< : What is RAID:
disk array (redundant Arrays of Independent Disks raid " The redundant array of independent disks "
The disk array is made up of many inexpensive
The disk array can also take advantageof the concept of Parity check Array any one of the Hard Drive Failure , the data can still be read, and when the data is refactored, the data is computed and re-placed into the new hard drive.
< Two;: Implementation method: Hardware Implementation, Software Implementation
In the actual production environment, almost the use of soft raid, almost all hardware-implemented raid.
Common type: Raid-0,raid-1,raid-5,raid-10,raid-50,jbod
< three: Types of RAID:
1>. RAID0:
Working Style : Divides the source data into pieces, and these pieces of data are stored sequentially on each disk, so the read and write performance of the entire disk is definitely improved.
Features: 1>. Read and write performance improvement 2> free space N*min (s1,s2,...) 3>: no fault tolerance 4> minimum number of disks: 2
Application Scenarios: store some unimportant data
2>. RAID1:
How you work: Store your source data on disk in a mirrored way. The same data is available on multiple disks. Implement fault tolerance.
Features: 1> Read and write performance improvement, write performance slightly decreased 2> free space: 1*min (s1,s1,..) 3> have redundancy ability 4> minimum number of disks 2
Application Scenario: Store data that is important and speed requirements for data read and write
3>. RAID4:
Working mode: at least three Disk , which has a disk to do parity check disk, the remaining the data in the disk is made XOR and the parity check is stored in the parity disk,
Features: 1> the ability to read has improved 2> free space: (N-1) *min (s1,s2,..) 3> Regardless of which disk the data is written to, the disk that holds the parity code is written, so it becomes a bottleneck for the entire RAID-4 mechanism.
Application scenario: Almost no, if you want to make RAID4, it's better to RAID5
4>. RAID5:
Working mode: RAID5 compared to the RAID4 principle is similar, but RAID5 compared to the RAID4, parity check bit is not stored on a disk, but distributed on each disk, in fact, compared to RAID4, reducing the pressure of that piece of parity check disk.
features: 1>. Read and write performance improvement 2> free Space: (N-1) *min (s1,s2,...) 3>. fault tolerance, allowing damage to 1 disk 4>. minimum number of disks: 3.
Application scenario: RAID5 is a common raid mechanism.
6>. RAID6
RAID6 has two parity discs compared to RAID5, with higher security levels
specific: 1>. read-write performance improvement 2> RAID6 has two parity discs compared to RAID5, so it requires at least 4 disks 3> free space: (N-2) *min (s1,s2,....) 4>. have fault tolerance, allow bad 2 disk 5>. minimum number of disks: 4
7>. RAID7
RAID 7 full Name " Optimized asynchrony for hi i/O Rates as well as high data Transfer Rates (optimized asynchronous hi I/O rate and high data transfer rate) " , which is clearly different from the raid level we've seen before. RAID 7 can be fully understood as an isolated storage computer with its own operating system and management tools, which can be run independently.
8>. Jbod:just a Bunch of Disks
Function: Combines the space of multiple disks into one large contiguous space for use
Free space: sum (s1,s2,...)
9> mixed type
1>>. RAID10:
Implementation method: First the disk 22 into RAID1, RAID1 the upper layer and then made RAID0.
features: 1>. The bottom is RAID1, the upper layer is RAID0 2> read-write performance improvement 3> free space: n*min (S1,s2,...) /2
650) this.width=650; "src=" Http://s3.51cto.com/wyfs02/M02/85/47/wKioL1efBx6SN3vwAAAKFBHzQqM972.png "title=" Picture 1.png "alt=" Wkiol1efbx6sn3vwaaakfbhzqqm972.png "/>
2>>. RAID01:
650) this.width=650; "src=" Http://s3.51cto.com/wyfs02/M01/85/48/wKioL1efCM_QgirzAAARXSW3zzo632.png "title=" capture. PNG "alt=" Wkiol1efcm_qgirzaaarxsw3zzo632.png "/>
Implementation: The number of disks is divided into 2 parts, some of which are first made RAID0, and then the two parts are made into RAID1.
Features: The bottom is RAID0, the upper layer is RAID1.
From the above figure, the comparison of RAID10 and RAID01:
1>: When a disk in different parts of the RAID01 is broken, the whole RAID01 may not work.
2>: When the RAID10 two parts have disk damage, then the entire RAID10 may not be used.
3>. RAID10 to be better than RAID01
3>>. RAID-50: The bottom is RAID5, the upper layer is RAID0, so the data is more secure.
A little memory of RAID types under Linux