RAID (Redundant array of independent disks, formerly redundant array of inexpensive disks) is in short a failover system for hard drives. A certain number of hard drives (2–4) work together as one logical drive. If anyone of those disks fails at any time none of the data is lost, you simply have to replace the failed drive and the system will automatically rebuild the array.
RAID comes in many flavours.
RAID 0 (also known as a stripe set or striped volume). This is essentially a JBOB (Just a bunch of disks) configuration and provides no data redundancy and is not recommended for maintaining data integrity. Essentially you add up the space of two hard drives (or more) and it will show up as a single big drive.
RAID 1 creates an exact copy or mirror of a set of data. This is great for ensuring that data integrity is maintained because if any disk fails the other one immediately takes over and is completely transparent to the user.
Raid 5 uses stripping to both provide performance and redundancy. By using a min of 3 disks you can achieve super fast speeds and complete data redundancy, as long as only one drive dies. If multiple hard drives fail you will lose all your data. You generally get 75% of your total drive capacity in order to maintain speed and redundancy.
