Plex server randomly becomes unresponsive

[Tabmowtez]

Server Version#: v1.40.5.8921
Player Version#: Android Players & Web

I recently migrated to docker from a TrueNAS plugin and this issue was occurring previously. Every now and then it seems like the server is unresponsive. I can’t load anything, can’t connect at all, sometimes when I’m watching a show it happens between playing episodes. If I wait long enough it just comes back. I’m not sure if the below logs are relevant. If I need to provide something else just let me know…

When I migrated I did a complete refresh of all my libraries as the paths changed, I optimised the database as well.

I posted some extra logs from this morning after using the PlexDBRepair tool on auto. As it suggested, I then scanned my 3 media libraries and I got those extra error messages in the logs.

Logs:
==> Plex Media Server.log <==

Aug 30, 2024 19:09:44.540 [135625621646136] ERROR - [Req#2da4] HTTP 408 downloading url http://127.0.0.1:32400/library/metadata/53735?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&offset=-1&X-Plex-Incomplete-Segments=1&X-Plex-Session-Identifier=aa20cdf8406d3c6c-com-plexapp-android
Aug 30, 2024 19:09:44.540 [135625814584120] ERROR - [Req#2da4] TranscodeUniversalRequest: unable to get container: /library/metadata/53735?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&offset=-1&X-Plex-Incomplete-Segments=1&X-Plex-Session-Identifier=aa20cdf8406d3c6c-com-plexapp-android
Aug 30, 2024 19:09:44.540 [135625814584120] WARN - [Req#2da4] Failed to obtain a streaming resource for direct play of part 136951
Aug 30, 2024 19:10:04.544 [135625609063224] ERROR - [Req#2d95] HTTP 408 downloading url http://127.0.0.1:32400/library/metadata/53735?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&offset=-1&X-Plex-Incomplete-Segments=1&X-Plex-Session-Identifier=aa20cdf8406d3c6c-com-plexapp-android
Aug 30, 2024 19:10:04.544 [135625705532216] ERROR - [Req#2d95] TranscodeUniversalRequest: unable to get container: /library/metadata/53735?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&offset=-1&X-Plex-Incomplete-Segments=1&X-Plex-Session-Identifier=aa20cdf8406d3c6c-com-plexapp-android
Aug 30, 2024 19:10:04.544 [135625705532216] WARN - [Req#2d95] Failed to obtain a streaming resource for direct play of part 136951
Aug 30, 2024 19:10:24.552 [135625554537272] ERROR - [Req#2dab] HTTP 408 downloading url http://127.0.0.1:32400/library/metadata/53735?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&offset=-1&X-Plex-Incomplete-Segments=1&X-Plex-Session-Identifier=aa20cdf8406d3c6c-com-plexapp-android
Aug 30, 2024 19:10:24.552 [135625688755000] ERROR - [Req#2dab] TranscodeUniversalRequest: unable to get container: /library/metadata/53735?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&offset=-1&X-Plex-Incomplete-Segments=1&X-Plex-Session-Identifier=aa20cdf8406d3c6c-com-plexapp-android
Aug 30, 2024 19:10:25.931 [135625776835384] WARN - Held transaction for too long (/home/runner/actions-runner/_work/plex-media-server/plex-media-server/Statistics/StatisticsManager.cpp:294): 0.240000 seconds
Aug 30, 2024 19:10:26.918 [135625646811960] INFO - [Req#2dec] AutoUpdate: no updates available
Aug 30, 2024 19:10:27.258 [135625613257528] INFO - [Req#2e02] AutoUpdate: no updates available
Aug 30, 2024 19:10:41.228 [135625776835384] WARN - Took too long (0.160000 seconds) to start a transaction on /home/runner/actions-runner/_work/plex-media-server/plex-media-server/Statistics/StatisticsManager.cpp:277
Aug 30, 2024 19:10:41.228 [135625776835384] WARN - Transaction that was running was started on thread 135625776835384 at /home/runner/actions-runner/_work/plex-media-server/plex-media-server/Statistics/StatisticsBandwidth.cpp:40
Aug 30, 2024 19:10:41.358 [135625814584120] WARN - [Req#2dfa] SLOW QUERY: It took 400.000000 ms to retrieve 20 items.
Aug 30, 2024 19:10:41.393 [135625797806904] WARN - [Req#2d67] SLOW QUERY: It took 740.000000 ms to retrieve 1 items.
Aug 30, 2024 19:10:41.428 [135625676172088] WARN - [Req#2db9] SLOW QUERY: It took 220.000000 ms to retrieve 1 items.

==> Plex Media Server.log <==
Aug 31, 2024 11:43:11.399 [123138820025144] ERROR - Waited over 10 seconds for a busy database; giving up.
Aug 31, 2024 11:44:12.847 [123138815830840] ERROR - Waited over 10 seconds for a busy database; giving up.
Aug 31, 2024 11:44:12.847 [123138815830840] ERROR - Failed to begin transaction (/home/runner/actions-runner/_work/plex-media-server/plex-media-server/Statistics/StatisticsManager.cpp:277) (tries=1): Cannot begin transaction. database is locked
Aug 31, 2024 11:44:43.561 [123138815830840] WARN - Took too long (0.140000 seconds) to start a transaction on /home/runner/actions-runner/_work/plex-media-server/plex-media-server/Statistics/StatisticsManager.cpp:277
Aug 31, 2024 11:44:43.561 [123138815830840] WARN - Transaction that was running was started on thread 123138723556152 at /home/runner/actions-runner/_work/plex-media-server/plex-media-server/Statistics/StatisticsManager.cpp:277
Aug 31, 2024 11:45:45.000 [123138820025144] ERROR - Waited over 10 seconds for a busy database; giving up.

[Tabmowtez]

So, just in case someone else has this issue - after reading through a bunch of similar issues other people were having, I used the DBRepair tool, it didn’t really find any issues but considering this server has been upgraded for around ~4 years now it was probably due.

The original issue I had was every now and then the server would be somewhat unresponsive when I was running it as a plugin on TrueNAS Core. Considering they are no longer supporting the plugins anymore I decided to migrate my installation to a VM and docker, this is where a lot of my issues were amplified. I copied my Library directory to a NFS share, and this was the culprit for most of my issues it seems. I saw some posts from others saying this is a big no no and I can definitely confirm that!

I moved my Library directory to an iSCSI disk and my Plex instance has been rock solid for the last ~12 hours. Library scans finish within minutes compared to hours on NFS and everything is super snappy now.

So I guess the TLDR is, do not use Network File Systems for your Database
I’m also assuming the random issues I had when it was a TrueNAS plugin was probably due to issues with the Database that DBRepair may have sorted too but time will tell for sure.

Hope this helps someone else!