I do something similar, but more automated.

I have a bunch of NAS servers setup. None of them run RAID of any type whatsoever. They are all single drive setups.

All my desktops run SecondCopy (Windows) or rdist (Linux) to a NAS server each day to back up designated files to one of the NAS volumes. They do partial backups every day, and once per week a full backup.

Then, the NAS server(s) rdist the files to volumes on different physical drive(s) each night. There are dedicated backup drives that are spun down except when receiving the backup.

This has it's benefits.

(1) No trying to recover a RAID-anything when a drive fails. I've seen RAID sets logically corrupted due to a bug, etc. I'd rather have the drive fail and have it mirrored somewhere else and simply have to re-mount a filesystem.

(2) In the case where I hear someone in the other room yell "Oh, crap! I just deleted file YYY, now I'm hosed!" - I can always to one of the servers before the nightly backup and restore the file.

(3) The backups can also be done on a weekly basis, in a cascaded effect. So, each desktop backs up each 24 hours. Each NAS server duplicates each 24 hours. Each NAS server can also duplicate each 7 days to somewhere else. Ultimate redundancy.

The only thing this doesn't cover is catastrophic failure of the entire house. Still working on an off-site method for that (I'm thinking to setup a NAS server at my BIL's house and have them sending an encrypted backup automatically to each other weekly.)

This may seem like overkill, but it's very brain-dead simple to setup and very robust, gets the job done, and my data is not just backed up, but also layered so I get some sort of recoverability for the once in a while user-induced "oops!"

Broken does not imply that it doesn't work at all. Broken means that it doesn't function properly. It is not an absolute term.