Recovering cloud space

I use the CLI version to make backups to B2 storage. The size of what is stored on B2 keeps on growing, and I’d like to prune some files from there.

I found the ‘prune’ command, which I think removes ‘backup revisions’ from the database itself. But what is not clear to me is if this is also meant to permanently remove files from the cloud storage.

If not, is there way to do this other than deleting the repository and making a fresh backup?

Pardon the ‘stupid’ question!

Yes, it will remove chunk files that are not referenced by any revisions as well. You can specify retention policies with multiple -keep parameters.

For example

duplicacy prune -keep 31:360 -keep 7:90 -keep 1:14 -all

Will keep monthly snapshots for those older than a year, weekly snapshots older than 3 month, daily snapshots older than two weeks and all snapshots younger than that. -all means do that for all snapshot ids in the storage. Alternatively, you can specify a specific snapshot id.

As a result of deleting the snapshots some chunks may become unnecessary — if that bits of data no longer part of any remaining snapshots — and will be deleted in the end.

Thank you for clarifying. Am I right to understand that a chunk file is one of the (in my setup set at) 50MB files that contains actual data being backed up?

Is there a way to see which chunk files are associated with which revision?

Yes, the encrypted and hashed piece of the original sausage. If the chunk is not referenced by any snapshot revision — it will be fossilized and then deleted.

more details

Fantastic. Thank you for the help!

1 Like