PMS for Windows crashes during launch every time (after password change)

Thanks, that key is not even in my registry. I guess it would default to off though?

It should be on by default.

Thanks again. Added it as a DWORD and set it to 0.

But PMS still crashes like before, it was worth a try though!

I have made some more tries to get it going but nothing has improved.

Some questions that perhaps someone can help me with:

  1. The errors in the PMS log, are they at all important or can they be ignored? See below for reference from latest run. NB, the log goes on a long time after these errors.
  2. In the beginning, when PMS crashed it generated a Windows crash dump and an Application Error in Windows event log. It happened many many times but it no longer happens, though PMS seems to crash in just the same way and the PMS logs look very similar. Is this expected, or could it be a clue to something?

Errors in PMS log:

Oct 11, 2025 11:57:30.751 [7624] ERROR - Grabber: timeout waiting for Plex Tuner Service to finish starting. 
Oct 11, 2025 11:58:29.418 [13464] ERROR - BufferingLineReader: failed to read line (error: -2) 
Oct 11, 2025 11:58:29.418 [13464] ERROR - Timeout or error reading status line from plug-in pipe [com.plexapp.system], we're killing it. 
Oct 11, 2025 11:58:29.418 [13464] DEBUG - Stopping [com.plexapp.system]. 
Oct 11, 2025 11:58:29.418 [13464] DEBUG - [com.plexapp.system] Killing. 
Oct 11, 2025 11:58:29.419 [13464] ERROR - Error received reading configuration for com.plexapp.system 

So, I’m trying to tread on here to perhaps be able to run PMS again without having to rebuild all of the databases…

What I have tried now was to temporarily set HKCU\Software\Plex, Inc.\Plex Media Server\LocalAppDataPath to another (newly created, blank) location. When I do this, Plex seems to start up without issues and does not crash like when pointed to the existing LocalAppDataPath folder.
Of course, I would be extremely happy to not have to recreate everything from scratch, so I try to use the information from this to pinpoint what is going wrong otherwise.

Here are my main observations so far:

  1. With my existing metadata, the Plex Tuner Service.log ends very soon after being started. It never reaches device_manager_t::init, but up to that the logs look similar. Could this be something causing a domino effect leading to PMS crashing? The logs clearly complain about the Plex Tuner Service not running (see relevant PMS log excerpt #1 below). This happens early on when trying to launch PMS. What could cause the tuner service to fail like this?
  2. Something that seems to lead up to the PMS process stopping (with existing metadata) is that it notices that the com.plexapp.system plugin is not responding as expected (see relevant PMS log excerpt #2 below). The com.plexapp.system.log for the ā€œworking oneā€ (the one where I cleared metadata folder) has a couple of critical errors, but it still seems to work so I guess that should not be the issue. The logs for the ā€œfailing oneā€ are never created though, so it seems the com.plexapp.system plugin is never started or not able to start? (In fact, there are no new logs at all created in the Logs\PMS Plugin Logs folder since the issues began.) It seems the system is waiting for ~3 minutes for com.plexapp.system (after trying to spawn it and get prefixes) to be available. For the ā€œworking oneā€ it seems that com.plexapp.system is available after ~2 seconds. What could cause com.plexapp.system failing or not being available as expected?
  3. The ā€œfailing oneā€ tries to initiate something related to a play queue (ā€œBPQā€) (see PMS log excerpt #3 below) which does not happen on the ā€œworking oneā€, but I don’t see any errors related to this. I guess that since the ā€œworking oneā€ is ā€œcleanā€ there is just no queue to process?
  4. A lot of local HTTP requests that start happening on the ā€œworking onā€ never happens on the ā€œfailing oneā€, I guess these could be cascading issues from services/plugins not being available?

Thanks to anyone with insight into what might be going on who wants to try and help me!

PMS log excerpt #1:

Oct 13, 2025 11:18:00.938 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:18:11.026 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:18:21.026 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:18:31.027 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:18:41.027 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:18:51.027 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:19:01.028 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:19:11.029 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:19:21.030 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:19:31.030 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:19:41.030 [9804] DEBUG - Grabber: Plex Tuner Service not running; retrying in 10 seconds.
Oct 13, 2025 11:19:51.031 [9804] ERROR - Grabber: timeout waiting for Plex Tuner Service to finish starting.

PMS log excerpt #2:

Oct 13, 2025 11:20:51.424 [22452] ERROR - BufferingLineReader: failed to read line (error: -2)
Oct 13, 2025 11:20:51.424 [22452] ERROR - Timeout or error reading status line from plug-in pipe [com.plexapp.system], we're killing it.
Oct 13, 2025 11:20:51.424 [22452] DEBUG - Stopping [com.plexapp.system].
Oct 13, 2025 11:20:51.424 [22452] DEBUG - [com.plexapp.system] Killing.
Oct 13, 2025 11:20:51.425 [22452] ERROR - Error received reading configuration for com.plexapp.system
Oct 13, 2025 11:20:51.425 [22452] DEBUG - Stopping [com.plexapp.system].

PMS log excerpt #3:

Oct 13, 2025 11:19:47.075 [20128] DEBUG - BPQ: [Starting] -> [Processing]
Oct 13, 2025 11:19:47.076 [20128] DEBUG - BPQ: generating queue items from 2 generator(s)
Oct 13, 2025 11:19:47.077 [20128] DEBUG - BPQ: generated 1 item(s) for queue
Oct 13, 2025 11:19:47.078 [20128] DEBUG - PlayQueue: Added 1 items in 0.0 sec.
Oct 13, 2025 11:19:47.078 [20128] DEBUG - PlayQueue: total generated: -1, before cursor: 0, after cursor: 0
Oct 13, 2025 11:19:47.078 [20128] DEBUG - PlayQueue: Inserting 0 new items, starting at 0.000000.
Oct 13, 2025 11:19:47.078 [20128] DEBUG - BPQ: [Processing] -> [Paused]

Now I have also taken a look at some more available log files.

Currently no new logs are generated in Plex Media Server\Logs (and subdirectories) except Plex Media Server.log and Plex Tuner Service.log. This also means that some older logs from when the issues started are still available.

I started with the com.plexapp.system.log files of which there are six available from around the time when the issues started. The two oldest ones (starting 2025-09-10 01:45:50 and 01:46:20) looks similar to the com.plexapp.system.log files generated with a blank setup and also seems to exit gracefully(?) with ā€œSaved the dictionary fileā€ on the last line (~5 seconds after log start). The oldest one has two critical errors: Exception in thread named 'refresh_servers' → HTTP Error 503 and exception getting hosted resource hashes → HTTP Error 404. The later one has only the first of the errors.

After these first two logs the errors seem more severe. The three following logs that start 01:47:57, 01:54:51 and 01:58:19 are almost identical. The first critical error still appears, but ~2m in we can see Starting timed thread named 'refresh_servers' and ~20s later there is again Exception in thread named 'refresh_servers' and now with another result, timed out, and then the log abruptly ends.

The last available log file, starting 02:06:29, is almost identical again, except a weird(?) additional ending - after the now familiar Exception in thread named 'refresh_servers' → timed out at 02:08:50 there is a blank line and then one more line timestamped almost one hour later, 2025-09-10 03:06:30.339 (4d98) : DEBUG (runtime:1156) - Starting timed thread named 'invalidation_timer' (the ā€˜invalidation_timer’ thread is seen being scheduled an hour earlier, like in the other log files, but only here is it seen to be started).

The only other plug-in logs generated at this time are com.plexapp.agents.imdb.log files. These were only generated at some of the launch times, specifically starting at 01:45:52 and 01:47:58. Both are very similar, showing a critical error Exception getting hosted resource hashes → HTTP Error 404 and ~20s in ending with WARNING (core: 541) - Unable to retrieve the machine identifier or server version. The critical error does not seem to be important since it also seen in the com.plexapp.agents.imdb.log files from before the issues started.

Regrettably, the main PMS and Plex Tuner Service logs from this time (or older) does not seem to be available anymore, since I made so many restarts before being aware of the log deletion policy. There are however several Plex Media Scanner.log files available from shortly before the issues started, though they show no errors or hints of what could have been causing issues (as far as I can see).

Based on this I am thinking:

  1. What could be the changes that caused new behavior as seen in com.plexapp.system.log files?
  2. What could be the changes that caused com.plexapp.system.log files no longer being generated?
  3. Is the ending warning in the last two com.plexapp.agents.imdb.log files indicative of anything?
  4. What could be the changes that caused com.plexapp.agents.imdb.log files no longer beign generated?