I have been backing up with Duplicati for a while and recently I was going to replace a modular PSU. Obviously the existing cables fitted perfectly to the new PSU but it was plugged in a different manner, so I burned 8*3TB WD Red hard drives.
Of course that was annoying as h*, even if the insurance covered a lot of it it does not recover the files.
So I had my files backed up with Duplicati on a USB HD and in the cloud. So I went ahead with the restore, one for smaller file sizes, which is just some few GB large (text documents etc). That went smooth and as expected. But then I tried to restore my large file backup, with a total size of 1TB (photos and videos). This started as expected but it ended up stalling at “Recreating database” as the local one was blown away with the hard drives, so I waited… several days and the progress bar barely moved at all, but it did move. I got mad and googled the forums like crazy and saw this was a rather common thing, so I tried multiple things I found but to no help. My backup uses SMART retention, so the versions was at about 9. I just wanted to get all my recent data out asap (version 0).
At the end I ended up starting the backup, waiting until it’s stalling point at the progress bar (probably something like 95% - until it stopped moving for about 1 hour), then I cancelled the operation and just waited a bit. And then it suddenly started the restore process, so I thought I just wait and see what will happen. Maybe I just get garbage, or maybe wrong version or files. Maybe just a partly restore.
As it turns out though, is that it seemed to have recovered everything - at least I can’t detect anything missing, and that took just like ~1 day, not 5+. Right now it does verify the files it just restored, but for what I’ve seen browsing through the files manually it looks like everything is there with file permissions and all.
Can anyone explain what might have happened here? Or what might happen if I don’t wait fully for the entire database recreation? Is “this happening” an option that could be implemented somehow in Duplicati for a -way- faster restore in the future? Like “restore all latest version” and the just create the database based upon that, I dunno what’s exactly happens in the background - but maybe it can help someone desperate restoring their data.