Had some time this morning to experiment with this a little. I turned on crash reporting and debug logging then did this:
- Refresh metadata on an artist with a lot of albums. Got 2 in and crashed. Restarted server. Refresh metadata again. Now got through about 4 albums and crashed. Restarted. Now got through all but the last one and crashed again. Restarted and then got through all albums fine.
- Once I get to that state I can now refresh metadata on that artist with no further crashes…
- Picked another artist with a decent amount of albums and did the same thing with similar results… Crashed about 3-4 times with each time getting farther through the albums. Now refreshing on that artist goes through with no crashes.
Logs attached for the experiment above. Send crash to plex was turned on the whole time as well so not sure if you see anything on our end also.
