Hi all,
I’ve been using Duplicati for a couple months and I’m quite happy with it. I mainly use it to backup docker containers and their data. While configuring the backup, I included all my docker volumes which also contain the Duplicati files and databases.
Does it somehow make sense to backup the sqlite databases of Duplicati ?
I do. I have a primary backup job that backs up all my data. I then have a secondary backup job that backs up the sqlite database of the primary job. The thought is that if my primary sqlite db gets corrupt, I could restore it from the secondary backup. It should be faster than doing a full db rebuild of the primary database.
I run Duplicati in a docker container. I do not bother saving the sqlite DBs, as they can be regenerated from the remote backup files, and I would not want to run into consistency issues in case the sqlite DBs and the remote backup files do not match – which would kind of defeat the purpose of sqlite backups, since their value in a recovery scenario is basically zero.
I had to recreate the sqlite DBs once, which is a mild nuisance, but very manageable.
I do, however, keep a copy of all the backup configurations in a Keepass container, as well as the command that was used to create the docker container (so I wouldn’t lose volume mounts and so on), together with the credentials/paths to the remote backup files and a short description what’s included in the backup.
This Keepass container is backed up by Duplicati itself, automatically sync’ed between all my machines and my NAS, and periodically saved to a bunch of external media (read: hard disks, that are also physically spread). Losing that would be a catastrophe, which is why I’m extra paranoid – but that’s just a few KB in size.