Linux PMS error transcoding TrueHD, fails to play on ATV4K or web

I posted this problem to the wrong place apparently and I posted again on a thread about transcoding errors, but never got a reply. So here I go again.

Problem occurred after updating to server version 1.17.0.1709. Issue at the time was a spinning wheel on ATV4K, web, or PMP when attempting to play certain files. After trying a multitude of things I determined that the issue was ONLY with files that had TrueHD and thus needing transcoding. Full history can be found here.

I was finally able to update my docker (unRAID) to try the 1.18.1.1788 beta. It took a bit of work though.

In any case, the problem still persists in Web and Apple TV 4K. I still see errors in transcoding TrueHD. This just keeps repeating:

[Transcoder] [truehd_eae @ 0x925380] EAE timeout! EAE not running, or wrong folder? Could not read ā€˜/tmp/pms-e03d062b-74bc-461c-b6a4-f7191f7d8d75/EasyAudioEncoder/Convert to WAV (to 8ch or less)/2l4p8s4ltddb0f04nrijtslz_508-0-2.wav’
Sep 26, 2019 11:48:47.000 [0x14f92d9bb700] ERROR - [Transcoder] [truehd_eae @ 0x925380] error reading output
Sep 26, 2019 11:48:47.001 [0x14f8b35fa700] ERROR - [Transcoder] Error while decoding stream #0:1: Input/output error
Sep 26, 2019 11:48:50.000 [0x14f8b2df6700] ERROR - [Transcoder] [truehd_eae @ 0x925380] EAE timeout! EAE not running, or wrong folder? Could not read ā€˜/tmp/pms-e03d062b-74bc-461c-b6a4-f7191f7d8d75/EasyAudioEncoder/Convert to WAV (to 8ch or less)/2l4p8s4ltddb0f04nrijtslz_508-0-3.wav’

And again, InfusePro works no problem.

Latest logs attached. Any chance of getting this looked at?

Plex Media Server-092619.txt (1.4 MB)

Any ninja or dev want to help or should I just downgrade the server?

Hi @ionblue I run my main PMS on Unraid in our offical docker container with HW transcoding. I am also running the same build as you without any issues transcoding TrueHD for my Apple TV 4K.

No I know that is not helpful to hear on your case, but does mean that the issue is likely something particular to your setup or a rare bug. I’ll take a look at your logs now and see if I can provide some next step diagnostics.

Thank you. I’m just trying to get some eyes on this.

Other than the PMS update nothing on my side changed. It just up and started throwing the spinning wheel of death. If you look at the other thread I went down one path without success and started digging more into the logs. Every other movie I play that does not have TrueHD plays everywhere. It’s baffling and pisses me off unfortunately. Especially since it seems to just be me. :neutral_face:

And I know it doesn’t help, but the same ones that fail play in Infuse Pro. :frowning_face:

And I’m back on Version 1.17.0.1709 since the other didn’t help I didn’t see why I should be running it.

Also, other that PMP these fail on every other client so iOS, web, and Apple TV 4K.

Just to make sure transcoding was working I started a sync on a movie to my iPad. It’s cranking right along on the server and quite frankly this is way faster than PMS running on a Mac Mini.

Anyway, it’s at least another data point for you. Let me know if you need anything else.

@ionblue for your Docker container are you using /transcode as the Transcoder temporary directory and also mapping that in your Docker config to /tmp on the host?

based on your logs it looks like it’s still using /tmp inside the docker which should work, but worth testing using a volume mapping instead.

Plex setting

Unraid Docker Setting

After doing that can you see if you still experience the issue?

These are my mappings:

/transcode → /mnt/user/appdata/PlexMediaServer/transcode
/data → /mnt/user/
/config → /mnt/user/appdata/PlexMediaServer

The appdata share is on my SSD cache drive.

I don’t have an option in the config for /tmp (see attached). Should I manually add that? And I’ve never seen that and didn’t have issues before so why now?

Well crap. I manually added that and the docker completely stopped and doesn’t show up in the list at all. Switching to advanced view I see 2 orphaned images for Plex, but I can’t edit them. The only option is delete.

Guess I’ll install it again. At least my config should still work.

Yeah… I try adding that mapping and the damn thing fails every time.

Someone else over in my other thread is seeing this error, but with EAC3. Just an FYI.

Since I’ve managed to make several orphaned dockers trying to add that path in there, I went ahead and simplified my mappings. I’m waiting on Plex to update with the new paths to folders and then I’m going to run a Check Permissions cycle. This is one of the first things I tried after having problems, but it didn’t help. Figure it’s worth a shot again.

Oh yeah… I saw that you have ā€œprivilegedā€ checked so I even tried that. Still failed to build the docker.

This went from one problem to several problems. About to give up on this crap.

Jumping in on this thread from the other one. Here’s my original message:
https://forums.plex.tv/t/spinning-wheel-and-unable-to-play-video/465428/33?u=kevin.oconnor7

Attached a copy of my logs but the errors the exact same as OP’s here, just with the eac3 codec instead of TrueHD: Plex Media Server Logs_2019-09-27_14-20-44.zip (3.8 MB)

Also worth noting that I don’t use unraid, but I do use the official docker image.

Here are my transcoder settings:

And /transcode is just a normal docker volume, not a bind mount.

I think I know why this kept failing when I tried to add that path. I believe you’re telling me to use /tmp for /transcode ONLY. I left my other mapping in there (see pic) so I bet they were conflicting. Maybe?

I think Plex is done updating so I’ll give this a try.

I tried what I mentioned and still got the errors as before. New logs attached. They are probably messier than the others. I searched on ā€œ[Transcoder] [truehd_eaeā€ and found them throughout and also near the end which would have been the last attempt to play.

Plex Media Server-092719.txt (3.1 MB)

I tried something else based on your recommendation and it failed as well. I tried this:

/transcode -> /mnt/user/appdata/PlexMediaServer/tmp/

No joy. Plex did automatically setup the Transcode directory and sessions directory, but that’s as far as it got. So this is the same as how I had it.

I’ll ask something stupid though, has the docker container itself been updated recently? I dunno. I can’t think of anything else…

ChuckPA has been working on a new installer for linux systems.
Have you put any eyes on this thread?

Before introducing another element into this mix (the new packaging), I would first like to see if the issues I see in the existing logs can be addressed because the new installer adapts to what it finds and if what it finds isn’t functional then it won’t function.

My findings are:

  1. The transcoder temp specification is getting hard I/O errors whenever it tries to operate. Where is this temp directory? Is it on a network share without POSIX file locking? If so, it will fail (network shares for transcoding are ill-advised under any circumstances).
Sep 27, 2019 14:20:12.001 [0x7f5baeffd700] ERROR - [Transcoder] [eac3_eae @ 0x173c040] error reading output
Sep 27, 2019 14:20:12.001 [0x7f5bbf7fe700] ERROR - [Transcoder] Error while decoding stream #0:1: Input/output error
Sep 27, 2019 14:20:12.510 [0x7f5c027fc700] DEBUG - HTTP 200 response from GET https://plex.tv/servers/142bb34d64cf6714dff5f7796ee6f8ff6d7f4b22/access_tokens.xml?auth_token=xxxxxxxxxxxxxxxxxxxx&includeProfiles=1&includeProviders=1
  1. The network link is always changing. Every time it does, the link between server and client is terminated ( How TCP works )
Sep 27, 2019 14:20:20.525 [0x7f5c15ffb700] DEBUG - Network change notification but nothing changed.
Sep 27, 2019 14:20:20.525 [0x7f5c15ffb700] DEBUG - NetworkInterface: received Netlink message len=1440, type=RTM_NEWLINK, flags=0x0
Sep 27, 2019 14:20:20.525 [0x7f5c15ffb700] DEBUG - NetworkInterface: Netlink information message family=0, type=1, index=29539, flags=0x1003, change=0x0
Sep 27, 2019 14:20:20.525 [0x7f5c15ffb700] DEBUG - Network change.
Sep 27, 2019 14:20:20.525 [0x7f5c15ffb700] DEBUG - NetworkInterface: Notified of network changed (force=0)
Sep 27, 2019 14:20:20.525 [0x7f5c15ffb700] DEBUG - Network change notification but nothing changed.
Sep 27, 2019 14:20:20.731 [0x7f5c15ffb700] DEBUG - NetworkInterface: received Netlink message len=520, type=RTM_DELLINK, flags=0x0
Sep 27, 2019 14:20:20.731 [0x7f5c15ffb700] DEBUG - NetworkInterface: Netlink information message family=0, type=1, index=29538, flags=0x1002, change=0xffffffff
Sep 27, 2019 14:20:20.755 [0x7f5c15ffb700] DEBUG - NetworkInterface: received Netlink message len=312, type=RTM_NEWLINK, flags=0x0
Sep 27, 2019 14:20:20.755 [0x7f5c15ffb700] DEBUG - NetworkInterface: Netlink information message family=7, type=1, index=29539, flags=0x11043, change=0x0
Sep 27, 2019 14:20:20.755 [0x7f5c15ffb700] DEBUG - Network change.
Sep 27, 2019 14:20:20.755 [0x7f5c15ffb700] DEBUG - NetworkInterface: Notified of network changed (force=0)
Sep 27, 2019 14:20:20.756 [0x7f5c15ffb700] DEBUG - Network change notification but nothing changed.
Sep 27, 2019 14:20:20.756 [0x7f5c15ffb700] DEBUG - NetworkInterface: received Netlink message len=312, type=RTM_NEWLINK, flags=0x0
Sep 27, 2019 14:20:20.756 [0x7f5c15ffb700] DEBUG - NetworkInterface: Netlink information message family=7, type=1, index=29539, flags=0x11043, change=0x0
Sep 27, 2019 14:20:20.756 [0x7f5c15ffb700] DEBUG - Network change.
Sep 27, 2019 14:20:20.756 [0x7f5c15ffb700] DEBUG - NetworkInterface: Notified of network changed (force=0)
Sep 27, 2019 14:20:20.758 [0x7f5c15ffb700] DEBUG - Network change notification but nothing changed.
Sep 27, 2019 14:20:20.758 [0x7f5c15ffb700] DEBUG - NetworkInterface: received Netlink message len=312, type=RTM_NEWLINK, flags=0x0
Sep 27, 2019 14:20:20.758 [0x7f5c15ffb700] DEBUG - NetworkInterface: Netlink information message family=7, type=1, index=29539, flags=0x11043, change=0x0
Sep 27, 2019 14:20:20.758 [0x7f5c15ffb700] DEBUG - Network change.
Sep 27, 2019 14:20:20.758 [0x7f5c15ffb700] DEBUG - NetworkInterface: Notified of network changed (force=0)

These two points along are, imo, the first two problems to resolve. When there are no errors at this level you might just find everything works as it should.

First, the transcode temp directory is on the cache drive and is local. It’s in the main PMS directory that Plex obviously has access to.

For the second part, I have no idea where to look for whatever this is. Up until recently I had been running PMS without any front facing issues. The docker is setup for host networking. Should I try another method? Or do you think this is something deeper and maybe on the unRAID side or the router side?

The server is directly connected to my router and has a hard coded IP if that makes any difference.

Can you give me an idea of where to start looking for the network part of this?

I dug through the unRAID logs and router logs. I can’t see anything out of the ordinary anywhere. There was one webgui error int the unRAID logs but have been told it’s harmless.

The only other network related oddity is that the remote access section of settings says it isn’t available out side the network when it actually is. I had reported this a wile back and was told it was a known issue and can be ignored. If I refresh it it always says fully available and I’ve tested on a cellular network and it is indeed available.

So I still don’t know where to go with any of this.