NAS server OMV 5.0
Duplicati 2.0 runs in Docker
Backup via ssh
backed up files approx. 2.9 TB
Linux Debian 10
I have a problem restoring a backup.
I made a backup of my hard drive five days ago. A complete backup was then completed after just under 3.5 days.
Since the “old” hard drives were defective, I bought new hard drives and wanted to restore the data today. I went to restore - direct restoration of backup files. After what felt like an “eternity” a new database was set up and I could see my saved data. Unfortunately, when I wanted to initiate the recovery, my workstation computer “ran out”.
I then logged in again and found that there was no database and performed a restore - direct restoration of backup files - again. After a significantly longer “waiting time” than before, the restoration ends without displaying any data or paths. On the backup computer, however, I can see that all of the “zip files” are obviously there.
I am now a little at a loss as to how I can restore my data.
When you do “direct restore from backup files” it’s going to create a temporary database for just the one restore operation. It’s ok for one off restores on temporary machines, but if this is your primary machine whose data you are trying to restore, I would instead create a backup job but don’t schedule it. Then go to the database section of the job and click Recreate. This will still take a while but the database will be “permanent” and you can perform multiple restores more quickly. To do the restores just click on the job and click Restore. (Don’t use the “direct restore method” …)
Once everything is restored, you can reconfigure the job (make sure the backup selections are set right) and turn on scheduling.
You tried the direct restore twice the same way, and got different results? Or was there a run difference?
I don’t know how hard it is to trace what’s where, but does that include the drive Duplicati’s DB was on?
If the Recreate isn’t working well for some reason (and a few reasons might be below), is old DB there?
If you have the backup files, it should be possible to recreate the DB. Ordinarily a direct restore builds a “partial temporary” database, so is faster, if you asked for all the files, I’m not sure if that would benefit…
Do you recall how far right progress bar moved? The 70% - 100% range (especially 90% - 100%) range mean Duplicati is doing more than usual. You could also watch About → Show log → Live → Verbose.
For larger backups, increasing blocksize helps performance. The default 100 KB is a whole lot of blocks. Unfortunately, this can’t be changed on an existing backup, so your DB Recreate might run more slowly.
Or a regular Recreate might work, and life will be good. You have some things to watch while waiting too.
If it winds up giving an error, look around well, e.g. <job> → Show log and About → Show log → Stored.