RAID
Redundant Array of Inexpensive Disks, is the organization of multiple disks into a large, high performance logical disk.Disk arrays stripe data across multiple disks and access them
in parallel to achieve:
- Higher data transfer rates on large data accesses and
- Higher I/O rates on small data accesses.
There are a number of different RAID approaches; these are the most popular methods.
Raid 0
RAID 0. Disk stripping offers superior speed because it simultaneously
access all drives through concurrent parallel read and writes; however,
it does not offer redundancy. This RAID level is suitable, for example,
for temporary synchronised video files.
Raid 1
RAID 1. Disk mirroring. Two identical disks operate side-by-side
in parallel, providing a complete, continuously updated backup of all
data. Should one of the drives fail, the surviving mirrored drive
provides read and writes. A replacement module (with new drive or power
supply) can be hot-swapped. Data is backed-up to the new drive and full
redundancy resumes. Disk mirroring offers the highest level of "mission
critical" redundancy. This RAID level is suited for moderate-size file
servers.
Raid 3
RAID 3. Parallel Data Access offers the potential for extremely high data exchanges rates because rates because every disk in the array is used for each read and write. Only one I/O request can be active at a time but RAID 3 achieves exceptional speed because, rather than writing all the data to one disk, the individual record’s data is striped across multiple disks. For example, if you have a five-disk array, the data stream is logically broken up and continuously written or read across the four drives while the fifth drive provides redundancy by tracking parity. At each write, the parity is recalculated and rewritten on the fifth drive. The parity drive provides the missing numbers if one of the data drives fails; or, if the parity drive fails, the data drives can provide the data directly.
When a new module containing a configured drive or power supply
is hot-swapped in, the array automatically rebuilds its redundancy.
This RAID level might be used for larger file servers or applications
requiring large sequential transfers.
Raid 5
RAID 5. Independent Data Access is designed for high transaction
throughput and concurrent access of data. RAID 3, with parallel data
access, kept the data on four disks with the parity on the fifth. RAID
5 puts the entire record on one of the five disks and the related
parity information on one of the four remaining disks. If any of the
five disks fails, the parity information from the other four disks
enables the system to mathematically rebuild the missing data. The data
on the undamaged disk is simply read and written directly. This RAID
level is ideal for database severs where there are greater amounts of
independent access.
