Admin's RAID Data Recovery Guide

HomeBasicsStorage TechniquesRaid LevelsRaid Data RecoveryArticles
Printable Version Site Map

RAID Levels and RAID Data Recovery

Berkeley researchers defined five types of RAID: RAID 1, 2, 3, 4, and 5. Since then however, many more levels have surfaced. Companies have come up with their own proprietary RAID levels; new breeds of RAID have been created by combining RAID levels, and mutations of existing RAID levels have engendered aberrant stepchildren.

Unfortunately, we won't be discussing any of these more esoteric RAID solutions. Instead, we'll be focusing on the five original RAID levels, and one RAID level that isn't really a level at all.


RAID 0 is considered by many purists not to be a true RAID level because it lacks the all important "R." RAID 0 provides no redundancy, and as such, should never be used for applications where data is critical. If a single hard drive fails in this configuration, RAID recovery may be necessary, because the loss of even one drive will result in all data in the array being lost.

Because it only involves striping, RAID 0 is one of the simplest levels of RAID to implement. It requires at least 2 hard drives, but as long as both drives are identical, no storage space is wasted. RAID 0 delivers the best performance and data storage efficiency of any RAID level.

Figure 0. In RAID 0, data is is broken down into stripes which are written across all the drives in the array.

RAID 0: Striped Set


RAID 1 employs the mirroring technique. As a result, it uses storage space very inefficiently. Fifty percent of your disk space will always be wasted in a RAID 1 configuration. However, it does offer the advantage of 100% redundancy. If one disk fails, there's no need to call a RAID recovery company to recover your data, simply rebuild your lost data from the mirror.

RAID 1 requires at least 2 hard drives, and additional hard drives must always be added in pairs. It is ideal for applications where data is critical.

Figure 1. In RAID 1, data from one hard drive is mirrored onto a second hard drive, so that there are two identical copies of the data.

RAID 1: Mirrored Set


RAID 2 is the black sheep of the RAID family in that it doesn't use one or more of the standard striping, mirroring, or parity techniques. It does however, use something similar to striping with parity, which we'll read when we cover RAID level 3.

Because of its high cost and complexity, RAID 2 never really caught on. In fact, it isn't even used commercially today. RAID 2 uses byte level striping with a form of error correcting code (ECC) known as Hamming code. The number of hard drives required for a RAID 2 configuration may vary, but a typical setup may use as many as 14 disk drives: 10 data disks and 4 ECC disks.

Figure 2. In RAID 2, data is split at the bit level over a number of data and ECC disks. Every time data is written to the array, the Hamming codes are calculated and written to the ECC disks. When the data is read from the array, these ECC codes are read as well to confirm that no errors have occurred since the data was written. If a single-bit error occurs, it can be corrected immediately.


Next Previous  |  Next: More RAID Levels