The main error that is concerning me is the “database schema has changed” and “slow query”.
The only real problem I’m seeing with the application is that PlexPy/Tautulli is having issues connecting even tho it says it is connect to it.
These message relate to the database. The “schema” issue is usually when PMS cannot update the database to the latest. The “slow query” means the DB is responding slowly, which is typical if you have not run the “optimize database” in a while.
Can you restart PMS, don’t do anything for 3 minutes, then pull a new log?
I have ran Optimize Database a few times over the past couple weeks, even today before this log file. But here’s another log file after a PMS restart. I am seeing slowness in library scanning also.
That log does not look like what it should after restarting PMS. Are you sure PMS stopped before you restarted it?
Here’s a fresh one. deleted the log before starting.
Would it be possible to get the entire set so I may see if there is anything else to be seen? (settings - Server - Help - Download Logs)
To add to the other feedback you had,
Logged lines like these are normal behaviour and expected
Mar 14, 2018 15:42:02.065 [0x7f0ce03ff700] INFO - SQLITE3:0x10, 17, statement aborts at 57: [select * from metadata_items limit 1] database schema has changed
Mar 14, 2018 15:42:02.071 [0x7f0ce03ff700] INFO - SQLITE3:0x10, 17, statement aborts at 57: [select * from metadata_items limit 1] database schema has changed
Mar 14, 2018 15:42:02.076 [0x7f0ce03ff700] INFO - SQLITE3:0x10, 17, statement aborts at 57: [select * from metadata_items limit 1] database schema has changed
They arise during a database schema migration or after an optimize of the database - no need to worry about them
However, these could be a problem with some transactions taking several seconds
Mar 14, 2018 15:40:46.488 [0x7f0ce03ff700] WARN - Held transaction for too long (../Library/FullTextSearch.cpp:17): 9.200000 seconds
Mar 14, 2018 15:40:49.145 [0x7f0ce03ff700] WARN - Held transaction for too long (../Library/FullTextSearch.cpp:27): 2.390000 seconds
Mar 14, 2018 15:40:56.856 [0x7f0ce03ff700] WARN - Held transaction for too long (../Library/FullTextSearch.cpp:38): 7.640000 seconds
Mar 14, 2018 15:40:59.574 [0x7f0ce03ff700] WARN - Held transaction for too long (../Library/FullTextSearch.cpp:48): 2.590000 seconds
In your second log - after the optimize, there were two warnings for slow transactions over a second each - retrieving 50 items in each case
Mar 15, 2018 15:02:35.848 [0x7fcdb83fe700] WARN - SLOW QUERY: It took 2360.000000 ms to retrieve 50 items.
. . .
Mar 15, 2018 15:02:38.276 [0x7fcda77ff700] WARN - SLOW QUERY: It took 6780.000000 ms to retrieve 50 items.
Unfortunately you have disabled debug logging - so cannot tell what these queries related to
So before you upload the full set of logs, please enable debug logging on the server (settings / server / general / show advanced ) and restart the server so there would be complete debug logs from launch and if you continue to have issues, upload zipped logs and indication of when the issue arose.
If you get a failure to connect reported by Tautulli - please indicate what time so logs can be checked. A database optimize action may lock the database for many seconds
I see you are using Plex Relay - that would not give you high speeds and need to get remote access to work so it does not switch to the slower relay backup feature
Here’s the log download for ChuckPA
I had hoped you’d enable DEBUG as SA2000 had requested. That information is needed so we can see what PMS was doing prior to the problems.
Do you mind enabling and then recreating?
