High Performance Website Architecture Design cache Chapter (4)-Master-slave replication

RAID is an abbreviation for redundent array of inexpensive disks, literally "Redundant array of Inexpensive Disks" or "Disk array". Later, the letter I in the raid was changed to Independent,raid as a "stand-alone redundant disk array". The disk array is made up of a number of inexpensive, smaller, more stable, slower disks, combined into a large disk group, to improve the performance of the entire disk system by providing the added effect of data generated by individual disks. At the same time, when storing data, using this technology, the data is cut into many sections, respectively, stored on each hard disk. The disk array can also take advantage of the same-bit checking (Parity check) concept, in the array of any hard disk failure, can still read data, in the data reconstruction, the failed hard drive data, after the calculation of the new hard disk. The disk array enclosure is an external raid that is equipped with many hard drives.

RAID Benefits:

(1) high transmission rate. In partial raid mode, by dividing data into multiple chunks (blocks) to write/read multiple disks in parallel to increase the speed of access to the disk, many disk drives can transmit data simultaneously, and these disk drives are logically a disk drive. So using RAID can reach the rate of a single disk drive several times. Because the CPU is growing fast, and the data transfer rate of disk drives cannot be greatly improved, a solution is needed to resolve the contradiction between the two.
(2) Higher security. Compared to ordinary disk drives, many RAID modes provide a variety of data repair functions, and when a hard disk drive in a raid fails to be used, the data in this drive can be recovered through the other disk drives in the raid, and ordinary disk drives cannot be implemented.

RAID classification:

RAID technology is divided into several different levels, each of which offers different speeds, security and cost-performance. Depending on the actual situation, selecting the appropriate RAID level can meet the user's requirements for storage system availability, performance, and capacity. The commonly used RAID levels are the following: Nraid,jbod,raid0,raid1,raid0+1,raid3,raid5,raid10 and so on. RAID10,RAID5 and RAID (0+1) are often used today.

RAID0 Technology

The RAID0 is the first RAID mode, data stripping, which is a striped technology. RAID 0 is the simplest form of a disk array, requiring more than 2 hard drives and low cost to improve performance and throughput across the entire disk. RAID0 does not provide redundancy or error-repair capability, it is the lowest implementation cost.
The simplest implementation of RAID0 is to create a large volume set in the form of hardware that is used by the same hard disk drive in a smart disk controller or by using a disk driver in the operating system in a software way. In the use of computer data is written to each block of hard disk, its greatest advantage is that it can be multiplied to increase the capacity of the hard disk. If you use a four block 80GB hard disk set to build RAID0 mode, then the disk capacity will be 320GB. In terms of speed, the speed of each individual piece of hard disk is exactly the same. The biggest drawback is that any piece of hard disk failure, the entire system will be destroyed, reliability is only a single piece of hard disk 1/n.

RAID1 Technology-Mirror (mirroring)
RAID1 is called disk mirroring, the principle is to mirror the data of one disk to another disk, that is, the data is written to a piece of disk, the other disk will be idle to generate a mirror file, without affecting the performance of the maximum guarantee system reliability and repair, As long as there is at least one disk in any pair of mirrors in the system can be used, and even in half the number of hard disk problems when the system can operate normally, when a hard disk fails, the system will ignore the hard disk, instead of using the remaining mirror disk to read and write data, with good disk redundancy capability. While this is absolutely safe for the data, the cost is significantly increased, with a disk utilization of 50% and a disk space of only 160GB for four 80GB capacity drives. In addition, the hard disk failure of the RAID system is no longer reliable, should promptly replace the damaged hard disk, or the remaining mirror disk also has problems, then the entire system will crash. The original data will take a long time to synchronize the image after the replacement of the new disk, the external data access will not be affected, but the performance of the entire system has declined. As a result, RAID1 is used in situations where critical important data is preserved.

Matrix RAID Technology

Matrix disk array. is a new Intel-created patented RAID mode for SATA interfaces, characterized by the ability to simultaneously implement RAID0 and RAID1 two modes on 2 disks, which works by dividing some disk space of each disk in 2 disks into RAID0 or 1. The remaining space is composed of RAID1 or 0.

Matrixraid also features a "hot backup" hard drive that supports RAID1 array partitioning. Motherboards that typically support the Matrixraid feature have four SATA interfaces, while a set of matrixraid requires only two hard drives, using two SATA interfaces. The other two idle SATA interfaces can be plugged into the hard drive, starting the "Hot backup" function. When a hard disk in the Matrixraid system fails, the "Hot backup" hard disk will immediately take over its work to ensure the security of the data in the RAID1 array partition. Because the data in the RAID 0 array partition has been destroyed when a hard drive crashes, the "Hot backup" drive is not valid for the RAID0 array.

RAID3 Technology

RAID3 is to divide the data into multiple "blocks", according to a certain fault-tolerant algorithm, stored in the n+1 hard disk, the actual data occupied by the space of n hard disk sum, and the first n+1 hard disk stored on the data is to verify fault-tolerant information, when this n+ When one of the 1 hard drives fails, the data from the other n hard drives can also be restored to the original data. That is, RAID3 uses a single disk to hold parity information. If a disk fails, the parity disk and other data disks can regenerate the data. If the parity disk fails, data usage is not affected. RAID 3 provides a good transfer rate for large amounts of continuous data, but for random data, parity disks can be a bottleneck for write operations. The use of a separate check disk to protect the data although there is no mirror security is high, but the hard disk utilization has been greatly improved, for (n-1)/n.

In general, the use of RAID3, security can be guaranteed. Compared to RAID0, RAID3 is relatively slow in terms of reading and writing speed. The use of fault-tolerant algorithms and the size of the block determines the application that RAID uses, under normal circumstances, RAID3 more suitable for large file types and high security requirements of the application, such as video editing, hard disk broadcast machine, large database and so on.

RAID5 Technology

RAID5 is a storage solution that combines storage performance, data security, and storage costs. RAID5 can be understood as a compromise between RAID0 and RAID1. RAID5 can provide data security for the system, but the level of protection is lower than mirror and disk space utilization is higher than mirror. The RAID5 has a data read speed similar to that of RAID0, but with a more parity information, it is slower to write data than to write to a single disk. At the same time, because of the multiple data corresponding to a parity information, RAID5 disk space utilization is higher than RAID1, the storage cost is relatively low, is a more application of a solution.

RAID5 stores the data and the corresponding parity information on each disk that makes up the RAID5, and the parity information and the corresponding data are stored on separate disks, where the full data is stored on any N-1 block disk, meaning that there is space equivalent to a disk capacity for storing parity information. Therefore, when a disk in RAID5 is damaged, the data integrity is not affected, thus ensuring data security. When the damaged disk is replaced, the raid also automatically rebuilds the data on the disk with the remaining parity information to maintain the RAID5 's high reliability.

RAID7 Technology

RAID7 This is a new RAID standard, with its own intelligent real-time operating system and software tools for storage management, can be completely independent of the host running, do not occupy the host CPU resources. RAID7 can be seen as a storage computer (Storage computer), a real-time event-driven operating system that is used primarily for system initialization and for arranging all data transfers for RAID 7 disk arrays and converting them to the appropriate physical storage drives. Through the array computer board in its own system to set and control the read and write speed, the storage computer operating system can make the host I/O transfer performance to achieve the best. If a disk fails, recovery operations can be performed automatically, and the rebuild process for the backup disk can be managed. RAID7 has breached the technical architecture of previous RAID standards, and has adopted non-synchronous access, greatly reducing the data write bottleneck and increasing the I/O speed.

The so-called asynchronous access, that is, each I/O interface of RAID7 has a dedicated high-speed channel, as the flow path of data or control information, so can independently control the data access of each disk in its own system. If the RAID7 has n disks, then removing a check disk (used as a redundant calculation) can simultaneously handle the random read/write instructions from the N-1 host system, thereby significantly improving the I/O application. The RAID7 system's built-in real-time operating system also automatically optimizes the read/write instructions sent by the host, intelligently reading the data that may be read into the fast cache, thus greatly reducing the number of heads rotated and increasing the I/O speed. RAID7 can help users to effectively manage the increasingly large data storage systems, and make the system more efficient operation of at least one times, to meet the different needs of various users.

RAID7 all the I/O transfer is synchronous, can be controlled separately, so as to improve the parallelism of the system, improve the speed of the system access to data, each disk with high-speed buffer memory, real-time operating system can use any real-time operation of the chip, to meet the needs of different real-time systems. It is important to note that it introduces a high-speed buffer memory, which has advantages and disadvantages, because once the system loses power, the data in the buffer memory is lost, so it needs to work with the UPS. Of course, the price is very expensive for such a fast thing.

RAID10 Technology-- high reliability and efficient disk structure

This structure is nothing more than a band structure plus a mirror image structure, because the two structures have advantages and disadvantages, so they can complement each other, to achieve both high efficiency and high speed can also be mirrored with each other. We can combine the advantages and disadvantages of the two structures to understand this new structure. The price of this new structure is high and the expansibility is not good. It is mainly used in databases with small capacity but requiring speed and error control. RAID10 is the first mirror and then partition data. is to divide all the drives into two groups, considered as the lowest combination of RAID0, and then treat the two groups individually as RAID1 operations. The RAID10 has a good reading speed and has a higher data protection than RAID0. Outlined as "RAID0 's high-speed and RAID1 security".

Minimum number of disks required for each RAID level

---------------------------------------------------------------the END----------------------------------------- ------------------------

The above content is organized in the network article:

Reference: http://www.pcguide.com/ref/hdd/perf/raid/levels/single.htm

Http://blog.chinaunix.net/uid-16728139-id-3823924.html

           http://blogread.cn/it/article/4295?f=wb1