Detailed deployment of RAID disk arrays and soft RAID implementations

In the current era of large data explosion, more and more enterprises pay more attention to the data security of enterprises, the integrity of data, then how to protect the enterprise data bank security, lossless? Then we use RAID technology, which is now a very hot virtualization, cloud computing, the bottom of the raid storage is used to complete. So what is RAID? We will learn through soft raid cases and theories.

First of all, what is RAID?

Raid simplicity allows a fault-tolerant disk array to consolidate multiple disks into a larger hard disk. The consolidated disk then has a large storage capacity and a hard disk fault-tolerant function. RAID is also divided into soft raid and hard raid;

Partitioning of RAID ratings?

RAID0

The best effect, but the lack of fault tolerance, this mode of raid in reading the data, the first will be written to the first disk and then write to the second disk, so that the data will be written to the two hard drives. If RAID0 has two pieces of data, when the data is written to disk A and then writes to disk B, the pressure on each disk is reduced, and the more disk performance is better for RAID0, the less complete the data will be when a hard drive fails. Therefore, the risk of this approach is high, there is no security to speak of.

RAID1

It can be so understood that he is to make up for the shortcomings of RAID0, the model of the hard disk requires two identical hard drive to make up, if two hard disk is not the same size, he will be small hard drive to the main. The biggest benefit of this raid is the security of the data, but the space usage of the hard disk can only use 50%, because one disk writes the data, one disk data does the backup data simultaneously, so the data of two disks is identical.

The advantage of RAID1 is that it can guarantee the security of data, the disadvantage is that the disk usage is lower and the writing speed is slow.

The best use in the enterprise is

Raid0+1 or raid1+0.

RAID0 data is not secure and does not guarantee the integrity of the data, but RAID1 can guarantee the security of the data, raid0+1 can increase the security of RAID usage.

RAID0+RAID1 Mode of work:

First the disk into RAID0 so that two groups, each group of two hard drives, and then the two groups of RAID0 composition raid1.raid1+0 and raid1+0 the same way. When we need to write the data, we start with the RAID1, and the RAID1 is composed of raid0, so we write the data on average, so we don't care which group of RAID0 disk is corrupted, as long as we have a group of RAID0 is not afraid of data security. So this is the safest.

RAID5

RAID5 requires at least three more disks to be able to be made into a disk array. The way this disk works is that every disk in the writing process, there will always be a disk in the check code, so as long as not two disks at the same time, the entire data is not a problem. At the same time, two pieces of the word that is the character broke out. In the enterprise combat generally will raid5 in the addition of a backup disk, when one of the disk is broken, another disk will automatically switch past. Other raid in the enterprise is not used a lot, we do not introduce.

RAID arrays are classified as soft raid and hardware RAID, hardware RAID is relatively expensive, hard raid is used in large data, and soft raid is rarely used, and below we use a RAID from Linux to simulate raid, which is basically the same as hard raid.

First, deploy disk array raid

1, first on the server to add a hard disk, we simulate in this case, such as the figure is we add the hard drive SDB.