What are the ramifications of editing the snapshot id in the preferences?

Support
#1

I’ve just switched from Ubuntu to Manjaro. Currently planning to keep running the backups I’d already defined (CLI duplicacy). But that leaves me with one minor niggle - the names I used originally contain the OS name eg
"id": "ubuntu_home",
so I’d like to rename it eg
"id": "manjaro_home",

But its not clear whether this will create an entirely new backup set of snapshots or build on the existing. Also not clear whether I have to change anything on the server.

I’m backing up over sftp to a single storage shared between several computers, and multiple snapshot ids on each computer.

2 Likes
#2

You can change the snapshot id for the repository in the preferences file, but this will start a new chain of snapshots under a different id. However, that’s not a bad thing, as the chunks already shared between the id’s will be mostly identical and you’ll benefit from de-duplication.

Your first backup under the new id will take longer as it will use the -hash option.

What you can’t do is change the id’s at both ends, so as to maintain a continuous history under one id - the snapshot revision files in the storage e.g. 1, 2, 9, etc… have the id hard-coded in the metadata, so you can’t rename the snapshots/id directory in the storage.

So if you want to keep those revisions around, I’d just let them gradually expire during the regular prunes and just keep in mind those older snapshots were created under a different id, if you ever need to do a restore.

The only other thing to note, is that prunes won’t consider collecting fossils for chunks referenced in those old snapshots, until 7 days have elapsed - because you’ll have stopped creating any new backups under that id, and Duplicacy is cautious not to delete fossils that may need resurrecting. (See last paragraph.)

4 Likes
#3

Thanks for clarifying. I’ll change the ID’s :slight_smile: