I have a several clients that do not use a hierarchical backup strategy and just backup there data to a single external drive on their servers. That is fine until the customers says to me ,"I need a documents that I modified 2 weeks ago that had some important data." and my response it,"unfortunately your current backup system on allows you to retrieve what was modified the previous night and is not a historical backup system." basically I tell them they are "SOL". Many clients are going through hard economic times and on a fixed budget. My recommendations are lost cost NAS devices

http://www.buffalotech.com/products/...n-es-ts-xelr5/

that have multiple RAID (0,1,5,10), the "0" ??? and just create scripts that backup the appropriate data to the NAS. I would like to implement a historical backup system that is cost effective and have the NAS devices be the repository like a SAN. Any suggestions?

Not really sure about the NAS solution but, this might help with a cheap historical backup solution.

http://www.mikerubel.org/computers/rsync_snapshots/

It has a lot of good ideas and explains things pretty well.

That mikerubel link is the historical reference on how to do that. But there is also open source software that implements those details. Check out rsnapshot, BackupPC, Flyback for Ubuntu, or the Writeup at Backup Central.

A little time with google will turn up more.

A different approach would be to implement Amanda and backup to the NAS or SAN. That is also open source, a relatively small footprint on your server, and offers lots of options for backup strategies.