Incremental backup script for CEPH Volumes

[Norman Uittenbogaart]

I am making a backup script for Ceph volumes.
It will make a backup of the rbd volume, then it will continue making diff backups.
This will make it a lot easier to store offsite, as you only have to transfer the main backup once.

The next step will be to merge the differences after a month orso on both sides, so you won't get 1000's of diff files.

I'm not a programmer or something, so any comments are welcome

Here you can find the script,
https://github.com/normanu/rbd_backup

 

[MACscr]

How has this worked out for you? How are you creating the rbd snapshots in the first place? Right now I only have a single ceph pool, so definitely looking for way to do more efficient backups.

 

[robhost]

Looks interesting. Wondering how the merging could work

 

[MACscr]

It hasnt worked well for me. Not sure if its an NFS thing or what, but it never seems to finish. Maybe I just limit its loop to one so that it only tries it out on one small vm first.