We recently updated our Jellyfin server to version 10.11.0, and we’ve noticed that the initial media scan in Symphonium (approximately 55,000 tracks, a mix of Opus and MP3) is significantly slower than it was a few weeks ago. It’s so slow, in fact, that after running it overnight, it didn’t even reach 50% completion. This is surprising, as the original scan within Jellyfin itself, when the library was first added, was much faster than what Symphonium is currently doing. We’re wondering if Symphonium might be having difficulty with how the new Jellyfin version organizes media compared to the previous Jellyfin database structure.
Logs:
Upload description: janik
Additional information:
Reproduction steps:
Following a complete reinstallation of Jellyfin on our NAS and Symphonium on the mobile device (Samsung smartphone), and the subsequent addition of Jellyfin as the source, the initial scan is commencing but is experiencing significant slowness.
I knew that eventually Symfonium users would experience these issues too, thought it would take longer though
10.11 is just too slow, even if Symfonium doesn’t have to sync that often.
As a fellow client dev, my condolences Tolriq
I have updated to versions 1.11.1 and 1.11.2, and unfortunately, this issue continues to persist. I would prefer not to revert to an earlier version, so I will be using Finamp for the time being. However, I do find myself missing Symphonium whenever I use Finamp, and I sincerely hope this will be resolved in a future update as soon as possible.
You’re doing great work and I’m sorry you’re facing backlash because of this update!
Same on my pre-production jellyfin server with 100k songs, 1.11.3, still didn’t fix that, also it eat 80% of the CPU for 6h until it end it’s “Media Segment Scan” + “Media Library Scan”.
Sad
x.11.x versions are totally broken at this stage for music.
You can follow this thread until we find a working one:
@Tolriq Do you see from Symfonium side at all what the root cause is? I want to open a issue with Jellyfin to see if they can fix it but dont know what the actual issue is
No problem I thought maybe there was something obvious other than timeout/errors which you could see. Im sticking with 10.10.7 anyways until the day that “hopefully“ the backend gets implemented into the Open Subsonic API.
The normal requests timeouts or errors, Finamp also have reports and say don’t use 10.11