I’ currently thinking about a backup strategy for my linux server. I have several subversion repositories und home directories with imap maildirs that I definately don’t want to loose in case my server disc crashes. Actually I have been very lucky so far with not backing up most of my data for years. Maybe that’s why I feel, that I have to do something in that area.

Anyway, I have about 1-2GB of data, most of which doesn’t change a lot, but it should be stored daily. I’am looking for some kind of network backup solution for storing backups on my workstation or friends’ servers on a VPN. While searching, I came across DIBS: Distributed Internet Backup System which uses a peer to peer approach to store recovery data on serveral distributed servers. From the manual it looks promising and quite mature but I’m not sure if I’ll find enough reliable servers with a decent internet connection to use it.

Bacula looks really interessting but might be overkill in my situation. I like the fact that it provides clients for different OSes so that you can use it as a single backup solution even for the windows clients in your network. It is mainly built for larger networks with distributed backup and storage servers for a large number of clients. I don’t think I’ll invest the time to set up the different server components for this one, though.

Possibly I’ll end up with a tar, scp or rsync solution that copies backups to my windows workstation or other peers in my VPN. linux-backup.net and this IBM article of automation of backups on linux give some useful example for such a do-it-yourself backup solution.

Update: I found one more popular opensource backup application which is called Amanda which provides a similar feature set like Bacula.