Stewart Stremler wrote:
A double-disk failure in a mirror is just as bad.
Not necessarily. RAID 1+0 (mirror first, then stripe across the mirrors) can handle quite a few disk failures.
I installed an 8 drive RAID 1+0 for a client once. Two drives *could* take it out, but it *could* take as many as 5 drives. When I went back 2 years later, they had lost 3 drives and were still running quite blissfully in spite of the 3 bright blinking red lights.
Which was suprising. Drives are supposed to automatically handle certain sorts of failures (bad sectors), up to a point, which would seem to create a bathtub curve.
Well, a bathtub curve just implies an age related breakdown. Disks have *so* much error correction that it just gradually degrades. Also, don't some of the Linux filesystems map out bad sectors, too?
The whole reason that drive manufacturers went from 512 byte sectors to 4096 byte sectors was so that they could run the ECC over a longer number of bits.
-a -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
