Is Synology DS920+ good enough for hosting a Jellyfin server?

[RelatableQuery]

I've tried setting up Jellyfin multiple times in a docker container and it always went OK. But, I'd never fully get around to using it and then I'd come back a few weeks later and find that the jellyfin container is hogging a lot of the resources. As soon as I reboot and remove the container, the system seems to be running OK.

So, my my hardware the limitation here? I would imagine that Synology DS920+ should sufficient.

[TheDreadPirate]

How is it hogging resources? CPU or memory?

Which image are you using? Linuxserver or the official image?

[Efficient_Good_5784]

I also have a DS920+. If you're using the system for multiple things at the same time with Jellyfin, the CPU might be slow enough. Especially if Jellyfin is doing its first time scan of all the files. Unless you filed your NAS with all SSDs, keep in mind that everything will be running off of slow HDDs. Jellyfin needing to scan everything will further slow down how fast other programs can use the HDDs themselves. Also, if you haven't set HWA for trickplays and have those enabled, the CPU will indeed be under stress while it's working on that.

The unit being a NAS with NAS software, it also provides you with file upkeep tasks such as file scrubbing. If your NAS is running a scrub on the files on the HDDs, your whole system will also slow down to a crawl.

After you got your whole library scanned and all resource-intensive tasks are not running, the DS920+ will provide a relatively smooth experience.
Just make sure to also setup HWA for transcodes.

Finally, my piece of advice to speed up Jellyfin (if you haven't already) is to install a 2nd piece of RAM into the system. Synology sells you the unit with a single stick of RAM. Once you add another stick, you'll benefit from dual-channel RAM for faster memory bandwidth. When I added a 2nd stick to mine, my HWA transcodes speed jumped up by 15-20fps on average.

[RelatableQuery]

@TheDreadPirate,
It's the CPU that's getting bogged down, even without me actively using it.
I am using lscr.io/linuxserver/jellyfin:latest. Is there a difference between that and the official one?

@Efficient_Good_5784, I will definetly add more RAM to it. But, I have no interest in transcoding at at all. I only plan to use this from within my home network on devices that will be capable of playing videos directly.

[TheDreadPirate]

We'd need to see your logs to try to figure this out. I'm wondering if a scheduled job is running, like chapter image extraction, or keyframe extraction, etc.