 |
||
 |
 |
|
 |
||||
 |
||||
|
|
|
|
|
|
The term RAID (Redundant Array of Inexpensive Disks) was coined in a 1988 paper by Patterson, Gibson and Katz of the University of California at Berkeley. In that article, the authors proposed that large arrays of small, inexpensive disks could be used to replace the large, expensive disks used on mainframes and minicomputers. Their analysis showed that the cost per megabyte could be substantially reduced, while both performance and fault tolerance could be increased. In the years since the publication of the Patterson article, the arguments in favor of disk arrays have gotten much stronger. The capacity and performance of small drives have increased, while their cost and physical size have decreased. Over the same period, of course, mainframes and minicomputers have been largely relegated to the status of living dinosaurs or, more politely, "legacy systems". Most disk arrays are now targeted for client/server systems, network servers, and high performance workstations. In these microprocessor-based systems, the performance of the CPU has increased so rapidly that the disk I/O bottleneck is now clearly recognized as the obvious location for innovative and cost-effective performance improvement. While CPU performance has increased by a factor of about 20 over the past nine years (from 16 MHz. 386's to 200+ MHz Pentium Pro's), disk performance has only increased by about a factor of 5. Even in 1988, disk performance was widely recognized as the bottleneck in computer systems. Today, that bottleneck is many times worse. The solution to the performance bottleneck is RAID disk arrays. Let's look at the origins of RAID technology, and the different types of RAID solutions that have been proposed. |
||||
 |
||||