Yesterday I updated to v.1.0.5_stable_2025-03-04. I ran my daily backup and all worked as expected. (Win10, backing up to local NAS)
This morning I attempted to restore one file to test and am getting this error
2025-03-27 09:55:31 +01 - [Error-Duplicati.Library.Main.Controller-FailedOperation]: The operation List has failed with error: database is locked
database is locked
SQLiteException: database is locked
database is locked
I re-ran the backup, that is still working, it’s the restore that gives error.
I have another backup job which runs monthly, I did not run a manual backup for this one - tested restore and that works. I’m wary of running a backup on this job in case it also locks the database.
I ran a database repair which reported as successful. But I still cannot restore files.
The error message seems to indicate the something (most likely Duplicati) is locking the database.
I do not understand how the backup can run and the restore fail, if they are using the same database. (It is either locked or not locked).
Can you elaborate on how you are running the backup and restore? Are these for the same backup (and using the same database)?
A quick-fix might be to just reboot the machine as that will release any locks taken, but of course this should not be needed.
After reboots, closing duplicati processes and lots of testing it seems that:
During the restore process (triggered from both the "Restore’ option in the mail menu and via the backcup job) the backup times-out when attempting to fetch path info. This locks the database which, after a delete & repair, has become huge (from 650mb to over 900mb - source data around 70Gb) and either resolves if you leave the machine alone for a wee while or stays locked until you close/kill duplicati. The times that it resolves itself, the restore works - eventually.
So I have just set up a new backup job using the same config just updating the destination folder and a new database and that works perfectly.
I will keep the old backup for a month or 2 in the unlikely case I need to restore something and continue using the new one as my daily backup.
Yes! If it was just super slow it would be OK, but half the time the database does not unlock and error “getting file list failed” and then " Unable to find scope in lookup table, this may be caused by attempting to transport call contexts between AppDomains "
The NAS I use is old and slow & the database large so I assume that doesn’t help matters.
This backup has become my old, keep it just in case backup and I’m using the new replacement one with no issues.
Ok, I am aware that some combinations of search are super slow. We will be looking at that shortly.
That issue was fixed in the canary builds (but not the speed issue unfortunately). You can try the canary build if you like: Release: 2.1.0.112 (Canary) 2025-03-26