EAE timeout! EAE not running, or wrong folder?

Plex Media Server NAS & Devices
1

Server Version#: 1.19.4.2935
Player Version#: 4.36.1

I have a docker installation.

I can’t play EAC3 audio.
A few months ago, the change in the inotify table had solved the problem but, this is no longer the case, even after increasing this value!

I also came out of the container / tmp folder but this changes nothing.

Jun 30, 2020 19:26:54.000 [0x7ff99e7fc700] Erreur — [Transcoder] [eac3_eae @ 0xda8680] EAE timeout! EAE not running, or wrong folder? Could not read '/tmp/pms-6c30145b-db08-44da-8d75-433e9650a969/EasyAudioEncoder/Convert to WAV (to 8ch or less)/7e2ej4cglzyrvl30i5fvtob8_493-0-31.wav'
Jun 30, 2020 19:26:54.000 [0x7ff99e7fc700] Erreur — [Transcoder] [eac3_eae @ 0xda8680] EAE timeout! EAE not running, or wrong folder? Could not read '/tmp/pms-6c30145b-db08-44da-8d75-433e9650a969/EasyAudioEncoder/Convert to WAV (to 8ch or less)/7e2ej4cglzyrvl30i5fvtob8_493-0-31.wav'
Jun 30, 2020 19:26:54.000 [0x7ff99e7fc700] Erreur — [Transcoder] [eac3_eae @ 0xda8680] EAE timeout! EAE not running, or wrong folder? Could not read '/tmp/pms-6c30145b-db08-44da-8d75-433e9650a969/EasyAudioEncoder/Convert to WAV (to 8ch or less)/7e2ej4cglzyrvl30i5fvtob8_493-0-31.wav'
Jun 30, 2020 19:26:54.000 [0x7ff9257fa700] Erreur — [Transcoder] [eac3_eae @ 0xda8680] error reading output
Jun 30, 2020 19:26:54.001 [0x7ff99e7fc700] Erreur — [Transcoder] Error while decoding stream #0:1: Input/output error
Jun 30, 2020 19:26:54.000 [0x7ff9257fa700] Erreur — [Transcoder] [eac3_eae @ 0xda8680] error reading output
Jun 30, 2020 19:26:54.001 [0x7ff99e7fc700] Erreur — [Transcoder] Error while decoding stream #0:1: Input/output error
Jun 30, 2020 19:26:54.000 [0x7ff9257fa700] Erreur — [Transcoder] [eac3_eae @ 0xda8680] error reading output
Jun 30, 2020 19:26:54.001 [0x7ff99e7fc700] Erreur — [Transcoder] Error while decoding stream #0:1: Input/output error
2

How much media has been added since the table size was increased?

Please restart PMS,
wait 2 minutes - completely idle
start playback - let fail
Download Logs ZIP
attach please.

3

Thanks for the reply, about 3TB I think.

the command: find / folder -type d -print | wc -l

Show 62721 on my media files
Indicates 1244746 on the “Plex Media Server” and “Transcode” folder

I indicated fs.inotify.max_user_watches = 524288 and it worked until I changed the volume of my plex installation, I am now at 1343488 but that does not change anything …

here are the logs:-----

Edit :I have the impression that this also makes the maintenance run in a loop because the intro detection does not manage to do this!

4

How many directories exist ?
Do you need 1,344,488 directory slots? That is over 512 MB of memory.

5

Value of the commande “find / folder -type d -print | wc -l” for my media files is 62721 ( I do not know the exact value of the total number of all files (video, music, photo, …) )

Folder Plex Media Server, 743286 files, and 1138044 folder.

6

I believe I found your error?

That’s not the value you should base on.

find /media-directory-1 /media-directory-2  -type d -print | wc -l

Do not count the directories within the metadata structure. This will confuse Plex and lead to kernel failure. This is why media should always be kept separate from the metadata storage (/config)

You may list all the media directories on the command line before the -type d predicate or repeat the command as needed and add the totals.

I think you asked the kernel to use more memory than it has. The kernel will refuse

7

Okay, I had a doubt whether to include the config file in the calculation or not, and so I tried both ways.

only my media / photos files =

Edit :
I’m change inotify table to 98304, same problem

8

You will also need restart the system OR

  1. Stop Plex
  2. sysctl -p
  3. Start Plex.
9

I’m in docker, the docker must necessarily be on the way to make this order, and therefore also plex.

Unless the inotify setting should be done on the main system, not in the docker?

10

inotify must always be done on the host.
Containers rely on host resources.
The container is abstracted from the kernel but does use its services.

A VM, running in a different NameSpace on the host CPU, has its own resources. Memory, IO and other hardware is shared.

11

I’m modif /etc/sysctl.conf on the host and in ssh : sysctlp -p

Restart Plex, same problem

12

stupid question time: (on my part). Docker container (/config) is resident on the local host or a network share?

13

Yes, /config is resident on the local host.

14

Hello,
I’m indicated fs.inotify.max_user_watches = 524288 in the host ( /etc/sysctl.conf ) and start Plex.
EAC3 transcoding is active again!
Thank you for the time spent and your help!

15

Sorry to bring life back to a dying thread but I’m not 100% sure on what to do.
Should I go to /etc/sysctl.conf and ad a line saying ‘fs.inotify.max_user_watches = 524288’?

16
1 Like
17

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.