Initially, as envisaged in 1987 by Patterson, Gibson and Katz from the College of California in Berkeley, the acronym RAID stood for a “Redundant Array of Cheap Disks”. In brief a bigger variety of smaller cheaper disks might be used instead of a single way more costly giant exhausting disk, and even to create a disk that was bigger than any presently accessible.
They went a stage additional and postulated a wide range of choices that may not solely lead to getting an enormous disk for a decrease price, however might enhance efficiency, or enhance reliability on the identical time. Partly the choices for improved reliability have been required as utilizing a number of disks gave a discount within the Imply-Time-Between-Failure, divide the MTBF for a drive within the array by the variety of drives and theoretically a RAID will fail extra shortly than a single disk odzyskiwanie danych poznań.
Immediately RAID is often described as a “Redundant Array of Unbiased Disks”, expertise has moved on and even the costliest disks usually are not significantly costly.
Six ranges of RAID have been initially outlined, some geared in the direction of efficiency, others to improved fault tolerance, although the primary of those didn’t have any redundancy or fault-tolerance so may not really be thought of RAID.
RAID zero – Striped and not likely “RAID”
RAID zero supplies capability and velocity however not redundancy, information is striped throughout the drives with the entire advantages that offers, but when one drive fails the RAID is useless simply as if a single exhausting disk drive fails.
That is good for transient storage the place efficiency issues however the information is both non-critical or a duplicate can be saved elsewhere. Different RAID ranges are extra suited to important techniques the place backups may not be up-to-the-minute, or down-time is undesirable.
RAID 1 – Mirroring
RAID 1 is commonly used for the boot gadgets in servers or for important information the place reliability necessities are paramount. Often 2 exhausting disk drives are used and any information written to at least one disk can be written to the opposite.
Within the occasion of a failure of 1 drive the system can change to single drive operation, the failed drive changed and the info transferred to a alternative drive to rebuild the mirror.
RAID 2 launched error correction code era to compensate for drives that didn’t have their very own error detection. There are not any such drives now, and haven’t been for a very long time. RAID 2 is just not actually used wherever.
RAID three – Devoted Parity
RAID three makes use of striping, right down to the byte degree. This provides a overhead for no obvious profit. It additionally introduces “parity” or error correction information on a separate drive so a further exhausting disk is required that offers higher safety however no extra area.
RAID four – Devoted Parity
RAID four stripes to the block degree, and like RAID three shops parity info on a devoted drive.