How much data is downloaded during the pruning process? This becomes an issue for backends that charge more for downloading, like Amazon or Backblaze.
I know that restic users will prune far less often than they backup. Monthly or weekly are very common pruning schedules for restic users. But using the duplicacy web version scheduler, daily is the least frequent schedule.
For each backup revision, the prune command only needs to download several chunks in order to determine the list of chunks the revision references. Moreover, these chunks are cached in the .duplicacy/cache directory, for each chunk you’ll only download it once if you run the prune command from the same computer.
The prune command also need to list the chunk list on the storage but this should be a very efficient operation.
Sorry to bump a bit of an old thread, but I thought I’d add something that I feel is also helpful. gchen answers by saying pruning is not expensive. That’s great to know, but lance1 mentions you have to prune daily, it’s the least frequent schedule.
I don’t believe this is true! It may say ‘daily’ still, but it’ll only be daily on the one ticked day. So this effectively gives us weekly.
