RAID Devices
RAID technology was developed in 1987 at the University of California at Berkeley.
The original intention of the RAID development team was to reduce the cost of mass
storage by combining small inexpensive disks to replace larger expensive disks.
The Berkeley engineers also wanted to provide a level of protection by including
redundant information to ensure that a disk failure would not cause the loss of
access to data. The idea behind raid is to spread the data across a number of
independent disk drives and I/O buses, in a process called ‘striping’, in such
a way that consecutive accesses to the normally contiguous data are not slowed
down by the speed of a single device. Also, enough redundant information is stored
across the disks so that in the event of one disk failing, the data that is on
it can be retrieved from the remaining disks instead.
Protection against disk failure has been achieved through RAID technology. Unfortunately,
RAID arrays no longer reduce the cost of data storage. In fact, the cost of a RAID
array is usually higher than standard disk drives because of their increased complexity,
redundancy and the decreasing cost of large capacity disk drives. However, RAID
technology does provide better performance, data integrity and data availability
than standard disk storage. While I/O has always lagged behind CPU performance,
the disparity is greater today, as developments in technology have enabled CPU
speed to be increased faster than that of I/O systems. The appropriate RAID solution
can significantly help to close that gap.