|
Not
a bug spray? Not a sudden swift attack? RAID originally stood
for Redundant Array of Inexpensive Disks. Current usage often
reads Redundant Array of Independent Disks (or Drives). Now, both
the "I words"–independent, and, especially with IDE drives,
inexpensive–are appropriate. A RAID controller manages
an array of physical hard drives into a simple scheme of logical
drives which the system sees.
Basics
RAID
systems address one or both of two basic data storage problems:
data safety, and read/write performance (speed).
First,
safety. Data stored on magnetic disks is, as most of us know only
too well, subject to loss. As some say, it is not a question of
if your hard drive will fail, it is only a question
of when. And though IDE drives have improved astonishingly
in physical size (smaller is better), in capacity (larger is better),
and in reliability (longer MTBF is, of course, better) since their
introduction, it is still a question of when. At
the simplest level, you can have two IDE drives in a RAID array,
with exactly the same data on each. If either drive fails, the
other takes over. Of
course, in the event of a disk failure, in addition to managing
the reconstruction of data, the RAID controller sends an alert,
so that steps may be taken to replace the failed hardware.
Second,
speed. If you have a ton of data coming in (or going out) every
second, such as with streaming uncompressed, full-screen, hi-res
video, with stereo sound, it is still a challenge for a hard drive
to keep up. Dropouts, or worse, do happen. But, again at the simplest
level, if you have two hard drives sharing the load (you could
think of it as splitting the stream), they can keep up.
It's another case where two heads are better than one, or (with
four read/write heads in a typical IDE drive) eight are better
than four.
|
