I’m looking for a linux program that can mirror a directory to another drive and move a subdirectory in the mirror.
Perhaps git would work?
Syncthing is also an option.
Have a look at rsync-sidekick maybe.
lsyncd does just this. It’s intended use is to sync directories between systems over slow(ish) uplink, but it can work locally as well. It takes some fiddling to set up, but once set up it just does it’s thing seamlessly at the background. However if you’re just looking for a backup solution I’d might look for something else, like a plain rsync script.
This is the best answer. I use it regularly to keep hundreds of TB in sync across nodes. Works extremely well and is pretty much hands off once setup.
But
rsynccan do this fine with--recursive --delete. The mirror will remain an exact replica.I found an answer on StackExchange that refers to a tool called rsync-sidekick which looks like it could achieve your aims.
Edit: There is also another answer on that same StackExchange page referring to a script called rsync-prepare which is capable of working with a remote destination.
Why not maintain a soft or hard link?
I am trying to create a crontab to backup multiple directory in which files change their location, but I don’t know if links would solve it
You can also have rsync exclude specific files, so you could sync the main dir first, then the special subdir specifically
I use RealTimeSync.
It just… Works. I love it. It’s free.
There are plenty of features that could effectively let you do what you want.
Mine is set to sync my server with a backup whenever my main PC is idle for 20 mins. Once it determines how to sync, it lets me review before syncing.
borg does this
Why not exclude the folder you want to move from the initial sync, and sync that folder separately to the final location?
rclone sync might be a acceptable?
It doesn’t move files from one dir to another but it would delete the old directory and recopy to the new directory.
I think rsync has the same functionality but could be wrong
As long as the backup system performs deduplication, it should be able to store the backup data efficiently even if the source files are constantly moved around or renamed.
ZFS snapshots can perform deduplication. There is also ZBackup, which also has deduplication support.
Restic, kopia, and Borg are all pretty good backup tools with deduplication built in, so they might be a good option if you’re doing this for backup purposes?
