2023-10-18, 06:13 PM
I have just installed Debian 12 on my Intel NUC 11 N5105 today, due to a upgrade on the internal SSD and have always had this issue using Intel Quick Sync, but QSV works great!
First I moved my docker folder to my new instance and ran it with docer-compose up -d (from inside the folder) and it launched fine and runs fine, but when it comes to transcoding im having a hard time understanding it. The reason is, if i reboot, then i need to run "sudo chmod -R 666 /dev/dri/renderD128"
Why do i always need to do that, to get access to use the GPU?
I have added both UID, user and jellyfin for render in /etc/group
In the docker-composer.yml, i've added this, for the GPU access:
devices:
- /dev/dridev/dri
But, when i restart, i need to run:
sudo chmod -R 666 /dev/dri/renderD128
I use UID 1000 and GID 1000, why it should use the primary user as i understand it..
First I moved my docker folder to my new instance and ran it with docer-compose up -d (from inside the folder) and it launched fine and runs fine, but when it comes to transcoding im having a hard time understanding it. The reason is, if i reboot, then i need to run "sudo chmod -R 666 /dev/dri/renderD128"
Why do i always need to do that, to get access to use the GPU?
I have added both UID, user and jellyfin for render in /etc/group
In the docker-composer.yml, i've added this, for the GPU access:
devices:
- /dev/dridev/dri
But, when i restart, i need to run:
sudo chmod -R 666 /dev/dri/renderD128
I use UID 1000 and GID 1000, why it should use the primary user as i understand it..