Redundant Array of Inexpensive / Independent Disks (RAID) is a method of storing multiple disk drives which appear like a single disk drive to the software. Data is distributed over the disk drives for parallel operation. It is done to improve the performance of the disk drives. RAID can be organized in 5 different levels as given below:
- RAID level 0: In this organization, consecutive strips are written in drives in round robin algorithm. The method of distribution of data in multiple drives is known as striping. This is done for performing operations in parallel. Level 0 is generally used for large requests. In this organization, performance is increased but it is not very reliable in case of disk failures.
- RAID level 1: In this organization disks are duplicated. Since data is written twice the write performance is decreased. However, the fault tolerance is increased. If a disk failure occurs then backup disk can be used for restoring.
- RAID level 2: Here striping of data is done in byte level by using Hamming codes and parity bits. Performance is greatly increased in this level. The disadvantage of this method is that all the drives should be in rotational synchronization.
- RAID level 3: It is a simplified version of Level 2 where only one parity bit is calculated for each data word. The parity bits are then stored in a separate parity drive. In this level, drives should be in rotational synchronization.
- RAID level 4: Here, striping is done in block level and parity bits are then stored in a separate parity drive. Synchronized drives are not required in this method.
- RAID level 5: Here, striping is done in block level and parity bits are distributed uniformly over all the drives in round robin method. Synchronized drives are not required in this method.