PMS makes the server swap+heavy load with errors from TranscodeOutputStream

Plex Media Server
1

Hello,

Since a few days, at random times, the PMS suddenly takes heavy load and I have a tons of logs like these one :

Jul 29, 2020 22:19:16.398 [0x7fa4a0ff9700] DEBUG - [TranscodeOutputStream] Changed instance plex-transcode-yaf5ejgedctv7k97ebxwu5vc-f5d5ade5-152f-4415-8928-9ba98b364756 which had to -1; was at chunk -1 with offset -1.000000 now at chunk -1

When this happens, the RAM (8GB) and swap (4GB) become full and the server get heavy loads (more than 30). I have no other solutions to restart the plexmediaserver service. Looks like a memory leak.

How can I fix this?

Thanks!

Infos:

  • plexmediaserver on Debian 10, Intel NUC CPU J5005, 8GB RAM, SSD 256GB
  • Server Version#: 1.19.5.3112
  • Player Version#: web - 4.39.1
  • I optimized the database
  • only one user uses Plex
  • Same error here

Logs when the problem happens around Jul 29, 2020 22:19 : Plex Media Server Logs_2020-07-30_07-51-39.zip (1001.0 KB)

2

Thank you for the logs.

I see only one thing in your logs, unfortunately, and that’s line after line of the transcoder error.

It would have been nice to capture the logs 30 seconds after playback starts so we can see what it was trying to do.

Assuming the J5005 is likely to be troubled like the J4xxx CPUs are,
I am going to suggest you try the VAAPI bypass setting. Doing this will bypass a known issue with the Intel iHD driver.

  1. Stop Plex
  2. Edit Preferences.xml ( vi , nano, etc works fine )
  3. Add VaapiDriver="i965" before the closing /> in the file
  4. It will look like: some_other_pref="value" VaapiDriver="i965" />
  5. Save, Exit
  6. Start PMS
  7. Test

The J4xxx is a known problem but it’s unfortunately low priority for Intel.

3

Thanks for your answer, i’ll try this out.

It would have been nice to capture the logs 30 seconds after playback starts so we can see what it was trying to do.

I’m not sure what logs you’d like to see. Currently I can’t reproduce the issue (I’m not the one using Plex “intensively”), So I exported all the logs I got.

What I know for sure is that after every update I need to remove the driver with this bash script I made :

#!/bin/bash
mv /usr/lib/plexmediaserver/lib/dri/iHD_drv_video.so /usr/lib/plexmediaserver/lib/dri/iHD_drv_video.old
service plexmediaserver restart

Without this, the playbak can’t even start.

The issue is recent (few days), and recently I installed docker on the server and running Home-Assistant in a container. I don’t think it can affect Plex, but why not. Docker don’t throw any errors except the timeouts to the socket, due to the server’s heavy load.

4

Hello,

Thanks, I confirm this fix is working.

5

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