I tried reinstalling Sonarr and restore from a backup, same problem. I tried reinstalling Sonarr completely and start from scratch, same problem.
Sonarr runs for 24 hours, then this error appears and it starts crashing:
ReleaseModule Episode search failed: Episode with ID 4020 does not exist
CommandExecutor Error occurred while executing task EpisodeSearch: Episode with ID 4020 does not exist
No dice, Sonarr went haywire again today at 18:23, i have uploaded the log so you can see the error.
Anti-Virus have been completely uninstalled just to be sure.
Could it be deluge that causes this crash?
Sorry, i recovered sonarr with the latest backup i have (before the crash), trimmed down unwanted series, and made a fresh log (with debug enabled this time!). I will bet it happens within 24 hours.
Furthermore, Sonarr is not exposed thru WAN.
Basicly what happens is, 2/3 of all series disappears, and it starts downloading everything again (Past seasons, everything)
This is the first indication that somthing is wrong since my Pushover spams me with notifications.
Then the log gets spammed with the CommandExecutor error.
Very strange, there appears to be two things happening as searches are occurring and series are being removed, which leads to series/episodes not being found in the DB.
Nothing in Sonarr will automatically remove a series and the only time a search would happen automatically would be in response to a failed download (which would require the series to be in Sonarr still).
((Debug logs)) will show API requests, which hopefully points to something.
and the only time a search would happen automatically would be in response to a failed download (which would require the series to be in Sonarr still).
@markus101 Actually, no, it happens when a series metadata gets updated during which new, already aired, episodes get added. Which kicks off a search for that episode.
@AtomArne My guess is that something torpedoes the nzbdrone.db-wal file, which would leave you with a potentially corrupted db with old data in it. It would be a good idea to check the Created and Last modified timestamps for those files. And also check the windows Event Log (Event viewer) for possible related errors.
If you’re technically adept you could setup procmon to monitor all IO activity on that particular file (or with a similar tool).
But first you need to make sure that your starting point is a working database with no corruptions. And be aware that the .db and .db-wal file are a pair, change one and everything will be corrupt. This is particularly relevant when doing backup restore operations.
At 17-11-23 18:16:41.3|Debug|PendingReleaseService|Removing failed releases from pending it goes haywire.
You might be right, there is a miss match between the .db and db-wal.
Guess il reinstall Sonarr from scratch, but im pretty sure i tried that.
It flips out, 9/10 of the series it tries to update before the crash, i don´t even own.
No crashes…