ZFS can automatically repair checksum errors if the drive with the error is part of a pool set up with redundancy.
What about the case when the checksum error is on a pool without redundancy, but there is a replicated snapshot elsewhere? Can the broken block be copied with zfs send or similar?
For example, let's say there is a primary pool tank that has automatic snapshots and replication to a backup pool backup. Think of backup as a single drive. A scrub of backup finds an unrecoverable checksum error. The valid blocks are still available on tank. Is it possible to recover this checksum error from tank?
This manual file copy method is the closest I've found, but it leaves much to be desired. Is there a cleaner way that can be automated?
