RAID technology introduction
RAID is called Redundant Array Of Disks, which is a "Independent Disk Redundant Array" (initially "cheap disk redundant array") abbreviation. In 1987, Patterson, Gibson and Katz were defined in an article in the University of California University. RAID array technologies allow a range of disk to be packetized to implement data redundancy required for data protection, and data strip distributions formed to improve read and write performance. RAID originally used in the high-end server market, but with the rapid development of computer technology, RAID technology has penetrated into various fields of computer all over. Today, in the household computer motherboard, the RAID control chip is also visible.
Generally, the RAID system can exist in various interface interfaces, as we are currently, PATA, SATA, and SCSI have corresponding hard drives to form RAID. With the release of the Intel 865/875 series chipset, the hard disk interface of the home market begins to switch SATA, while the RAID method will also transition from PATA to SATA.
RAID technology has experienced a series of changes and development experience with people's use. In the household market, we can only see the RAID 0, RAID 1, and RAID 0 1 of these disk array. However, from the birth of the DFI LANPARTY motherboard, we ushered in the fourth disk array, that is, RAID 1.5.
From practical applications, most of the formation of home RAID is to further improve the read and write performance of the disk, and the data backup can be reached in other ways (such as burning). Therefore, in the case of only 2 hard drives, people are willing to try RAID 0, but the birth of RAID 1.5 makes us change this concept. What is the performance of these two relatively inexpensive disk arrays? Let us notice everyone.
RAID 0:
RAID 0 uses a technology called "striping" to distribute the data to each disk. Every "strip" is dispersed to a continuous "block" (block), and the data is divided into several blocks from 512 bytes to Duzab, and then write to the disk. The first block is written to the disk 1, and the second block is written to the disk 2, so that this is pushed. When the system reaches the last disk in the array, it is written to the next segment of the disk 1, so.
Segmentation data can assign an I / O load average into all drives. Since the driver can be written or read simultaneously, performance is significantly improved. However, it has no data protection capabilities. If a disk is malfunction, the data will be lost. Therefore, RAID 0 does not apply to a critical task environment, but it is very suitable for video, image production and editing.
RAID 1:
RAID 1 is also known as a mirror, because data on a disk is completely copied to another disk. If the data of a disk is incorrect, or the hard disk has a bad track, then another hard disk can rescue data loss and system interrupts caused by the disk failure. In addition, RAID 1 can also achieve duplex - you can copy the entire controller so that your data can be protected when a disk failure or controller failure occurs. The disadvantage of mirroring and duplex is to need more than one count of drives to copy data, but the system's read and write performance will not increase, which may be a small expense. RAID L can be implemented by software or hardware.
RAID 2:
RAID 2 is an array of sea-coded check-in-vault for mainframes and supercomputers. The first, second, fourth, 4th, 4th ... of the second N times of the disk drive group is a special check-in, for check and error correction. As shown below: RAID 2, 1, 2, 4 disk drives (red) of seven disk drives are the error correction disc, and the remaining (purple) is used to store data. RAID 2 has a very high performance of the reading and writing of large data, but the performance of a small amount of data is not good, so RAID 2 actually uses less. Due to the particularity of RAID 2, as long as we use the disk drive, the less the percentage of the verification disk is in it. If you want to achieve a relatively ideal speed and better disk utilization, it is best to increase the hard disk that saves the check code ECC code, but this will pay more of the purchase cost of the hard disk to ensure data redundancy. For the design of the controller, it is simpler than the RAID 3, 4 or 5 described below.
RAID 3:
RAID 3, which is a strip with a special parity (Parity). Each strip piece is equivalent to a "block" that is so large to store redundant information, ie phenomenon. The parity bit is encoded information. If the data of a disk is incorrect, or the disk is faulty, it can be used to recover data. In a data-intensive environment or a single user environment, the establishment of RAID 3 is advantageous to access longer, but the performance will decrease when accessing shorter records, like RAID 2.
RAID 4:
RAID 4 is a stand-alone disk structure with parity code. It is very similar to RAID 3, and the difference is that RAID 4 access to data is performed by data block. RAID 3 is a horizontal strip, while RAID 4 is a vertical bar. So RAID 3 often allows all hard drives in the array, while RAID 4 is only required to access the useful hard drive. The speed of reading data is greatly improved, but in terms of write data, the old data to be restored from the data hard drive and verify the hard drive is required, and then write the updated data and inspection blocks. The hard drive, so the processing time is longer than RAID 3.
RAID 5:
RAID 5 is also called a band with distributed parity. Each strip is equivalent to a "block" that is used to store the odd bits. Unlike RAID 3, RAID 5 is also distributed on all disks in all disks, which is not a disk, greatly reduces the burden of parity discs. Although there are some capacity losses, RAID 5 can provide a more perfect overall performance, which is also a disk array scheme that is widely used. It is suitable for input / output intensive, high read / write ratios, such as transaction processing, etc.
In order to have a redundant degree of RAID 5, we need at least three disk arrays consisting of at least three disks. RAID 5 can be implemented by disk array controller hardware or through some network operating system software.
RAID 6:
RAID 6 is a stand-alone disk structure with two parity check code with two distributions. It uses a second parity assigned on different disks to achieve enhanced RAID 5. It can withstand multiple drivers at the same time, but it is more time to calculate the time spent on parity and verification data, which causes the system's load, which greatly reduces the overall disk performance, and the system needs one Extremely complex controller. Of course, since the second parity value is introduced, we need N 2 disks.
RAID 7:
RAID 7 own with intelligent real-time operating system and software tools for storage management, it can be run independently of the host, do not occupy the host CPU resources. RAID 7 Storage Computer Operating System is a real-time event-driven operating system, primarily for system initialization and arrange all data transfer of the RAID 7 disk array and convert them to the corresponding physical storage drive. To set and control the read / write speed via Storage Computer Operating System, you can make the host I / O pass performance optimally. If a disk is faulty, the recovery operation can be automatically executed, and the reconstruction process of the backup disk can be managed. RAID 7 uses non-synchronous access methods, which greatly reduces the data writing bottleneck and increases I / O speed. (The so-called unconductive access, that is, each I / O interface of RAID 7 has a dedicated high-speed channel, as a traffic path for data or control information, so it can independently control the data access of each disk in your own system.) If RAID 7 has n disks, remove a check disk (used as redundant calculations), can handle the read / write instructions randomly emitted at the same time simultaneously, thereby significantly improvement I / O applications. The RAID 7 system has a built-in real-time operating system that automatically optimizes the read / write instructions sent by the host. The data that may be read in advance will be read in advance, which greatly reduces the number of rotations of the magnetic head. Improve the I / O speed. RAID 7 helps users have effectively managed increasingly large data storage systems, and increase the operational efficiency of the system to at least double, meeting the different needs of various users.
RAID 10 (RAID 0 1):
RAID 10, also known as the mirror array strip, now we generally call it RAID 0 1. RAID 10 (RAID 0 1) provides 100% data redundancy, supporting larger volume sizes. To form a RAID 10 (RAID 0 1) requires 4 disks, two of which provide the band data distribution, which provides the read and write performance of RAID 0, and the other two is the image of the front two hard drives, ensuring the complete data. Backup.
RAID 30:
RAID 30 is also known as a dedicated parity array strip. It has the characteristics of RAID 0 and RAID 3, composed of arrays from two sets of RAID 3 disks (3 disks per group), using dedicated parity, and these two disks have a RAID 0 array to implement cross-disk extraction data . RAID 30 provides fault tolerance and supports larger volume sizes. Like RAID 10, RAID 30 also provides high reliability because there are two physical disk drives fail (one in each array), and data is still available.
The RAID 30 minimal requires 6 drives, which is best for non-interactive applications, such as video streams, graphics, and image processing. These applications processes large files and require high availability and high speed.
RAID 50: RAID 50 is called a distributed parity array strip. Immortal with RAID 30, it has the common characteristics of RAID 5 and RAID 0. It consists of two sets of RAID 5 disk (at least 3 each), each group uses a distributed odd-even bit, and the two sets of hard drives are set to RAID 0, and data is extracted across disk. RAID 50 provides reliable data storage and excellent overall performance and supports larger volume sizes. Even if the two physical disks have failed (one in each array), the data can also recover smoothly.
RAID 50 requires a minimum of 6 drives, which is best suited for applications that require high reliability storage, high read speed, high data transmission performance. These applications include transaction and office applications with many users to access small files. RAID 53:
RAID 53 is called an efficient data transfer disk structure. The implementation of the structure with the Level 0 data strip array, where each section is a RAID 3 array. Its redundant and fault-tolerant ability with RAID 3. This is beneficial for systems that require high data transfer rates, but it is expensive and low efficiency.
RAID 1.5:
RAID 1.5 is a new disk array method, which has the characteristics of RAID 0 1, and the difference is that its implementation only requires 2 hard drives. From the surface, the disk after RAID 1.5 is formed, and both have the same data. Of course, RAID 1.5 is also a disk array mode that cannot fully utilize disk space, so two 80GB hard drives are the same as RAID 1.5, that is, only 80GB actual use space, and another 80GB is it. Backup data. If the two hard drives are separated, they run them in the original system, and they are unobstructed. However, by practical applications, we found that if the two hard drives are running separately, the slight changes in their data will cause the reorganized disk array, and complete data recovery cannot be achieved, but is subject to less data.
Since RAID 1.5 and RAID 1 have a very similar effect, how is it experimenting with RAID 0 strip read and write operations? So far, we have not yet true materials prove that the following imaginary: The disk array control chip has advanced control functions, allowing two disks to record the same data in the way, but when reading, control But you can distinguish between the program strips that need to be read, and then read different strips from different hard disks, respectively, to improve performance RAID 0 effects.
