Crash every time multiple streams are happening

[bconor33]

Hi all,

I'm running the linuxserver.io latest docker image for Jellyfin. Running a single stream works fine but when a second stream is opened, within 5 minutes the process will become unresponsive but keep running on :8096. 
Running a docker down doesn't release it, I have to go in manually and find the pid and kill the process to be able to restart. Looking in the Jellyfin and FFmpeg logs, nothing really jumps out.
This is the closest thread I've been able to find to my issue and this isn't solved: https://forum.jellyfin.org/t-jellyfin-st...ew-minutes

Here are some logs:
The crash happens after 14:58:07.055 +00:00 on the jellyfin log

  log_20240430.txt (Size: 199.41 KB / Downloads: 20)

  FFmpeg.Transcode-2024-04-30_14-55-10_ad12cff220d06bd4aa368ec9823172aa_ce5fbb7b.txt (Size: 136 KB / Downloads: 39)

  FFmpeg.DirectStream-2024-04-30_14-58-05_48e8ef7732ca336a107b644ea13c76fd_ac5c3797.txt (Size: 20 KB / Downloads: 32)

Let me know if there's anything else I can provide.
Thank you!

[TheDreadPirate]

Both of your ffmpeg logs have weird blocks of characters at the end. Are you running out of disk space? Transcodes are written to disk. The fact that only when multiple streams at the same time, many/all appear to be transcoded, results in crashes kind of points towards running out of disk space.

It looks like you may have thought of that since you moved transcodes from container path /config/transcodes to /transcodes. But are they on the same host volume?

The later crashes appear to be due to zombie jellyfin processes preventing new instances from starting.

[bconor33]

All the media and logs are on a drive with ~10TB of extra space.
I moved the transcode directory to a different drive (SSD) with about ~200GB free with to try and troubleshoot but didn't notice a difference.
Do you think attempting multiple direct streams instead would be a good troubleshooting step?

Thanks for the help

[TheDreadPirate]

Direct streams are also written to disk. How much memory does your system have? What are the full specs and can you describe the rest of your setup?

[bconor33]

I believe that everything's working now because of the new transcode folder. Previously the media and the transcode folder both lived on the same 16TB HDD, it's possible that the drive was simply too slow to support multiple streams.
The moving of the transcode folder also must not have completed before the last crash but I've been watching the disk sizes while running 5 transcodes currently and it's writing the transcodes to the SSD correctly now.

Thank you for the help!