RAID stands for Redundant Array of Independent Disks - this fun acronym refers to digital storage units which use multiple disk drives to improve performance or for data redundancy. Keep in mind that this means that RAID units are designed to operate faster and ensure that data is available if a drive fails. It does not mean that a single RAID unit operates as a back-up archive. A RAID unit will not account for problems when files are corrupted or deleted. So if a file on one of the RAID disks is deleted, it will be automatically removed from the mirrored disks in the same unit.
There are several different kinds of RAID configurations, and they are each assigned a number.
RAID 0 - employs striping to distribute the contents of a file evenly across the multiple disks in a RAID unit. This lets concurrent read or write operations run more quickly, but if one disk fails, the entire volume is lost.
RAID 1 - does not use striping - so each file is entirely on one disk, but there are redundant copies of the file on the other disks, so if one disk fails, the data can be recalled from one of the others. The problem is that data can only be written at the speed of the slowest drive since all of the drives must be updated.
RAID 2 - in this configuration, the spindles on each drive are synchronized, and each sequential bit of a file is stored on a different drive. This is a largely obsolete configuration, and is not used on any newly manufactured RAID units.
RAID 3 - in this configuration the spindles are also synchronized, but the data for a file is spread across drives at the byte level. RAID 3 is very rarely used.
RAID 4 - this configuration uses block striping - a block usually being 512 KB. [8 bits in a byte].
RAID 5 - The data is distributed so that if one drive fails, the data can be assembled from the others. A minimum of three drives are required. It also uses block level striping.
RAID 6 - like RAID 5 this also uses block striping, but allows for recovery even if two drives fail, and requires a minimum of four drives.