On Tue, Nov 26, 2002 at 08:22:34PM -0500, Joey Parshley wrote: : : I want: : To implement RAID. Which would be better for disaster recovery 0 or 1 (I'm : thinking mirroring)? RAID 0 is pure striping. No data recovery in case of a disaster. RAID 1 is mirroring. If one drive dies, at least you have its mirror as a redundant spare. : I have 4 120 Gb drives. Would it make sense to mirror : the last two and leave the first two as is? It depends on how important your data is. Ask yourself some questions. Given configuration X, if drive X1 dies, what would happen, and what do you need to do to recover from the incident? Just think things through. : Everyone to authenticate via the server. I would like to have their "home : folder" be on the server so they can use any client. [...] : Please excuse the "basicness" of this post. I have several OS X books but : nothing strictly focusing on setting up a client server environment. If you're running a server, and a drive dies, your main concerns are to recover all user data with minimal/no loss and to restore the server as quickly as possible to normal. I guess your idea is to put all the user data onto a RAID volume. RAID 0 gets you twice the disk space, but you are toast if one drive dies (unless you have tape backups). RAID 1 will save you if one drive dies, and it's faster than tape backup, but you effectively lose half of your drives on disk space. Also, don't forget to enable HFS+ journaling! -- Eugene Lee eugene at fsck.net