Speaking from an IT management perspective, there's no such thing as an overkill recovery plan... However, in the scenario you're looking at, mirroring the servers themselves would be the best bet, preferably with the servers in two separate locations. I'm also an advocate of splitting services among as many machines as I can budget for, to near-eliminate the possibility of never having total downtime. Chris -----Original Message----- From: Matt Johnson To: SuSe Sent: 7/10/03 4:01 PM Subject: [suse-linux-uk-schools] bombproof or overkill recovery plan If I was starting from scratch... As a method of complete disaster recovery, would it be overkill to buy two identical (cheapish) servers. Server 1 would have two identical 40gb ide disks. I plan to 'dd' from one disk to another each night. Then I'd have a complete image nightly of our server, right? If things went horribly wrong, I could replace server 2's hard drive with either of the two in server 1 and it would work as the hardware in both servers is the same? Is two servers, and two hard drives in server 1, overkill? Or in a school with 600 pupils and staff relying on these systems, is this redundancy sensible, if not necessary? (I'd probably use server 2 to do a copy of any key files and home directories from server 1 as well - they'd be in different areas of the building. This would be for a teacher accidently deleting all their work, or the like). Silly question... Can you 'dd' from the drive your using at the time? ie. If I've booted from, and am using hda, can I use it as my 'dd' input? Thanks for any pointers. -- Matt ________________________________________________________________________ Want to chat instantly with your online friends? Get the FREE Yahoo! Messenger http://uk.messenger.yahoo.com/ -- To unsubscribe, e-mail: suse-linux-uk-schools-unsubscribe@suse.com For additional commands, e-mail: suse-linux-uk-schools-help@suse.com