There is a rule that you should regularly test restoring your #backups.
Do you have any processes/recommendations to do this in an automated(!) way? For more machines? Ideally w/ @restic?
What does "restore works" mean for you? Backup repository integrity checks? Successfully extracting a certain subset of files? Extracting all files (which takes quite some time/disk space)? Comparing checksums/contents of certain files? (But what about newer versions on the backup source?)
@frederic That's a good question, which I'm also asking myself right now.
I think I'll program a little script.
I also use #restic, but in addition to that, I use #resticprofile.
There is a configuration parameter called “run-after” as a command hook.
I'll give that a try when I have a little more time.
https://creativeprojects.github.io/resticprofile/configuration/run_hooks/index.html
Run commands before, after success or after failure resticprofile has 2 places where you can run commands around restic: commands that will run before and after every restic command (snapshots, backup, check, forget, prune, mount, etc.). These are placed at the root of each profile and are always considered. commands that will only run before and after specific restic commands. These are placed in supported sections of your profiles (currently supported are backup, copy, dump, find, ls, mount, restore, snapshots, stats and tag). Here’s an example of all the external commands that you can run during the execution of a profile:
@Kaffee_Junkie42 I'm using resticprofile as well.
So far `restic check --read-data` or `--read-data-subset` (mentioned by @unixtippse) seems to be a great fit:
https://restic.readthedocs.io/en/latest/045_working_with_repos.html#checking-integrity-and-consistency
I'm also using resticprofile's Prometheus integration to alert for failed backup runs and stale backups (with no added/changed files in a while).
Frederic
Rainer