What is RAID?

The acronym RAID stands for Redundant Array of Independent Disks. It is a method for storing data in multiple hard drives. In this way, if one hard drive fails the data is restored from the "redundant" storage on another disk. This feature is the "fault-tolerance" of the RAID. A RAID array commonly uses a system of "striping" to divide a large hard drive into smaller, easier to access sections.

There are at least nine types of RAID. The RAID types used by the NAS servers are:

JABOD
Just A Bunch Of Drives. This configuration isn't actually RAID. As its name indicates, JABOD is nothing but sequential hard drives. Data is not spread between drives.

RAID0
This configuration actually provides no redundancy. Information is divided up and spread over one or more drives. This makes data quickly accessible, but not fault-tolerant. RAID0 is possible with just a single hard drive. This is also known as striping.

RAID1
This configuration uses two drives to provide fault-tolerance. The data on one disk is duplicated on to the other. This is also called "disk mirroring". Information is quickly accessible (just at it would be for a single drive). For multiple users, RAID1 offers the best combination of fault-tolerance and performance. This is also known as mirroring.

RAID5
This configuration uses at least three drives to provide fault-tolerance. The data is spread out on two of the drives, and a parity check is stored on the third. Data is not duplicated, as in RAID1, but is reconstructed using the parity information. Because of the drive configuration, RAID5 is slower than RAID0 or RAID1 when writing data to the array. This is also known as checksumming.

A RAID appears as a single volume. The user is unaware that the information is being redundantly stored.

The FileZerver NAS server supports JABOD, RAID0, RAID1 and RAID5.

The CDH NAS server supports RAID0, RAID1, and RAID5.