I installed `nvidia-container-runtime` on an Ubuntu fork to enable hardware acceleration for Nextcloud running inside Docker containers. However, I'm facing issues and want to remove the NVIDIA runtime. I've modified the daemon.json file to switch to runc and even tried using youki, but the runtime hasn't changed. I attempted to pass the runtime directly to the container, yet it still behaves as if the NVIDIA runtime is active. I also checked the Docker systemd unit file, and it doesn't indicate any runtime changes. As a result, I'm unable to start my Nextcloud Docker containers due to this NVIDIA runtime issue. What am I missing?
4 Answers
It seems like you might be running into a runtime configuration issue. Can you provide the command you're using to start the Nextcloud container? Sometimes context settings can affect which runtime Docker is using.
Make sure to check that the command you're using is right. From what I can see, the error `unknown or invalid runtime name: nvidia` suggests Docker is still trying to use that NVIDIA runtime instead of the one you configured. Verify changes in your daemon.json are reflected properly!
It could be an issue with how you modified your Docker configuration. Make sure you restarted the Docker service after making changes to daemon.json. You can do this with `sudo systemctl restart docker`. This is often required for changes to take effect.
First, can you tell me how you're starting the Nextcloud container? Is your Docker client functioning properly? You can check the situation with `docker context show` to ensure everything is set right.
Related Questions
Can't Load PhpMyadmin On After Server Update
Redirect www to non-www in Apache Conf
How To Check If Your SSL Cert Is SHA 1
Windows TrackPad Gestures