Why would TV episodes not play from a remote client?

General Discussions
1

I’m running into an issue with Plex of late that seems strange to me. Some TV episodes won’t play on a remote Plex client, but the same episode plays fine with the same client connected to my WiFi network. After a friend with iPad/iOS and Roku players indicated the episodes wouldn’t play, I’ve been able to duplicate the issue on those same episodes with my Android phone. At the same time, other episodes play without issue when I’m not connected to WiFi.

I’m running PMS v1.10.1.4602 as a plugin on my unRAID 6.4.1 server. My main player is PMP 2.4.1.787-54a020cd for WIndows 10. My Android is running Plex 6.13.6.3452 (d5f0a62c) and unless I’m connected to my local network via WiFi, some episodes that play fine on Windows 10 just will never connect on my Android.

This started happening early in 2018, but I can’t recall the sequence of the following events that may play a factor:

  • I installed a PCIE sata expansion card in my server (StarTech PEXSAT2) in addition to a 3TB WD Green HDD (disk space was in short supply!)
  • I started using Sonarr in a Docker container to obtain TV episodes
  • I had a power outage and my server/Plex did not shut down “nicely”. I’ve had several instances where PMP couldn’t display the On Deck dashboard; I restored an older index and that has generally been resolved
  • Since the power outage, I’ve discovered that my CPU runs at 100% even when idle (I’m not certain this was the case before the outage, but I know the fan speeds are at full speed/especially noisy) ***** EDIT - could the CPU running at 100% prevent remote players from connecting and playing the episodes??

I’ve looked a bit at the Plex logs (which is attached for anyone to see/analyze) and have noticed the following entries:
: ERROR - [FFMPEG] - Cannot load libcuda.so.1
: WARN - avcodec_open2 returned -1313558101 for encoder ‘h264_nvenc’
: ERROR - [FFMPEG] - No VA display found for device: /dev/dri/renderD128.
------> (repeats a few times)
: WARN - CodecManager: EAE exited with code 13. Restarting.
: INFO - CodecManager: starting EAE at /mnt/user/Apps/tmp/pms-3aacf811-f592-4172-b8e9-7ed0fd004025/EasyAudioEncoder
: ERROR - JobManager: child process returned: 13 (Permission denied)
: ERROR - [Transcoder] [eac3_eae @ 0x14774e0] EAE timeout! EAE not running, or wrong folder? Could not read ‘/mnt/user/Apps/tmp/pms-3aacf811-f592-4172-b8e9-7ed0fd004025/EasyAudioEncoder/Convert to WAV (to 8ch or less)/1c220fdb563222cc-com-plexapp-android_25120-1-0.wav’
: ERROR - [Transcoder] [eac3_eae @ 0x14774e0] error reading output
: ERROR - [Transcoder] Error while decoding stream #0:1: Input/output error
: ERROR - [Transcoder] [eac3_eae @ 0x14774e0] EAE timeout! EAE not running, or wrong folder? Could not read ‘/mnt/user/Apps/tmp/pms-3aacf811-f592-4172-b8e9-7ed0fd004025/EasyAudioEncoder/Convert to WAV (to 8ch or less)/1c220fdb563222cc-com-plexapp-android_25120-1-1.wav’
------> (repeats a several/many times)
: WARN - Transcode runner appears to have died.
: WARN - PubSub: Received notifyConnectivity event with incorrect async identifier (9471c186-4139-44f1-80ba-40a9365375c5, expected f5f7d33f-1e93-44bd-89c3-1de079473b63)

Nothing in there tells me anything, but I hope someone can provide some assistance as to what the issue might be.

2

seems like an issue with transcoder PMP locally can play almost anything so doesn’t need server to transcode for compatibility or network speed

I would see if you can optimize database and if there is any corruption https://support.plex.tv/articles/201100678-repair-a-corrupt-database/

3

I did previously restore the database with one from prior to the power outage, but the problem persists. I may wipe the entire library and recreate it just to see if that makes a difference. Thanks for the input; I’m away for the weekend, but hopefully I can give this a try early next week.

4

I would just reinstall the server first. It might be something with the app/transcoder itself getting corrupted during power outage.

I am not an Unraid user so can’t provide specifics but if you save your PMS folder and preference.xml and just reinstall PMS then it might be good.

if you do not have a large library with lots of customization then yeah just wiping and starting over might be easiest

5

I’ve wiped the database and rebuilt it; only 20TB and I haven’t started the customization that I had done (that is not too critical at this stage, though). Unfortunately, the problem persists. :frowning:

Thanks for trying, BigWheel. Any other thoughts you or anyone else might have would be appreciated.

6

Some more manual searching and reading of log files leads me to believe the problem is with the EAC3 encoder. Searching the forums, I have found this thread:

I’ve issued the 777 chmod EasyAudioEncoder in the appropriate folder, even though I’m not running a docker. It seems to be working on my end, but I’ll have to wait a bit to see if other clients’ have their issues resolved. If the problem persists with other clients, I’m looking at finding the episodes with EAC3 audio, pulling that audio stream out, re-encoding to AC3 or similar, and stuffing it all back together. I’m sure it’s simpler to obtain another version of the episodes … :wink: