Memory for disk arrays
To increase the speed of parallel processing is applied. If placed a block of N
array disks certain size and organize the strike, so that there was the
possibility of simultaneous readers, this unit can be considered in N times
faster (without taking into account the emergence of the block). Because all
figures are transmitted in parallel, this architecture is called parallel-access
array (with a parallel array access). Arrays with parallel access typically used
for applications that require data transfer to the large size. Some tasks, in
contrast, is characterized by a large number of small bugs. These tasks include
Cham, for example, the task of handling databases. With record database on the
disk mass-sion, it is possible to distribute the download, regardless
positioning drives. This architecture made vat called independent-access array (with
an array of independent access). In addition to the type of access, different
levels of RAID way location, and the formation of information-tion. The
superfluous information can be placed on a specially-allocated by disk or
stirred among all drives. Methods for the formation of this information a few
more. The simplest of them - this is a complete overlap (100 - Interest
redundancy), or mirror-istrative. Moreover, the codes are used to correction of
errors, as well as the calculation of parity. disk array without fault tolerance
(Striped Disk Array without Fault Tol-erance) 3 RAID-0 RAID Level 0, strictly
speaking, not a redundant array and consequently does not provide the
reliability of data storage. Nonetheless, this level is widely used in cases
where the need for high performance storage subsystem. Oso and particularly
popular in this level workstations.
Next part
|