Server Version#: 1.14.1.5488 (Docker)
Player Version#: 3.83.1 [Firefox - Plex Web, Original Quality]
Issue: Attempting to play files in Plex Web result in no playback and a endless spinning buffering screen. This only affects some files.
Now before you say it’s the file (!!!) about 90% of my files refuse to play in Plex Web but Plex Media Player has zero problems playing back any of them, even the affected files.
Console log says…
EAE timeout! EAE not running or wrong folder? Could not read /tmp/pms-etcetc [Transcoder] Error while decoding stream #0:1: Input/output error
Permissions of /tmp (plex runs as my user, subfolders inside /tmp/pms-etc" are owned by my user)
I’m getting the same errors on certain files with Dolby audio, multiple clients; Sheild, iOS, PlexWeb.
Error is:
ERROR [Transcoder] [eac3_eae @ 0x2888f00] error reading output
ERROR [Transcoder] Error while decoding stream #0:1: Input/output error
ERROR [Transcoder] [eac3_eae @ 0x2888f00] EAE timeout! EAE not running, or wrong folder? Could not read '/tmp/pms-4298f458-6976-409f-a693-dedb2f9bb08e/EasyAudioEncoder/Convert to WAV (to 8ch or less)/cd7a200bf5a863d4-com-plexapp-android_7859-0-10.wav'
No system changes or security modifications to folders. Possibly just a docker Plex image update (can’t be certain)
======================================== Ubuntu 18.04 64bit / Linux 4.15.0-43-generic Plex Version 3.83.1 (docker)
Practically same setup as OP
Symptom is whenever offending files are played the client just spins the hourglass (plex circle) and nothing happens. Logs fill with above error.
Direct-play works fine. Only happens on transcode.
I have the same issue as OP. Anything Dobly does not play. The only way it would play is if I enable the experimental player on the iPhone. In our case the issue not related to max_user_watches as that was increased and did not solve the issue.
One other note is I moved Plex to another server which seemed to have triggered the issue. The files still play fine on the previous server. Both servers are playing the same identical file. The new server triggers the issue whereas the old server does not. Both servers are running v1.14.1.5488.
All directory permissions are the same. The only difference I see so far is (old) has ssd drives and (new) has nvme.
Edit: After doing some deeper looking. It seems like the /transcode dir was loosing permission somehow. /dev/shm was being used as the transcode dir. Not sure why that was happening but now know what was causing it.