2023-12-03, 03:56 AM
Just to preface, I'm very new to Jellyfin and NAS setups overall. This is my first time setting up and running anything more than a plex server on my personal computer. Sorry for the long post but I want to document the process of what I've done up till now.
I recently picked up a Synology 923+ and 4x4TB Seagate Barracuda (6Gb/s 5400 RPM 256MB Cache) HDDs during the black friday sales. For a while now I've been wanting to set up a NAS for my small discord server to host movie nights and provide general media access. I've spent the last few days getting it all up and running and finally got everything automated and running smoothly via Docker containers. Ombi for user requests. Sonarr/Radarr and my Usenet downloader/indexer/provider. Synology DSM based reverse proxy for outside access to Ombi and Jellyfin. All of this is currently functioning and running smoothly. Users can request media, and it is quickly retrieved and sent to Jellyfins library where users are able to access them for viewing, even a few users at a time.
Now, Thursday night, after getting this all going, I decided to put it to the test and host a movie night via SyncPlay, 8 users including me. Immediately there were issues, Jellyfin UI lagging for everyone and issues with getting everyone into the SyncPlay group, then once we were finally able to get everyone into the group, attempting to hit play brought everything to a halt. Fine, I didn't expect this to immediately work. Did some research and found that 1) transcoding can be an issue, especially with all of us on different browsers/apps and 2) the official jellyfin image provided to docker has optimization issues on DSM 7 and that the linuxserver version runs much better. I completely got rid of my Jellyfin container and remade it with the linuxserver version and the difference in responsiveness was immediately noticeable. Got it all up and running and connected to the automation and reverse proxy again and let everyone back in the next day(Friday).
This time, we tried again but with everyone using the Jellyfin desktop app to hopefully prevent transcoding. Everyone got in and joined the SyncPlay group with minimal issues. Hit play on the movie, and this time, rather than the server locking up, we made slightly more progress. After a couple minutes of loading, about half the users were able to view the movie normally and in sync. The other half were either stuck in the menu, on a loading screen, or with a black screen(one user had a black screen with synced audio to everyone). We waited about 10 minutes to see if anyone else would get the movie and unfortunately no luck. Checked the resource manager and cpu usage was pretty low (10% - 20%), but RAM usage was pushing 80%. Disk and network speeds were high but not high enough to seem like they would be the issue.
Did some more research today and looked into various upgrades for the 923+. Ram seems like the obvious one, I would probably go up to the max of 32gb. I also read about using an m.2 or attaching an external ssd, either as a cache or a separate volume to handle all of the docker containers and processes. Now before I go and buy anything else for this system my question is: Is this a problem that can be solved with these hardware upgrades, or is what I'm aiming for completely out of scope for the 923+. I've also read about people using a nuc to handle streaming, and solely using the 923+ as storage. I'd like to avoid this if possible but if its the only solution then I'm open to this. Has anyone ever tried SyncPlay with this many users? Or is this just completely a lost cause.
Sorry for the rambling post, and sorry for any missing details. Happy to provide any additional info to help solve this. Thanks!
I recently picked up a Synology 923+ and 4x4TB Seagate Barracuda (6Gb/s 5400 RPM 256MB Cache) HDDs during the black friday sales. For a while now I've been wanting to set up a NAS for my small discord server to host movie nights and provide general media access. I've spent the last few days getting it all up and running and finally got everything automated and running smoothly via Docker containers. Ombi for user requests. Sonarr/Radarr and my Usenet downloader/indexer/provider. Synology DSM based reverse proxy for outside access to Ombi and Jellyfin. All of this is currently functioning and running smoothly. Users can request media, and it is quickly retrieved and sent to Jellyfins library where users are able to access them for viewing, even a few users at a time.
Now, Thursday night, after getting this all going, I decided to put it to the test and host a movie night via SyncPlay, 8 users including me. Immediately there were issues, Jellyfin UI lagging for everyone and issues with getting everyone into the SyncPlay group, then once we were finally able to get everyone into the group, attempting to hit play brought everything to a halt. Fine, I didn't expect this to immediately work. Did some research and found that 1) transcoding can be an issue, especially with all of us on different browsers/apps and 2) the official jellyfin image provided to docker has optimization issues on DSM 7 and that the linuxserver version runs much better. I completely got rid of my Jellyfin container and remade it with the linuxserver version and the difference in responsiveness was immediately noticeable. Got it all up and running and connected to the automation and reverse proxy again and let everyone back in the next day(Friday).
This time, we tried again but with everyone using the Jellyfin desktop app to hopefully prevent transcoding. Everyone got in and joined the SyncPlay group with minimal issues. Hit play on the movie, and this time, rather than the server locking up, we made slightly more progress. After a couple minutes of loading, about half the users were able to view the movie normally and in sync. The other half were either stuck in the menu, on a loading screen, or with a black screen(one user had a black screen with synced audio to everyone). We waited about 10 minutes to see if anyone else would get the movie and unfortunately no luck. Checked the resource manager and cpu usage was pretty low (10% - 20%), but RAM usage was pushing 80%. Disk and network speeds were high but not high enough to seem like they would be the issue.
Did some more research today and looked into various upgrades for the 923+. Ram seems like the obvious one, I would probably go up to the max of 32gb. I also read about using an m.2 or attaching an external ssd, either as a cache or a separate volume to handle all of the docker containers and processes. Now before I go and buy anything else for this system my question is: Is this a problem that can be solved with these hardware upgrades, or is what I'm aiming for completely out of scope for the 923+. I've also read about people using a nuc to handle streaming, and solely using the 923+ as storage. I'd like to avoid this if possible but if its the only solution then I'm open to this. Has anyone ever tried SyncPlay with this many users? Or is this just completely a lost cause.
Sorry for the rambling post, and sorry for any missing details. Happy to provide any additional info to help solve this. Thanks!
![[Image: GitHub%20Sponsors-grey?logo=github]](https://img.shields.io/badge/GitHub%20Sponsors-grey?logo=github){kind=icon}