A disk array consists of multiple disk
drives under the command of an array controller. The disk array
incorporates features that differentiate it from traditional disk
storage devices.
Most types of disk arrays provide for one of two possible
options for protecting data in the event of a disk failure. This
becomes more and more important as the number of disks on a system
increases, since the chance of a disk failure also increases. Normally,
a disk crash brings the system down or prevents access to data,
removing it from service until the problem is located and repaired,
and the data is reloaded.
The first kind of data protection is called data
encoding. When a disk drive fails, the array controller
generates encoded data, which is similar to parity or checksum calculations.
This allows missing user data to be reconstructed using a mathematical
formula to rebuild lost data. As a result, the data remains accessible
and the system remains up and running without suffering any downtime.
The second method of data protection utilizes hardware
mirroring as a means of providing high data availability
by duplicating data on redundant disk drives. As a result, failure
in one disk still allows access to the data on an alternate disk.