My entire business (and thus my life) resides on my computer, so I have designed my backup procedures so that, in case of a hard disk failure, my critical data files are protected and I can get the operating system and my applications up and running ASAP.
I have two 120 GB hard disks, each partitioned into two logical drives. Drive 1 is split into a 20 GB partition ( C: ) and a 100 GB ( E: ), and Drive 2 is 100 ( D: ) and 20 ( F: ).
The operating system (Win 2000) and my programs are on C, and all my data are on D. (BTW, there are operational advantages to putting data on a separate physical drive from the OS and programs: things tend to run a little faster.) Most applications can easily be configured to store their data on a different drive, and for those few that can’t (MS Outlook), I have written a batch file that runs automatically at midnight every night to copy those files (and certain other critical files) to the D partition.
At 1 a.m., Windows Backup backs up the contents of the D partition (drive 2) over to the E partition (drive 1), running a full backup on Friday nights, and incremental Saturday through Thursday. The mix of one full and six incremental backups of the data fits (so far) onto a partition the same size as the source.
At 4 a.m. every morning, PowerQuest Drive Image 7 makes an image of the C partition (drive 1) onto the F partition (drive 2). As danceswithcats says, this program has to run at boot level, so the system reboots into DOS mode, makes the image, and reboots to Win2000 when it’s done.
Note that files from one physical drive are being copied to the other. This makes the process about as fast as it possibly can be. It takes about 40 minutes to image the C partition (about 7GB). I have no idea how long it takes to backup the data drive these days (about 40 GB), but I don’t care, since it’s happening while I sleep.
The intent is to guard primarily against hard disk failure, and my assumption is that it is highly unlikely that both drives would fail simultaneously. But I am not protected against loss of the whole machine to fire or theft. (These are just not risks I consider likely enough to be too concerned about. YMMV.)
I use Drive Image for the C partition because if Drive 1 fails I will want to get a new HD and restore the OS and my programs all at one fell swoop. Drive Image will do that quickly and efficiently, without making me spend hours or days reinstalling and reconfiguring the OS and dozens of programs. Backup won’t.
But I use Backup for the data because more likely than having to restore the whole drive is the possibility of needing to restore a couple of data files that have been accidentally lost or corrupted. Drive Image doesn’t work well for that. (It says it will, but it won’t. Trust me on this.) The advantage of Backup over ordinary copying is the ease of doing incrementals and the slight compression it offers.
The batch files, Backup, and Drive Image (along with Norton AntiVirus and a few other housekeeping programs) are all started automatically each night by Windows Task Scheduler.
Every week or two, I copy the C image from F onto a rewritable DVD, just for an extra level of protection. (I haven’t bothered copying the data onto DVDs, since it would take about ten of them.)
I’ve been using this method for more than a year, and although I haven’t had to restore either drive so far, the process is flexible, easy-to-use, and secure, and protects me from what I feel are my most likely disaster scenarios.