Thanks for the logs
I can see evidence of the tcp dynamic port exhaustion issue
This -138 error is one of the clues to the problem
Nov 10, 2020 17:38:04.586 [8624] ERROR - [Transcoder] [tcp @ 00acb0c0] Connection to tcp://127.0.0.1:32400 failed: Error number -138 occurred
The system ran out of tcp dynamic ports whilst scanning/refreshing the TV Shows library. We run out of ports when scanning / analyzing large libraries. This happened when we got to about half way through the TV Shows library
Nov 10, 2020 17:16:22.913 [0308] DEBUG - Activity: updated activity 7c2c5a1b-bb8c-472b-8ec0-8683a67954e2 - completed 44.9% - Scanning TV Shows
Nov 10, 2020 17:16:22.957 [0308] ERROR - [Transcoder] [tcp @ 00b3b3c0] Connection to tcp://127.0.0.1:32400 failed: Error number -138 occurred
An update to curl will be incorporated into a future release of Plex Media Server which should reduce dynamic port usage by 66%. This in itself will not resolve the issue for large libraries. The problem arises when the scanner process communicates with the server process as it goes through the shows and seasons
Most of the scanning is now in-line and so reducing the need for having sepearate processes running. The last to change was the Movies category and the next that will be brought in-line will be TV Shows and development work is in progress on that
The issue here is the size of the library and the fact that scanning is in a separate process and that we have a curl bug that increased port usage for requests three-fold
Suggested Actions:
- consider splitting the TV Shows library into 2
- Look into reducing Time_Wait and doubling the tcp dynamic port in windows - see Transcoder crashes on access errors - #44 by sa2000
- Reduce the frequency of periodic scans
and of course upgrade to the version of Plex Media Server that will have the curl fix (I have no information at this point as to how soon this will be )
