Redundant array of independent/inexpensive disks; a technology which utilizes multiple disks to provide increased reliability, performance, and/or redundancy. Can be implemented either in software or with a hardware controller.

RAID stands for Redundant Array of Independent/Inexpensive Disks, and is a technology which can provide increased reliability, performance, and/or redundancy by utilizing multiple hard drives simultaneously.

RAID can be implemented in a computer via either a hardware or software controller – the controller takes care of distributing the write and read commands to the drives in use.

There are several types of RAID configurations (also called "RAID levels"), depending on the configuration of hard disks. The most well known are:

  • RAID-0 ("stripeset") distributes the data over all available disks without redundancy. Improves performance by writing to multiple disks in parallel
  • RAID-1 ("mirror") duplicates the data by mirroring every drive, thus creating a 1:1 copy. Slight performance drop in writes, but parallel reads from the multiple copies can improve read performance.
  • RAID-3, RAID-4 are no longer relevant
  • RAID-5 stripes the data and spreads redundancy information on all drives. Can tolerate failure of 1 device.
  • RAID-6 stripes the data and spreads two copies of redundant information across the drives. Can tolerate a 2 device failure.

There are also nested RAID levels which combine the above RAID levels with several others.

Related tags

Tags for different RAID levels:

history | excerpt history