from Underwire@lemmy.world to jellyfin@lemmy.ml on 18 Jul 2025 06:21
https://lemmy.world/post/33126188
Hi
I started setting up a Jellyfin server for friends to allow them watch remotely after the recent changes from Plex.
The experience so far isn’t great. Jellyfin seems to have many advanced features than Plex but the essential ones do not work that well.
The library scan took 3 hours for around 6k files. Plex takes much less time than that. There are many people raising this issue for years now but it is still not fixed.
From the logs I see Jellyfin trying to extract images for chapters even though the option isn’t checked. I already found reports for such issues but they are closed without an answer.
The big problem for me is Jellyfin not starting and getting stuck (without meaningful logs) at startup each time the server is shutdown. I did the reinstall 5 times now and trying the native install now. From the debug logs it seems like Jellyfin is stuck on one sql query.
From what I have seen, many issues are getting closed because of non activity even when the issue is still there.
I know that the project is done by volunteers but I was just wondering whatever I should invest more time on trying to resolve the issues. Maybe my server specs are just not ideal for Jellyfin.
threaded - newest
What are your server specs?
The server has AMD Athlon II 2 220 Processor, 4GB of RAM.
But Plex is running on it without any issues and has no issues doing one transcode with the faster profile. I even ran plex on specs lower than this and was working without issues except of course for the transcoding.
How have you set up Jellyfin?
I started with docker install, first using the LinuxSever image and then the official image.
Now I tried the native install and will see tomorrow if it is still OK.
I am using Ubuntu.
Can you post your docker-compose.yaml so I can compare it to mine please. Make sure you obfuscate your IP addresses and ports.
It is really a standard config
services: jellyfin: image: jellyfin/jellyfin:10.10.7 container_name: jellyfin user: 30000:30002 environment: - PUID=30000 - PGID=30002 - TZ=Etc/UTC volumes: - /media/WD/tools/Jellyfin/config:/config - /media:/media ports: - 8096:8096 restart: unless-stoppedWhere does that
user: 30000:30002come from?Also that PUID/PGID seems wrong for a standard config.
Here’s mine:
jellyfin: image: jellyfin/jellyfin container_name: jellyfin environment: - PUID=1000 - PGID=1000 - TZ=Europe/London volumes: - /opt/jellyfin/config:/config - films:/data/movies - tv:/data/tvshows - jellyfinmeta:/config/metadata ports: - 8096:8096 restart: unless-stoppedTry dropping the user and changing your PUID and PGID back to 1000 and see how you get on.
user and PUID, PGID do the same thing and not all images support PUID and PGID. They tell docker to run the image using the user with those id. My user has ID 30000 and group 30002. This allows me to change files on my library with this user without causing permissions issues for jellyfin.
I would start the investigation there.
I’m running my Jellyfin on a Raspberry Pi, and it runs like a dream, so there’s definitely a config issue afoot.
Investigate what? That is a standard config and didn’t change anything. Your Raspberry Pi is maybe more powerful than my server. How big is your library?
Athlon II? What is this? A server for ants? ;-)
That seems quite underpowered to be honest. Jellyfin per default does quite a lot of things per default during startup or as a background job. Sometimes you need to dig deep into the config settings and also disable certain timed jobs.
What OS are you running on?
Yes it a low spec server but like I said Plex is running on it without any issues and it can even transcode.
I am using Ubuntu the latest LTS.
Why do you think they do not?
If you would look up what they are actually doing, you woulf realize that a lot of work is done to improve the underlying quality of code to make it easier to do major changes to core functionality. Quick and dirty fixes by the previously project, emby, has led to a very shitty code base that makes changes hard.
I thought that the specs are not ideal for Jellyfin as it get stuck at startup, and it takes 3h to do a full scan. My server has an AMD Athlon II 2 220 Processor, 4 GB of RAM, so I thought that maybe Jellyfin need more.
I saw that they are working on big refactoring to use EFCore instead of doing direct SQL queries. I actually was surprised when they were saying that the migration will take days for some, and you shouldn’t interrupt it.
Jellyfin has many advanced features compared to Plex, but the experience isn’t great for many. For example, Plex and Jellyfin allow you to match/identify your media. Jellyfin allow searching by IMDB/TMDB/… IDs, but Plex doesn’t (possible by hidden way). But Jellyfin doesn’t autofill the search and year fields, and Plex does. When you select a result, Plex instantly show the fetched details and does the rest in the background. Jellyfin will start rescanning the whole media or episodes, and you get a spinner until it finished checking all media/episodes (minutes later for me for some shows with many seasons). They are really small details but make the user experience not ideal.
That you should not interrupt a database migration is really standard procedure. If it takes days is unfortunate, but what should the devs do? Create a migration process with weeks and months of testing that can recover after a interruption, for those 3 ppl that run on slow hardware?
Pls do not get me wrong, that the database and everything related to it is slow and basically legacy code is not good, but exactly that is beeing worked on right now, instead of continuously pumping out new features. Complaining about the exact thing that is currently in the works feels very disingenuous.