Server Version#: current: 1.23.1.4571, problem version: 1.23.2.46xx
I am unable to update to any Plex Pass 1.23.2.46xx versions. I get an immediate crash dump and the server tries to restart, but can never do so. I haven’t been able to capture the dump log to look at because they vanish before I can see them, even if I stop the server immediately on crash.
I’m attaching a log file from the latest upgrade attempt this morning. I have tried full database repair as per the FAQ directions but still, have the same result.
I have a feeling I’m bumping into a strange use case, but I’m not doing anything out of the ordinary. The main errors I see are database schema errors from the metadata_items table.
The crash seems related to the DVR setup. When I remove the DVR setup entirely I am then able to update the server to 1.23.2.4656, but as soon as I add the DVR back while running 1.23.2.4656 it adds the tuner and starts downloading the EPG data but once it hits 100% the server crash dumps. The log file during that operation was also filled with errors related to the tuner.
I had to roll back to 1.23.1.4571 to get the server running again.
FYI: Turning off DEBUG logging makes it impossible to debug because that looks like nothing more than bad media files (which might be the case but need to prove that out first)
Please turn DEBUG back on. You’re not saving space with it off.
when back on & saved, recreate.
What we see at this point is:
Jun 06, 2021 09:08:28.272 [0x7f1ac08d9b38] ERROR - [Transcoder] [aac @ 0x7fda350934c0] channel element 0.0 is not allocated
Jun 06, 2021 09:08:28.273 [0x7f1abede3b38] ERROR - [Transcoder] Error while decoding stream #0:1: Invalid data found when processing input
Jun 06, 2021 09:08:28.273 [0x7f1abcedeb38] ERROR - [Transcoder] [aac @ 0x7fda350934c0] channel element 0.0 is not allocated
Jun 06, 2021 09:08:28.273 [0x7f1ac1219b38] ERROR - [Transcoder] Error while decoding stream #0:1: Invalid data found when processing input
Jun 06, 2021 09:08:28.273 [0x7f1abf305b38] ERROR - [Transcoder] [aac @ 0x7fda350934c0] channel element 0.0 is not allocated
Jun 06, 2021 09:08:28.273 [0x7f1abf1a7b38] ERROR - [Transcoder] Error while decoding stream #0:1: Invalid data found when processing input
Jun 06, 2021 09:08:28.273 [0x7f1abeed3b38] ERROR - [Transcoder] [aac @ 0x7fda350934c0] channel element 0.0 is not allocated
Jun 06, 2021 09:08:28.273 [0x7f1abde42b38] ERROR - [Transcoder] Error while decoding stream #0:1: Invalid data found when processing input
Jun 06, 2021 09:08:28.358 [0x7f1ac0b23b38] ERROR - IntroDetector: Failed to transcode file (69): /syno/Video/TV_Shows/Hell's Kitchen (US)/Season 14/Hell's Kitchen (US) - S14E03 - 16 Chefs Compete.mp4
Jun 06, 2021 09:09:36.788 [0x7f1ac274eb38] WARN - Took too long (0.320000 seconds) to start a transaction on ../Library/MetadataItem.cpp:1091
Jun 06, 2021 09:09:36.788 [0x7f1ac274eb38] WARN - Transaction that was running was started on thread 0x7f1ac0ab0b38 at ../Library/FullTextSearch.cpp:85
Jun 06, 2021 09:09:36.803 [0x7f1abe61ab38] ERROR - Error issuing curl_easy_perform(handle): 35
Understood. Here is the full log, with debug, from this morning when I installed 1.23.2.4656. As before, it installs, starts (I get connected over the WebUI for about 3-5 seconds), then the server crashes out with a dump.
One other error I’ve been getting for a few months, but haven’t been able to identify is the following:
Jun 07 08:39:30 shuttle Plex Media Server[205197]: ERROR: Empty fingerprint
Jun 07 08:39:39 shuttle Plex Media Server[205254]: ERROR: Empty fingerprint
I assumed it was maybe linked to some bad media files, but I haven’t been able to nail down anything specific. Typically after I encounter this error the server will no longer generate any chapter thumbnails or video previews until I manually restart the the PMS service.
I additionally included in the zip the Plex Tuner Service.log file because it had activity leading up to the crash after the install. No other log files were written to at that time. They are in a single zip with the dmp file.
Please let me know if you need anything else, I’m not sure why this is so angry, but I’ve been running an Ubuntu server version of PMS for at least 8 years now and this is the first update that’s ever done this.
According to my logs, /var/log/syslog, it looks like plex is crashing due to a segfault error when using cuda. Below is an excerpt from syslogs that’s been repeating hundreds of times since Plex is crashing every few minutes:
Jun 7 13:06:48 mail kernel: [ 9511.448317] resource sanity check: requesting [mem 0x000c0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000c0000-0x000dffff window]
Jun 7 13:06:48 mail kernel: [ 9511.448384] caller os_map_kernel_space+0x8d/0xb0 [nvidia] mapping multiple BARs
Jun 7 13:06:48 mail systemd-udevd[8081]: Process '/bin/systemctl start --no-block nvidia-persistenced.service' failed with exit code 5.
Jun 7 13:06:49 mail Plex Media Server[4009]: ****** PLEX MEDIA SERVER CRASHED, CRASH REPORT WRITTEN: /var/lib/plexmediaserver/Library/Application Support/Plex Media Server/Crash Reports/1.23.2.4656-85f0adf5b/PLEX MEDIA SERVER/1d306f57-7ab9-41b7-3c9a8e89-c5f0fd60.dmp
Jun 7 13:06:49 mail kernel: [ 9511.803604] Plex Media Serv[8071]: segfault at 69 ip 00007f156c26fcbb sp 00007f156fb34640 error 4 in libcuda.so.340.108[7f156c038000+d9c000]
Jun 7 13:06:49 mail systemd[1]: plexmediaserver.service: Main process exited, code=killed, status=11/SEGV
Jun 7 13:06:49 mail systemd-udevd[8087]: Process '/bin/systemctl stop --no-block nvidia-persistenced' failed with exit code 5.
<CUT>
Jun 7 13:08:19 mail systemd[1]: plexmediaserver.service: State 'stop-sigterm' timed out. Killing.
Jun 7 13:08:19 mail systemd[1]: plexmediaserver.service: Killing process 4048 (Plex Script Hos) with signal SIGKILL.
Jun 7 13:08:19 mail systemd[1]: plexmediaserver.service: Killing process 4215 (Plex Script Hos) with signal SIGKILL.
Jun 7 13:08:19 mail systemd[1]: plexmediaserver.service: Killing process 6804 (Plex Relay) with signal SIGKILL.
Jun 7 13:08:19 mail systemd[1]: plexmediaserver.service: Killing process 6805 (Plex Relay) with signal SIGKILL.
Jun 7 13:08:19 mail systemd[1]: plexmediaserver.service: Failed with result 'signal'.
Jun 7 13:08:24 mail systemd[1]: plexmediaserver.service: Service hold-off time over, scheduling restart.
Jun 7 13:08:24 mail systemd[1]: plexmediaserver.service: Scheduled restart job, restart counter is at 12.
I went into the Preferences.xml and set 0 for HardwareAcceleratedEncoders="0" and HardwareAcceleratedCodecs="0" and it’s been running smoothly for an hour now with no crashes. Plus playback of videos were successful.
Looks like the culprit is cuda for me. Check your logs to be sure.
Unfortunately not Nvidia-related. I’m using an Intel i7 7700K for QSV hardware acceleration. I can try to disable hardware acceleration and attempt the update again, but it’s definitely not CUDA.
I think I might have the same issue on Windows. Server is now crashing within seconds of startup and the last lines in the log are always about updating EPG.
