The docs say: “Restic supports storage and preservation of hard links. However, since hard links exist in the scope of a filesystem by definition, restoring hard links from a fuse mount should be done by a program that preserves hard links. A program that does so is rsync, used with the option –hard-links.”
Is this correct for NTFS on Windows as well as filesystems typically used on Linux?
How does restic backup process files that are referenced multiple times through hard links?
- Will the data of such files be read at most once per backup run?
- So no multiple reading that would cost additional time?
- And not even a single reading when restic comes across a hard link that is already known (path/name and timestamp) from the parent snapshot?