RAID Principles and usage

RAID: Redundant array of independent disks.

In the early days it was called the redundant Arrays of inexpensive disks, a "cheap Redundant array of disks"; it turned out that the inexpensive effect of such a device was really high. Therefore, change ' inexpensive ' to ' independent ', what we now call "independent redundant disk array".

Redundant array of independent disks is the same data stored in different parts of the hard disk, and do a backup, not to loss of data, that is, it has redundancy and fault-tolerant features.

Its role in the popular saying is: The combination of multiple devices to complete a task at the same time.

The implementation of RAID technology is tiered. Let's take a look at each level of RAID and how it is implemented:

RAID level: (only the disk is organized differently, no levels are available)

RAID 0:

Stripe technology, the data is stored in parallel on the disk, so there is high data transmission performance. But because it has no redundancy, there is no fault-tolerant ability, so the data can not be repaired once it is damaged. So, RAID0 simply improve performance, data security is very low.

Summary: Performance improvement: Read, write

Redundancy capability (fault tolerance): none

Space utilization: NS

Requires at least 2 plates.

RAID1:

Mirroring technology is to store the data on multiple disks and then generate a backup. When reading the data, it can be read from the original disk or read from the mirrored backup disk, but it cannot be written. When a disk is corrupted, you can also use a mirrored backup to restore the original data back. This gives it redundancy, so the number of disks it needs will be twice times the RAID0 store the same data (because mirroring is required).

Summary: Performance: Write performance degradation, read performance improvement

Redundancy capability (fault tolerance): have

Space utilization: 1/2

Requires at least 2 plates.

RAID4:

Parity code technology, but also the data stripe and distributed on different disks, but he needs to use a disk as a check disk, the data storage disk of the cyclic redundancy check code on this disk. When a disk is damaged, the corrupted data can be restored through a checksum disk and other data disks. Therefore, it also has redundant fault-tolerant capability. However, if two or more than two disks are damaged, the data will not be repaired. Therefore, this technique is less used.

Summary: Performance: reading, writing promotion

Redundancy Ability: have

Space utilization: (n-1)/n

Requires at least 3 plates.