Unmarking paths from backup - deleted?

Support
1

Hi,

I was wondering what happens when you remove a path from a backup… is this then removed from all prior backups, or is this kept until the files expire?

I removed a path with a rather huge amount of tiny files, and it’s been at “deleting unwanted files” for a few hours now.

2

Hello and welcome!

No, the data is not removed from prior backup versions. When you remove a folder from your backup selection list, the future backup versions won’t contain it. The data will still exist on the “back end” until all older backup versions that contained it are deleted per your retention policy settings, and a compaction event occurs.

That being said, there IS a way to delete data from past backup versions if you really wan to. (See the PURGE command for more info.)

It might be doing a compaction. You can go to About → Show Log → Live → Verbose and see what’s going on.

What is your retention set to? What is your “remote volume size” set to (default is 50MB)?

3

I read about this. I’ll have a look at it whenever this finishes.

Unclear… Get’ing and Put’ing and Delete’ing a bunch of dindex* and dblock* files, but it doesn’t really log anything that I can point to as “Oh, it’s that!”

Retention is 7D:1D,1M:1D,2Y:1W,99Y:1M. Remote volume size is default.

4

Almost certainly doing a compaction. If you see activity there, just let it continue doing its job. :slight_smile:

5

Purging a path, would that then be a commandline like

--include /directory/to/purge/

? Or something else?
I find the manual page not that clear… :frowning:

6

I’d be very careful before running the PURGE command as imprecise syntax could have bad consequences. Is it really important to you to purge the data from previous backup versions? If so, here are the steps:

  • Click the backup job in the main UI to expand options
  • Click the blue “Commandline …” link
  • Change the command dropdown to “purge”
  • Leave “target URL” box alone
  • Replace contents of “commandline arguments” box with:
--include=C:\PathToPurge\*

(or on Linux/MacOS:)

--include=/directorytopurge/*
  • If you only want to delete from a certain backup version, add another line to the “commandline arguments” box:
--version=X

(where X is the version # to delete. If you leave this off, the data will be removed from all backup versions.)

  • Finally, scroll down and click the blue “run ‘purge’ command now” button.
7

Well, it’s around 300gb of already deduplicated data (from my VM backup software) that should not have been backed up through duplicati. Multiply that by several versions, so it does add up.

Edit: dry-run reported saving around 450GB, which is a sizable chunk of my cloud storage :wink:

1 Like