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 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
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
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.
Previous | Next:
More RAID Levels