Server Version#: 1.18.2.2058
Player Version#: Irrelevant
I have been ripping my blurays, and the last couple of days when adding the movies it doesn’t attach an image automatically. This is what is showing in the logs:
Nov 25, 2019 18:57:30.416 [0x7f54c7447700] DEBUG - Item 314703 (Resident Evil: Afterlife) Scanning metadata graphic elements in XML file ""/Info.xml Nov 25, 2019 18:57:30.419 [0x7f54c4567700] DEBUG - HTTP requesting GET http://127.0.0.1:37114/system/agents/media/get?guid=com%2Eplexapp%2Eagents%2Eimdb%3A%2F%2Ftt1220634%3Flang%3Den&mediaType=1&url=metadata%3A%2F%2Fposters%2Fcom%2Eplexapp%2Eagents%2Eimdb_2fa7906e1c449ff465b48ca1dd268c856e4b0be5 Nov 25, 2019 18:57:30.431 [0x7f54c4567700] DEBUG - HTTP 500 response from GET http://127.0.0.1:37114/system/agents/media/get?guid=com%2Eplexapp%2Eagents%2Eimdb%3A%2F%2Ftt1220634%3Flang%3Den&mediaType=1&url=metadata%3A%2F%2Fposters%2Fcom%2Eplexapp%2Eagents%2Eimdb_2fa7906e1c449ff465b48ca1dd268c856e4b0be5 Nov 25, 2019 18:57:30.432 [0x7f54c7447700] ERROR - Error bringing media local (metadata://posters/com.plexapp.agents.imdb_2fa7906e1c449ff465b48ca1dd268c856e4b0be5)
The rest of the metadata adds fine, and the movie is properly matched, but for some reason the image doesn’t attach.
I am also able to open it up in edit, select an image and apply it just fine as well. Just not automatically.
My db got all corrupted for some reason (when I added a bunch of movies in close succession), and I had to revert the database back a day. Here are more logs to either help or hinder.
Loading a lot of anything all at once (more than 1000 at a whack) will make ANY synology struggle and nearly roll over on its back.
Adding tons (5000+) of images or music will outright kill it (non-responsive) for days on end.
a. Images get processed one by one
b. Music files get fingerprinted to find the album match.
Recommendation:
whatever section currently being added – Stop and delete it
If possible, partition the addition of new media in chunks. There is no limit on the number of folders or sub folders which may be added to a library section.
Observations on a DS1815 (Atom C2538 CPU):
700 movies, all perfectly named & structured takes about 12 hours due to the chapter images. This can be shortened considerably if chapter index image thumbnails are not desired.
3000 songs, all perfectly named and structured, takes about the 6 hours.
No problem, I was only sharing observed practical limitations of performance on something as small as a Synology DS1815+.
Database corruption indicates something else is happening.
PMS has a normal signalled ‘shutdown request’. It only takes a second or two to shut down. If, however, the database is extremely busy, or the host CPU very loaded, it may not shut down cleanly before the host terminates it . Synology waits 5 seconds from signalled shutdown until forcing it down. If this is happening then we have the root cause.
I’m running a 1019+, and as far as I’m aware the Synology didn’t shut down during that time.
I also haven’t seen the CPU under full load, and received no performance alarms.
I have however suspected that if you manually click “scan library files” more than once, this has coincided with database corruption in the past.
My Plex database is about 3 or 4 years old, and was transferred from a PC. I ran SQL statements on the DB to update the media paths, as to avoid rescanning and it all transferred over very smoothly.
The NAS isn’t used for anything other than storing my media and acting as a Plex Server.
Have you made any pragma adjustments to the database?
I’ve seen cases where some have tried, with the best of intentions, which resulted in database loss. It happens because there are collation methods in Plex’s version of the SQLite database which aren’t in the public utility. It causes corruption of some of the tables and leads to subsequent failure.
If you’re having issues with it becoming damaged by simply hitting scan a few times, there is already something wrong because PMS only locks as it updates individual records. A simple scan won’t cause it.
