One of my backups has gotten corrupted. I’ve tried various repair, and purge-broken-files that have failed various ways. What I want to do is get rid of all the remnants both local and backend , and just recreate it with the same configuration.
There may be multiple ways of doing this. One I can think of is through the retention settings, but that might only get you down to a single backup which is not quite good enough here.
I think the most “correct” answer is to use the “delete” option. See this link for syntax. You can also do this through the GUI using the same syntax guidelines. Note the “–allow-full-removal” option that you’ll need.
Also note that the “–version” option can take ranges and a comma separated list. So something like “–version=0-5,10” will remove versions 0, 1, 2, 3, 4, 5 and 10.
Assuming you don’t mind keeping the same configuration (be sure to turn off its schedule), local is Database managementDelete button, and backend is delete the duplicati files however you wish.
If it’s hard to get there somehow, you can export the job, delete it and say to delete the destination.
Then import the export to backup again. The delete-all-versions plan works too, but may be slower.
What broke anyway? We’re trying to reduce those, and helping to know what happens helps target.