After more testing, it looks like the failed upload is creating the file in ACD, so subsequent retries fail because of name collisions.
If I manually trash the file, the backup can continue.
I guess the right fix would be for acd_cli to update its file cache if the file has been created, but could you add in the duplicity backend an "acd_cli trash <file>" as a workaround?
After more testing, it looks like the failed upload is creating the file in ACD, so subsequent retries fail because of name collisions.
If I manually trash the file, the backup can continue.
I guess the right fix would be for acd_cli to update its file cache if the file has been created, but could you add in the duplicity backend an "acd_cli trash <file>" as a workaround?
Kind regards.