2024-05-18, 08:15 PM
Backup your data, uninstall + reinstall server and web. Make sure your repos are pointed properly.
Or...try to learn/use Docker. I'm lazy and created an alias of 'dcref' (docker compose refresh) that updates all my containers or 'dcrefs' for a single container. It essentially runs:
This pulls the new image, stops all containers, removes them, brings up my Pihole (required for connectivity), then brings up the rest of the project and removes any orphan containers. Standard practice for me when upgrading is:
Pulls new update, stops and removes the old container, brings it back up, and runs a tail on the container logs so I can watch for any errors or issues I need to fix.
No futzing with dependencies, no lost data, easy access to logs, persistent data, way easy to update or revert (as long as you don't engage in breaking changes like 10.8.13 --> 10.9 which are irreversible), easy to monitor, and easy to segment or swarm if you have some weird need for orchestration on your home server.
Running any software like Jellyfin requires *some* amount of research and troubleshooting, but Linux presents its own issues...frequently.
Or...try to learn/use Docker. I'm lazy and created an alias of 'dcref' (docker compose refresh) that updates all my containers or 'dcrefs' for a single container. It essentially runs:
Code:
docker compose pull && docker compose down && docker compose rm [all] && docker compose up -d pihole && docker compose up -d --remove-orphans
This pulls the new image, stops all containers, removes them, brings up my Pihole (required for connectivity), then brings up the rest of the project and removes any orphan containers. Standard practice for me when upgrading is:
Code:
dc pull [container] && dc stop [container] | dc down && dc rm -f [container] && dc up -d [container] && dc logs -tf [container]
Pulls new update, stops and removes the old container, brings it back up, and runs a tail on the container logs so I can watch for any errors or issues I need to fix.
No futzing with dependencies, no lost data, easy access to logs, persistent data, way easy to update or revert (as long as you don't engage in breaking changes like 10.8.13 --> 10.9 which are irreversible), easy to monitor, and easy to segment or swarm if you have some weird need for orchestration on your home server.
Running any software like Jellyfin requires *some* amount of research and troubleshooting, but Linux presents its own issues...frequently.
Jellyfin 10.10.0 LSIO Docker | Ubuntu 24.04 LTS | i7-13700K | Arc A380 6 GB | 64 GB RAM | 79 TB Storage