Memory for disk arrays
Features building. Drives 1 to RAID disk arrays The issue of increasing
productivity disk subsystem is complex. Growth calculation-oxidant capacity of
modern processors led to the observed apparent imbalance between the capacity of
hard disks and processor needs. However, if not enough possibilities single disk,
it may be partly solved this problem will allow the existence of several disks?
Of course, the existence of two or more hard disks on the computer or need to
make these discs work on servers decision does not change together (concurrently)
with each other in a manner that enhanced the performance disk subsystem
operations zapi-si/chteniya . In addition, there should not be using multiple
hard drives, thus improving not only the performance but also the reliability of
data storage that failure of a distribution whistleblower does not result in a
loss of information? This approach was proposed as early as 1987, the American
researchers-Pat tersonom, and Kattsom Gibson of the University of California-Berkeley.
In his article "A Case for Redundant Arrays of Inexpensive Discs, RAID" ( "Overweight
array of inexpensive disks"), they describe how to combine multiple hard drives
into a single logical device so that would-as a result of increased capacity and
system performance, and the refusal of the individual discs do not give him to
the failure of the entire system. Since the article has been 15 years, but the
technology build-RAID arrays not lost relevance of decoding today. The only
thing that has changed since then, it acronym RAID. The fact is that originally
RAID arrays are not built on cheap disk, so call Inexpensive (inexpensive) to
replace the Independent (independent), which is more consistent with reality.
Indeed, it is now RAID technology has been widely disseminated
Next part
|