Interesting case: shadow directory

Hello,

I have an interesting use case.
I’m backuping a veracrypt container - mounted to /mnt/container - to an USB drive, /mnt/usb.

The container is roughy 100GB.

The initial backup is ok.
But, the computer is rebooted, and for any reasons, the container is not mounted when Duplicati starts the backup. I have a 0 size backup, and when browsing that date, i have no file to restore, which seems normal.

Then the container mounted correctly. What would be the impact on backup size, and also on database size ? Will it find already backuped files, will it start a differential backup from previous one, and in that case, what would be the better way to not start the backup ? (i think i pre-event script checking a dummy file that would be in crypted container, or something similar).

Thank you !

Thank you !

If you look in your log that backed up 0 files, you’ll probably see all prior files marked as Deleted files, e.g.

image

When they show up again, they’ll probably be Added, but any blocks of data will likely be around unless retention keeps very few versions. Any existing version still in the backup should be able to keep blocks.

Basically, if there were no other changes (unlikely to be quite so good), block-level deduplication should prevent new block uploads from the files, however deduplication does need looking files over to be sure.

gives you some options on what to do:

1 Like

Thank you very much. I’ll adapt it to my problem (or soon to become one :p)