A redundant array of inexpensive disks is a data storage system that is typically used with computers, especially computer systems storing data that is important or accessed often. This type of system can be utilized in a number of different ways to ensure greater availability of data through redundancy or increased performance through distribution of data. Each of these two primary goals can be achieved in their entirety by sacrificing the other goal or in part as a combination of both goals. A redundant array of inexpensive disks is a common method of creating a data storage system and has come to be industry standard in some professional sectors.
Though often referred to simply as a “RAID,” the term “redundant array of inexpensive disks” has generally been replaced with the term “redundant array of independent disks,” but both names refer to the same type of system. The modern usage of “independent” rather than “inexpensive” has been utilized merely to dissuade potential clients or users from believing this type of system will be inexpensive. While the disks themselves may be fairly inexpensive, since there is an entire array of these disks, with at least two and potentially dozens or hundreds, a redundant array of inexpensive disks can still be quite costly.
The basic idea behind a redundant array of inexpensive disks is that multiple disks, each capable of storing data, are used together in a single system. Computers or terminals connected to this system see the entire array as a single storage device when writing to or reading from the array. By using multiple disks within a single array there is the potential for greater availability, being able to access data, and greater performance, the speed at which data is accessed.
Greater availability is created through the use of redundancy of information in a redundant array of inexpensive disks. With only a single disk, if data on that disk becomes corrupted then it is lost without other forms of backup. A redundant array of inexpensive disks, however, can be used to store the same data on multiple disks, and if one disk is corrupted the data can still be retrieved from another disk. This type of system is often referred to as a RAID-1 system.
A redundant array of inexpensive disks can also be used to increase performance. If data is stored in separate pieces across multiple disks, not as copies but as separate parts, then the data can be retrieved more quickly as each disk is accessed for one piece, rather than waiting for each piece of data to come from different parts of a single disk. This is called “data striping” and increases performance of a system, but not availability, and is typically referred to as a RAID-0 system. Other systems use a combination of both RAID types to improve performance, while still creating redundancy to ensure availability.