I would have left the CDR on another box and just pulled a drive image off the production server to back it up... less overhead on the production box. I wonder if your production server will have enough free resources to actually burn a CDR without errors.
Of course, if it was me I'd have redundant RAID arrays hooked up with a failover and I'd do a FlashCopy every 6 hours plus tape backup off the second array... but that's just me, I'm compulsive like that. Somebody shoot me.
So sorry. But hey - good news - a friend might be giving me two 60GB 7200 Rpm ATA/100 drives, which I *think* I can use to set up RAID. I'll test on my staging server and if it works, well, then the production server will be down again in the middle of the night sometime! And then you can get some sleep!