Resume backups after server failure

Hi there

First of all, THANKS RESTIC <3
My server died and I was able to recover more than 3.5TB of data from my remote repository in S3 to a new server.

My question is: now I have the recovered data + some new data, and I would like to resume my daily backups, nevertheless I’m not sure how to proceed.
When I list the available snapshots, there is a column called “Host” with the name of the previous server.

May I find troubles if I just launch a new backup job using the same repository?
Is restic smart enough to detect that the data is the same or it will start the a new snapshot from scratch?

Thanks for help!!

No, nothing in particular.

Every backup run is intended to create a new snapshot, so that is what you want. It will do that, but it will only transfer/upload metadata and parts of files that actually changed, so it will not have to upload it all again. Just go for it, run the backup :slight_smile:

Not the same thing, but related, just FYI: FAQ — restic 0.15.1 documentation

OK, great to know.
So, just to clarify, even if the “Host” is different (in this case different machine and different hostname) the new snapshot will only transfer the parts of the files that have changed. Right?


Restic (and you) are awesome :smiley: .

Thanks for the super fast support!!

1 Like

It’s only quick when you ask questions I know the answer to :rofl:


Now that your new backups will have a different “host”, you may need to keep this in mind when performing forget, as restic will consider snapshots with different hosts as separate groups. That is to say, if your forget keeps the last 100 snapshots, it will now keep the last 100 for the previous host and the current. If you wish forget to ignore the host field, use --group-by paths when running forget. (The default is --group-by hosts,paths).

1 Like

You might want to run your backup with --dry-run the first time to see what it will do.

Or use restic Tags to tell prune only this tag.