As always, please ensure you stop your Jellyfin server and take a full backup before upgrading!
Now, if only there was a simple, built-in way to backup/export and restore/import all settings and other data, so that all platforms could do this easily, without having to search the internet for which folders to back up...
FYI, this is the best we have atm (which is pretty terrible). Please correct me if there is a better way:
I'm surprised at the lack of enhancement request/PR addressing this.I really want to dust off my C# and try but I'm kinda scared that the reason it isn't yet a thing is because it's a mess to implement.
Based on some comments in recent PRs for requested features that seem to have gone nowhere, the devs are trying not to overly complicate the project at the moment with other people's code that they'd have to support, and instead leaving certain requests to be handled in some grand refactoring they're working on.
I believe they're suggesting just doing a full backup up of your system/Docker container. Which isn't ideal, but I think they're trusting people who can run a Jellyfin server to be able to use the scripts.
Sure. But what if Docker is not available on a machine? What if the import should happen on a Linux machine coming from Windows? What if I want to sync two installations on different OSs?
I know it's all doable, but not easy, let alone foolproof. It's so easy to install, but genuinely not easy to keep safe without tech knowledge.
Syncing two instances sounds like a fun challenge. I think there's some project to replicate an sqlite db over the network. Similarly, you could use ceph or other distributed storage for the media.
I built something like this for Nextcloud a few years back, fun times.
I run JF in a docker container, and although I don't have backups of my config files yet (because I don't really care about setting up from scratch if need be), it would be trivial to simply backup the mounted config volumes. Makes upgrading safe and easy, too.
That's probably how I would recommend going about this, personally.
Yes, it works that easy. I had to move hard drives, last time I did that without docker somehow it didn't recognize the library, might have been a mistake from my end though.
Now I did it again just a few weeks ago with a docker setup, all folders are on the hard drive. Could just mirror the drive, set it up at same mount point and there was no difference in the library, just worked.
If you run it on a container, it should be enough to just make a copy of the set up volumes, right? (with permissions and all the metadata kept of course)
Theoretically, support for that could be coming... Emby (where Jellyfin is based on) always used their own layer for interacting with a SQLite database. All that custom made logic is currently being swapped out for EF Core. EF Core is a DotNet library for interacting with databases and EFCore that also supports MySQL, PostgreSQL, SQL Server besides SQLite.
So my guess is that, once all that work is completed, support of other database can be added.
For a little bit of context. I am currently running Jellyfin on Btrfs and there is quite a performance impact due to CoW. If 2 clients decide to browse the libraries, both clients grind to a near standstill with regards to being able to see things. So I am following this work with quite some interest.
I am currently running Jellyfin on Btrfs and there is quite a performance impact due to CoW. If 2 clients decide to browse the libraries, both clients grind to a near standstill with regards to being able to see things.
CoW is not recommended for databases, all DB servers advise for turning it off for the actual database. You'll run into the same issue with a dedicated database if you leave CoW on I guess. You could also disable CoW for jellyfin's database right now and performance should increase.
I also follow the progress of a dedicated DB, but on the other hand I don't know how much sense it makes architecturally. The likeliness that you have multiple jellyfin server instances access the same database is low - after all, there is info very specific to the server in there like the file path. Just migration is already not easy, how likely is sharing the database live? And if each database is specific to an instance - why not use SQLite (like it's done right now) and allow for more specific parameter tuning, like used memory and the like?
You don’t actually need to run down first. Just a docker compose pull if you haven’t made any changes, then docker compose up -d will restart whatever needs to be restarted.
I wish there was some kind of place where we could croudsource impressions and fixes for new versions of docker images.
Manjaro does something like this for their releases. They also have a survey that indicates how well things went (although it's participant biased to some degree since folks who had a problem tend to vote more than people who didn't).
It would be amazing to be able to pop in and see that jellyfin had a couple of new releases, that one of them does much better than the others in terms of overall quality, and what kind of issues there are (and how to fix them).
Unfortunately I set mine up with synology’s own container manager (many regrets) and so far I’ve not been able to bump anything. Will have another go when I’m home but I may up up tearing it all down and starting again using docker compose. Seems like the better option
The latest version of Synology with the container manager allows you to update images from the registry and will restart the container for you.
But I still migrated to docker compose to enable hw transcoding with quicksync
That is also my understanding. But for whatever reason, after updating the image and restarting the container, Jellyfin is still reporting 10.8 on the admin dashboard 😢
I did not yet upgrade to the latest version, but to migrate to compose I only had to copy the volume paths and the environment variables from Synology.
I can share my compose yaml by the end of the day if you need.
Before I upgrade I will try putting the cache on a SSD instead, seems it can improve performance quite a bit
Not in any easily-readable format, no. You can look at PR’s and the like but that’s a bit of a pain. I imagine they’ll release something once they feature freeze in a week or so.
You can hide all of that on your sidebar customization settings, but yeah it’s annoying that it’s turned on by default. The Discover is occasionally useful, but I honestly use Overseerr for discoverability more than I use Plex’s built-in search.
My biggest complaint with Plex is the lack of support for .m3u8 playlists. I want to be able to give it a list of livestreams, and then tune into those via Plex. Plex obviously already has live-streaming support built in via their Plex channels, but they have actively worked against custom livestream playlists, (it used to be supported via an extension, but they removed extension support.)
The layout of Plex definitely fits my brain waaaaay better with respect to navigation. But I hardly use it because I keep running into playback stuttering which doesn’t happen on Infuse, which I point at Jellyfin in my Synology. Will give this version another try.
I was getting stuttering on my server the other day but i think it was caused by mismatched client/server versions. When i upgraded the server to the latest release it went away
Lacking a centralised server that even self-hosted instances must use to validate admins and will render your instance inaccessible if Plex's server goes down again?
Jellyfin server isn’t lacking, but the weak point is definitely client software, especially on iOS. Unfortunately they just don’t have the same kind of resources Plex do in that regard.
I run Plex and Jellyfin, with watched status between the two synchronised with jellyplex-watched.
Unless I’m missing something, you can’t have multiple user profiles.
I use Infuse on my Apple TV with Plex Media Server as the backend. I can work with that limitation on Apple TV because my wife and kids use the official Plex app while I use Infuse, but ideally I’d prefer if we all used Infuse.
Ahh yes multi user is unfortunately not an infuse feature. Swiftfin does however have multiple users on iOS and presumably tvOS but isn’t as polished as infuse.
LOL, same. I did a docker-compose pull and restarted, came here to look at the release notes, and almost started panicking whether I omitted some important upgrade steps. Turns out everything upgraded smoothly automatically though.
I got the surprise treatment.
Saw the announcement on the telegram channel (I use the app exclusively for notifications by my services) and noticed that my watchtower container already upgraded it :p
Cool. There's a db fail bug elsewhere that I didn't run into thankfully.
I had to switch from ConfusedPolarBear(archived on GitHub now) to this fork for intro skipping. Works great.
I'm still scanning for the new trickplay function but I've been hoping for something like that. It was disabled in the settings by default, plus you have to enable it on a per-library basis, and of course you can either wait for the schedule to start the scan, or start it yourself in scheduled tasks.
I guess not. To be fair, if I wanted intro skipping to be baked in to whatever I'm using, I'd pay $5 or $6/month for Plex or Emby. I paid for Plex Pass for years, but I'm switching away because I think Plex is getting too "commercial" if that makes sense.
Jellyfin is great so far, and if I have to spend 5 minutes installing an add-on to get intro skipping, that's fine with me.
Hoping that it gets included soon as the folks developing the Android TV client have said they don't want to support intro skipping until it's part of Jellyfin core. Until then the only option is auto skipping and I don't really want to do that.
I thought intro-skipping had been included in Jellyfin now but this doesn't seem to be the case (can't find it in release notes) … Thanks for the tip for the 10.9 plugin 😸
I'm actually pretty good at finding memory issues in .net programs, maybe I'll debug this a little on my setup if I get a moment. Dotnet trace was essential for this as it was a Linux arm device
jellyfin.org
Top