存储系统原理介绍（Introduction to storage system principle）.doc

存储系统原理介绍(Introduction to storage system principle)
-----------------RAID introduces concepts
--------------RAID 0
The full name of RAID0 is called Striped Disk Array without Fault Tolerance
Band disk array with fault tolerant design. A cylinder is a disk in the diagram
They are all in parallel. As you can see from the diagram, RAID 0 is stored
The data is divided into the same size by the RAID controller (hardware or software)
Write the disk in the array at the same time. If you play the imagination, you will feel the data
Like a tape across all array disks, the tape depth on each disk is
Same。 The depth of each strip depends on the type of RAID used
In the soft RAID 0 level of NT system, each band depth is only 64KB
Item, and in hard RAID 0 level, can provide 8, 16, 32, 64, as well
128KB and other depth parameters. Striped is a typical way of RAID,
In many RAID terms, Striped is directed to RAID 0. Read in
When the sequence is read from the array disk and bined by the RAID controller.
--------------RAID 1
Disk mirroring
The mirror image - the contents of the two hard disks are exactly the same
Backup each other from content. For example, there are two hard disks in the array
When writing, the RAID controller does not divide the data into strips
It writes data to two hard disks at once. In this way, the
What a hard disk data problem, you can immediately recover from another hard disk.
These two hard disks are not master-slave, that is to say, mirror / restore each other.
--------------RAID 2
At present, it is no longer used, instead of more advanced real-time testing RAID instead of, such as RAID 3, 5, etc..
It uses the early error detection and correction technology --- Hamming code
(Hamming Code) verification technique.
Hamming code is a kind of error code inserted in the original data to make mistakes
Coding techniques for error checking and correction. Encoded with typical 4 bit data
For example, Hamming code will add 3 check codes,