Problem on a backup. I repair or delete but it doesn't work

I often have a problem on a backup. I repair or delete but it doesn’t work.
Now I have a little bit clearer message:

Recreated database has missing blocks and 1 broken filelists. Consider using “list-broken-files” and “purge-broken-files” to purge broken data from the remote store and the database.

The backup is ~18go and I use:
Duplicati -

I don’t see where to do this action? in the browser settings?
Can you tell me how to do it?


What were previous messages saying? There might be a way to avoid them, or at least seek cause.

Recovering by purging files are the general directions.

Using the Command line tools from within the Graphical User Interface would be easiest way to do it.
You have to edit the default command, which has syntax for backup, to run desired command below.
You can leave most backup options, but clear the source paths from from Commandline arguments.


HI @ts678
Thanks for the answers.
I spent some time, I think I understood but it is not sure because there is a lot to read, it is technical and I do not speak English well
When running in graphical mode with the purge command (Using Duplicati from the Command Line - Duplicati 2 User's Manual) with clear the source paths from from Commandline arguments, the command never ends (several hours see screenshot)

This happens to me sometimes, is there anything to do to allow a command to finish?
Otherwise I give up. I have a backup on a disk (synchronization)
In this case, I will delete all the data from the target (in the cloud) and create a new backup!


There was no talk of The PURGE command, but you linked to The PURGE-BROKEN-FILES command, which I’ll assume is what you ran. The translation of what you showed on the status line might be below:

and Recreate is a variation of The REPAIR command done after a database is deleted. How you got into this situation is a good question, but sometimes recreation can take awhile. You can look for progress at About → Show log → Live → Verbose to see if anything is happening. In unusual error situations, it might download and process all destination files, attempting to locate some block of information that’s missing.

I can’t tell you if this is the case. You’d have to look. Alternatively, it’s possible that no activity is happening.

This is often easier for those who don’t mind losing old history. I’d still like to know errors you were getting. Backup errors often put no log in the usual log area, but you can also look in About → Show log → Stored.


From your screenshot and a ruler (centimeters is easier), it looks like it was doing the complete search for some missing information. This occurs from 90% to 100% on the progress bar. You’re at 95, so maybe it’s about half done 7 hours ago. Maybe it even made further progress and either finished or gave up with error.

Once information is missing from the destination, it stays missing. list-broken-files and purge-broken-files would be how to get limited deletion of only the affected source file backups, but a fresh backup is simpler.

In fact, I don’t know why this backup is not doing well (others don’t have problems with the same provider). Maybe a problem with the remote server.

About → Show log → Live → Verbose:

  • 29 avr. 2021 21:40: The Repair operation failed with the error : The database was attempted repaired, but the repair did not complete. This database may be incomplete and the repair process is not allowed to alter remote files as that could result in data loss.
  • 29 avr. 2021 21:40: Missing blocklisthashes repaired succesfully

As I said, if there is no simple solution, I will recreate a backup because I don’t have important data stored in the cloud.