Prune isn't cleaning remote storage

Hello everyone,
I’m trying to set up Prune in the Web Edition, so after a few days, old backups will be removed from my remote storage. in my case i’m using OneDrive. Sadly when i trigger a Prune with the command -all, on my remote site no data is removed, i’m not getting any storage back. (see image)

What am i doing wrong?

Duplicacy implements a two-step fossil collection algorithm to remove old snapshots. See the explanation here: Lock free deduplication algorithm

The first step, called fossil collection, deletes snapshot files and marks unreferenced chunks as fossils. This step doesn’t reclaim much space. The second step, called fossil delete, will actually delete fossils.

In your case wait until the next run to see if any space is reclaimed.

What do i have to add to my webui to make it happen that the backups older than 3 days are removed automatically?

You can add the option -keep 0:3 to the prune job.

the option -keep 0:3 adding to the prune job, isnt working. now my onedrive is full, im not able to upload backups anymore, (check screenshot)

duplicacy is saying that my backup is 829GB, and my recycle bin on onedrive is empty, so the prune job isnt working. i wanna buy duplicacy, but this isnt working properly…

Run the prune a second time, and it will clear the data. Prune runs in 2 steps, that’s why the first time nothing was deleted.

If running prune a second time didn’t work, you may have to run it with the -exclusive flag, making sure no backups are running while it does.

This is because for chunks to be deleted, the first prune collects prospective chunks to be deleted, then waits for at least one backup to be completed for all repositories to be sure they’re no longer needed (any that are, are either resurrected or recreated I’m not sure which).

The second prune finally deletes those collected chunks.

1 Like

Running a prune job with the -exclusive flag should release some space immediately. If you want to keep this flag for future runs, move the prune job to the same schedule as the backup jobs. This is because this -exclusive flag requires that no backups can run while the prune job is in progress.

now i changed my backup schedule to something like this:

hopefully it works :slight_smile: