Hey guys. I'm a fairly capable Linux admin, but my data storage game is weak and I just had a data loss event, so I'm looking to step it up.
I'm interested in moving from ext4 to one of the more modern filesystems, and have a policy of always using the option in the kernel when there is one, so I'm going with btrfs. I'm starting the learning on a homelab server so no warnings about btrfs' prod readiness are needed, thank you.
Now, installing and using btrfs is well documented, and that's not what I'm asking about. What I'm wondering is, how are you guys laying out your schemas? Are you using RAID? How are you automating snapshotting? Where are you storing your snapshots? Is / on btrfs?
If you care to speak to my specific setup, it's got a set of 3 3TB drives, from which I only need 3TB of storage. My current setup is 2 of them in a RAID1 array, and the third rsyncing their data over daily. This has a huge set of problems, mostly to do with KVM images, which I'm handling by just not backing those up at all (?!).
I've heard people say that btrfs makes RAID unnecessary, but I don't understand how this can be. Sure it prevents bit-rot, but I didn't think it did anything to protect from drive failure. Can anyone shed light on this claim?
Thanks for the input. This is the best community ever!
Edited for clarification
[link] [comments]