How is this, from a technical point of view, more advanced than running rsync in a regular fashion (with hardlinks to keep cheap snapshots)?
Just wondering about the technical aspects.
Just wondering about the technical aspects.
Although the application itself is not open source, the developer has open-sourced a restore tool, to help ease the "what if they go out of business" concerns:
https://github.com/sreitshamer/arq_restore
Also, somebody else has created a Go utility that can read the backup format:
https://github.com/asimihsan/arqinator
Finally, if you really wanted to do it all yourself, the data format is documented here:
I had an older version that I upgraded to 4.x because they added support for uploading backups to Dropbox and I was already paying for a Dropbox pro account that had plenty of unused space. The ability to have versioned backups on various cloud services that don't support rsync is a big plus.
if a few years ago somebody said "just copy all your private data to some place on the internet... but encrypt it first" you would take it for the crazy argument it is.
Yet. It can't be read yet.
Short of a vulnerability in AES (in which case we have more problems than a few copies of the Anarchists Handbook in our backup files), cracking proper encryption is simply not feasible.
Have a look at the `--link-dest` option to rsync.
In the larger scope of things, that level of deduplication has never been something I've found all that valuable; but of course most binary files I backup are images and music files; not ones which change frequently (if ever).