Borg supports local and SFTP backups while Restic supports more (a lot more). S3, Google Cloud, B2, etc. In fact, they integrate with rclone so anything you can access using rclone, you can backup to using Restic.

Borg uses compression while Restic does not. Restic just uses deduplication so your backups with Restic will likely be larger in size.

Anyway, that's what jumps to mind. They're both pretty great honestly (in terms of community support and reliability). There are a lot of other options too btw. The Restic repo has a pretty good list [1]

[1] https://github.com/restic/others

The rclone backend is a killer feature of Restic - using it in my home setup which is:

6 computers backing up using UrBackup client to a 6 drive 2U NAS running Raid Z2 and UrBackup server (which is nice as it stores incremental backups as ZFS child datasets).

I then have a post backup script which creates a snapshot of the latest backup, mounts it, backs up using Restic using the JottaCloud rclone backend.

Do you know if Restic supports block-level deduplication? I am backing up some database dumps and the backup filesystem has ballooned in size, even though the dumps are 99% identical every day. This makes me suspect that they only deduplicate on the file level, and that I would have some very significant savings if I zipped the dumps before backing up.

EDIT: Some details on deduplication are here, and I'm puzzled as to why my backups are so big: https://restic.readthedocs.io/en/latest/100_references.html?...

Restics dedupe window is tuned to a 1mb boundary - it's optimised for cloud provider IOPS.

Bup with it's rsync sized one always did much better for databases for me.

That's interesting, thank you. Do you have any info on how to enable that one?