See WikiPedia:Redundant_array_of_independent_disks for more information.
This is a system which changes the file structure of multiple hard drives at the same time, in order to create redundancy (except for certain RAID types, which aren't true RAID. Such as RAID-0). The redundancy is used to rebuild the file structure in case of a disk failure. My personal implementation of RAID-1 is IDE, which is the economic way of implementing a RAID structure in a personal PC.
RAID type 1 is a mirroring system. This is the simplest way to implement RAID, and the quickest to restore. Unfortunately, RAID-1 is the slowest at writing rates. It is quick at reading, though.
RAID type 0 spans data across multiple disks. It is not true redundancy, and requires being mixed with another level in order to obtain fail-safe abilities. This is used mainly for saving space, and increasing disk-access speeds.
This is the level running on /mega. This level incorporates both of the above levels in order to achieve high redunancy and reliability, while increasing the speeds of reading and writing by a large multiplier.