I have just implemented a different approach to backing up 2 very large volumes on my main file and print server. The "D" and "E" drives contain roughly 750 gigs each, and the full backup on Fridays was brining the server down to a crawl. I wrote a script that creates a snapshot of the drives on the storage array, mounts them to a dedicated backup server (just a file system client currently, but soon to be a dedicated 64bit media agent), runs the backup and finally dismounts the snapshot volume. Everything is working great and the speed is nearly 3 times as fast. I ran a full backup, with the snapshot, on Tuesday, 2 incrementals on Wednesday, and incremental on Thursday and 1 last night before the synthetic full. I was surprised to see this morning that the size of the synthetic full was only 205.63 gigs as opposed to the size of the true full backup from Tuesday which was 722.04 gigs. Does this make sense? I figured they would be roughly the same size or at least a lot closer than they appear to be.