I'm trying to figure out how to back up S3 buckets when using Rook-Ceph. In the past, with Minio or RustFS, I could just use rsync or restic to back up the data directory, but I'm not sure that's an option here. I'm curious to see how others approach bucket backups in this setup. Any suggestions?
5 Answers
Using rclone in a pod essentially mirrors how rsync or restic work for backing up data. Just set it up to target the RGW S3 endpoint instead of the backend filesystem, and you’re good to go!
I actually do this by using rclone! I have a cronjob that runs nightly, syncing my S3 buckets to a remote storage location without any hassle. It works great!
That's awesome! I didn't realize you could use restic alongside rclone. I only used restic for file system backups before.
We’ve had great success with Plakar.io. It’s an open-source tool that’s efficient and super easy to integrate into our setup.
Instead of relying solely on RGW for buckets, we switched to using garage-operator. It allows us to have everything backed by PVC on Ceph RBD, making backups with Velero straightforward. Much simpler than juggling two separate backup systems!
One approach we use is replicating to a second S3 provider at a different location. It's a solid strategy for redundancy.
I think the OP is referring to backing up the direct file system from a single server. With Rook and Ceph, you need something that can work with the S3 interface, which rclone does nicely.