Disk mirroring simply createsa mirror image of whatever partitions you want backed up onto a second drive.As has been said, you lose some capacity in trade-off for the mirror, but as far as performance issues go, the only thing I can think of is you might take a slight hit on the write speed to your drives as a mirror requires the same data to be written twice. On the other side, you might get a small boost to read speed as both drives can be read from simultaneously. It will also allow you to mirror system or boot partitions so as to protect your system from a disk failure. It mirrors partitions and not physical disks though, so if you have multiple partitions you want to mirror you'll have to choose to mirror all of the partitions you want backed up, It won;t just automatically set up the second disk as identical to the first. This will let you play with the capacity loss ass well, for example if you have a 2 120 gig drives and only have say 40 gig you want to mirror, you can just create a 40 gig partition, save all you want to back up onto that partition and mirror it. It should leave you with 80 gig left on both drive to use as you see fit, but just remember only what you actually mirror will be backed up.
It's a good system to run if you have a lot of data that you can't afford to lose and enough HD space to spare so as not to miss the loss of capacity. If you lose one disk, it's just a matter of breaking the existing mirror and creating a new one again.
Disk Striping (Raid 5) requires at least 3 drives to implement. It uses some type of math to keep track of your data across the drives. It starts writing dat on the first disk, jumps to the second, third, then back to the first disk and so on so you basically have all your data spread out over all your disks. It also can be a lot more expensice to set up since you need more drives.
In essance it something like if you have the formula 5+5+5=15, where each number in a disk drive, and you lose one, it can calculate the missing portion. For example, 5+(?)+5=15, it would recognize the missing data as 5. The problem with that is that if something happens to data on more than one disk, it can't make the computations and you still lose everything.
Performance wise, it takes a much bigger hit on write speed since the data is being written across multiple disks as well as the calculations required to ensure recovery. The upside is that you get better read speeds since the data is spread over multiple disks and can be read allows you better acess time. you still take a loss on storage capacity, but it's not as big as with a mirror set since it's more spread out. A thrre disk set would cost you a third of your available storage capacity, a four disk set would cost you a quarter and so on.
If this seems like it rambles, I haven't slept in over 24 hours, so forgive me. Also, it's been a few years since I worked in a network environment, so all this is based on my knowledge of Windows NT, and at the end we simply used ghost to keep a copy of everything on the server and simply ghosted the image down to whatever computer went down on us. they were all identical anyway seeing as the majority of the systems we dealt with were lab computers at a college, so we only had to keep about 4 ghosted images on hand and it took less than ten minutes to set a system back to it's original state that way.
I'm also not guarranteeing I didn;t miss something or that I know half what I;'m talking about in my current state.Like HLD said, it would be wise to talk to Ned since he's still actually in the field
