla T4 GPU
TrueNAS Scale Electric Eel 24.10.2

  1. I remember seeing the “install drivers” option in Apps → Configuration when I first upgraded. I clicked it and have not seen it again since.

  2. Confirmed the driver being used is vfio-pci and not nouveau

  3. Running nvidia-smi returns the failed to communicate message

  4. Originally, nvidia-smi was returning a not found then I ran the docker update command and restarted to get the communication error.

  5. I cannot see the GPU in the Jellyfin configuration

I am not clear where to go from here. Please help.

Sounds like you enabled GPU isolation in the advanced settings and you’re attempting to run a docker image of Jellyfin. Would that be correct?

If so, disable GPU isolation and reboot. You may need to reinstall the Nvidia driver by unchecking the box, clicking save, then checking it and clicking save again. Reboot again.

nvidia-smi should then be able to pickup the hardware as it’s not loading that vfio driver. You will then be able to passthrough the GPU to the docker container.

1 Like

The seems to have been part of the problem. I don’t even remember turning this on. I turned of GPU isolation and nvidia-smi is now returning as expected. However, when trying to add the GPU to Jellyfin it says unrecognized GPU and it errors when i try to save the configuration.

[EFAULT] Failed to render compose templates: Traceback (most recent call last): File “/usr/bin/apps_render_app”, line 33, in sys.exit(load_entry_point(‘apps-validation==0.1’, ‘console_scripts’, ‘apps_render_app’)()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File “/usr/lib/python3/dist-packages/catalog_templating/scripts/render_compose.py”, line 47, in main render_templates_from_path(args.path, args.values) File “/usr/lib/python3/dist-packages/catalog_templating/scripts/render_compose.py”, line 19, in render_templates_from_path rendered_data = render_templates( ^^^^^^^^^^^^^^^^^ File “/usr/lib/python3/dist-packages/catalog_templating/render.py”, line 36, in render_templates ).render({‘ix_lib’: template_libs, ‘values’: test_values}) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File “/usr/lib/python3/dist-packages/jinja2/environment.py”, line 1301, in render self.environment.handle_exception() File “/usr/lib/python3/dist-packages/jinja2/environment.py”, line 936, in handle_exception raise rewrite_traceback_stack(source=source) File “/mnt/.ix-apps/app_configs/jellyfin/versions/1.1.21/templates/docker-compose.yaml”, line 3, in top-level template code {% set c1 = tpl.add_container(values.consts.jellyfin_container_name, “image”) %} ^^^^^^^^^^^^^^^^^^^^^^^^^ File “/mnt/.ix-apps/app_configs/jellyfin/versions/1.1.21/templates/library/base_v2_1_16/render.py”, line 59, in add_container container = Container(self, name, image) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File “/mnt/.ix-apps/app_configs/jellyfin/versions/1.1.21/templates/library/base_v2_1_16/container.py”, line 94, in init self.deploy: Deploy = Deploy(self._render_instance) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File “/mnt/.ix-apps/app_configs/jellyfin/versions/1.1.21/templates/library/base_v2_1_16/deploy.py”, line 15, in init self.resources: Resources = Resources(self._render_instance) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File “/mnt/.ix-apps/app_configs/jellyfin/versions/1.1.21/templates/library/base_v2_1_16/resources.py”, line 24, in init self._auto_add_gpus_from_values() File “/mnt/.ix-apps/app_configs/jellyfin/versions/1.1.21/templates/library/base_v2_1_16/resources.py”, line 55, in _auto_add_gpus_from_values raise RenderError(f"Expected [uuid] to be set for GPU in slot [{pci}] in [nvidia_gpu_selection]") base_v2_1_16.error.RenderError: Expected [uuid] to be set for GPU in slot [0000:41:00.0] in [nvidia_gpu_selection]

There is a workaround for this in
https://www.truenas.com/docs/scale/24.10/gettingstarted/scalereleasenotes/#known-issues

1 Like

Thanks to everyone who helped me find a working solution. The GPU section in Jellyfin config still shows Unknown but it is checked and running. I enabled hardware encoding and will now confirm if that is working as expected.

Nice. :slight_smile:

Thanks! :heart: