Server Version#: 1.21.1.3830
Player Version#: AppleTV 7.11 (22422) / Kodi, maybe more
Hello all,
I seem to have a repeated issue with my PMS. Users/Friends on Apple TV regularly encounter this error message: “Conversion failed. The transcoder exited due to an error.”. Other users also have their active session halted.(File removed)
On looking at logs, i seem to have the same issue as mentioned in this thread, but i’m on Mac without MalwareBytes.
This also seems to relate to this issue , but the user does not appear to have found a solution.
Let me know if a thread is already open on this, i searched through the forums and did not find anything that could apply to this issue, for Mac specifically.
I’m attaching logs as well.
Thanks in advance!
Death’
Jan 03, 2021 21:20:48.664 [0x700008f2e000] ERROR - [Transcoder] [tcp @ 0x7faf71e5d500] Connection to tcp://127.0.0.1:32400 failed: Operation timed out
Jan 03, 2021 21:20:48.665 [0x70000a19a000] ERROR - [Transcoder] [stream_segment,ssegment @ 0x7faf7386dc00] Failed to open segment list 'http://127.0.0.1:32400/video/:/transcode/session/E0E3722C-2801-45E5-A387-D89A91FE56DD/e5fb3e9c-d817-455e-a733-601d77a169df/seglist?X-Plex-Http-Pipeline=infinite'
Jan 03, 2021 21:20:48.665 [0x700009552000] ERROR - [Transcoder] av_interleaved_write_frame(): Operation timed out
There is a possibility that this is happening because of running out of tcp port numbers from the dynamic tcp pool - as mentioned in this forum post SOLUTION: Conversion failed. The transcoder exited due to an error - but all previous cases I looked at were on windows
The errors arose after scanning the TV and Movies libraries - these scans used started at 21:20:18 and the errors started at 21:20:46
We need to establish if the errors are to do with the tcp ports
The issue I linked to is being addressed as follows:
update to curl used by Plex Media Server that will reduce the number of ports used by 66%
Advice to users to reduce Timewait system setting for tcp requests and to increase the dynamic tcp port if there are large libraries
Upgrade movies libraries to the new in-line Plex Media Server scanner that avoids use of the dynamic pool for scanning movies libraries
For TV libraries, work is in progress to change the scanner to be in-line and there is already a forum preview release for this development work
I would suggest that on getting this failure, you find out what tcp connections are in place (using lsof and saving output to a text file) and what the dynamic tcp port range is and what the timewait is before ports are released - this together with the server logs will help establish if it is the same issue
I think this tells me all I need to know.
I noticed in my settings that my library was scanned automatically + every 15 minutes. Not sure when this was set but this must be why the server is overloaded with tcp ports.
I’ve changed the policy and asked my friends/users to monitor for any reoccurrence of the issue.
I will report back with outcome of the aforementioned trial if my attempted corrective action fails.
Otherwise, I’ll close this issue as this could be a simple case of extreme settings.
I looked at this post on this and understand you mainly recommend to split the library if there is a large number of movies (I have 1236x so wouldn’t usually be considered too large). I’ll try this if my above attempt does not work.
