Disk Array (DiscArray) is composed of many disk drive or CD-ROM drive units according to certain rules, such as sub-section (Striping), block (Declustering), interleaved (Interleaving) composed of a fast, large capacity external memory subsystem . It is in the array controller's control and management, fast, parallel, or interleaved, and there is a strong fault tolerance. From the user point of view, although the disk array is composed of several, dozens or even hundreds of disc composition, it is still considered to be a single disk, the capacity can be up to several hundred ~ thousands of gigabytes, so this technique widely welcomed for multimedia systems.
disk array is the full name:
RedundanArrayofInexpensiveDisk, referred to as RAID technology. It was in 1988 by the University of California, Berkeley professor who raised DavidPatterson disk redundancy. Since then, the disk array technology developed rapidly, and gradually becoming more mature. Has now been recognized with the following basic eight series.
1.RAID0 (0 级 disk array)
RAID0, also known as data block, that is, the data distributed across multiple disks, there is no fault tolerance measures. Its capacity and data transfer rate is N times the stand-alone capacity, N constitute a disk array for the total number of disk drives,
barcelona jersey cheap, I / O transfer rate high, but the average time to failure MTTF (MeanTimeToFailure) only a single one-disk machine N Therefore, the reliability of disk array zero the worst.
2.RAID1 (1 级 disk array)
RAID1, also known as mirroring (Mirror) disk, use mirroring to improve reliability, fault-tolerant. That each job has a mirrored disk set, each write data must be written to the mirror disk, read data, read out only from the working plate. Once the job offer immediately transferred to the mirror disk failure, read data from the mirrored disk, and then by the system recovery disk and then the correct data. Therefore, data can be reconstructed in this way, but the work plate and mirror plate must be kept one to one relationship. The high reliability of disk array, but its effective capacity is reduced to less than half the total capacity. Therefore, commonly used in RAID1 error rate requirements are very strict on the applications, such as finance, banking and other fields.
3.RAID2 (2 级 disk array)
RAID2 also known as digital crossover, which uses Hamming codes for error test disc, without in each sector after the CRC (CyclicReDundancycheck) test. Hamming code is a (n, k) linear block code, n is the length of codeword, k is the number of bits of data, r is the number of bits used to test, so there is: n = 2r-1r = n-k Therefore, bit-interleaved
most beneficial for the Hamming code inspection. This dish suitable for large data read and write. However, the overhead is too much redundant information, to prevent the wide application of such disks.
4.RAID3 (3 级 disk array)
RAID3 parallel transmission for the single disk fault tolerant disk arrays. It is characterized by the test panel is reduced to a (RAID2 parity disk into multiple, DAID1 test plate for 1 to 1), in bits or bytes of data stored in the drive way (decentralized sector recorded the same number in the group on each disk drive). The advantage is that the bandwidth of the entire array can be fully utilized, the bulk data transfer time decreases; The disadvantage is that each read and write to affect the entire group, you can only complete one I / O.
5.RAID4 (4 级 disk array)
RAID4 a group can be independently read and write the disk array. The parity disk is only one.
RAID4 and RAID3 difference is: RAID3 is based on bit or byte interleaving, and RAID4 is based on the block (sector) access, can be individually operated on a disk, it did not like RAID3 as even if every small I / O operations must involve the whole group, only the group involved two discs (one data set, a test plate) can be. To improve a small amount of data I / O rate.
6.RAID5 (5 级 disk array)
RAID5 is a rotating parity array of independent access. It RAID1,
nba jerseys, 2,3,4 different points of the disk array is that it has no fixed parity disk, but according to certain rules to the redundant parity information is distributed evenly to all disks in the array on the respective . Then the disk drive on the same data also verify existing information. This change solves the problem of disk contention checking, so DAID5 in the same group within the allowed multiple concurrent writes. So that is suitable for large data RAID5 operations, but also for all transactions. It is a fast, high capacity and fault-tolerant and reasonable distribution of the disk array.
7.RAID6 (6 级 disk array)
RAID6 is a two-dimensional independent access parity disk array. It's redundant check, error correction information distributed evenly across all disks, but still variable size data blocks to cross the way to save on disk. Double disk array can be allowed such an error.
8.RAID7 (7 级 disk array)
RAID7 RAID6 is based on the use of cache technology, which allows transfer rates and response speed are greatly improved. Cache is a cache, that data is written to the disk array before the first write cache. Generally use the cache block size and disk array data block the same size, that is, a cache block corresponding to a disk block. When the data were written to write two separate cache, so even if one has a cache fails, the data is not lost. Write directly in the cache-level response, and then go to the disk array. Written to the disk array data from the cache, the same track data will be completed in one operation, avoiding a lot of questions written in block data several times to improve the speed. In reading, the host is also read out directly from the cache rather than read from the disk array, reducing the number of disk read operations, this is more fully utilize the disk bandwidth.
this cache and disk array technology combined make up less disk array (such as block write requests and poor response to defects), so that the whole system in an efficient, fast, large capacity, high reliability and flexible, and convenient storage system provides to the user, to meet the current technology development, particularly the needs of multimedia systems.
resolve key technical
disk array storage technology in the computer technology received extensive attention, the server storage technology is the industry concerned about hot spots. It comes to server storage technology, people almost immediately and SCSI (Small Computer Systems Interface) technology together. Despite the cheap IDE hard disk performance, capacity and other key technical indicators have been greatly improved to meet or exceed the original server storage devices. However, due to the rapid development and popularity of Internet, network servers are becoming increasingly large scale. Meanwhile, Internet not only for the web server itself, but also made the server storage demanding. Endless demand prompted the rapid development of server storage. The server storage disk array is a relatively mature technology, but also relatively more common in the market, one of the large-capacity peripheral.
at the high end, the traditional storage model in terms of scale, or safety, or performance, are unable to meet specific application needs of the expanding storage. Such as storage area network (SAN) and other new technologies or applications are constantly emerging, and new storage solutions for emerging architecture, server storage by direct-attached storage (DAS) to storage networking technologies (NAS) expansion areas. In the low end, with the continuous development of hardware technology, strong market demand in the promotion,
cheap kobe bryant jersey, localized, based on direct attached disk array storage technology, speed, performance, storage capacity and so continue to a new level. And, in order to meet the security of stored data users, access speeds and large demand for storage capacity, disk array storage technology from the emphasis on technical innovation, attention to system optimization, technology-driven technology solutions to promote the industrial stage into the stressed standards, focus on market size, the mature product as the leading product adoption period.
review the development process of the disk array, and SCSI technology has been closely related, some vendors have introduced proprietary technologies such as IBM, SSA (Serial Storage Architecture) technology, due to compatibility and upgrade ability is not satisfactory, impact on the market are far less extensive SCSI technology. SCSI technology compatibility as well, the market demand,
cheap barca jersey, making the SCSI technology has developed rapidly. From the original 5MB / s transfer rate of SCSI-1, has been developed to LVD interface, 160MB / s transfer speed Ultra 160 SCSI, 320MB / s transfer speed Ultra 320 SCSI interface, will occur in 2001 (see Table 1 .) From the current market to see, Ultra 3 SCSI technology and RAID (Redundant Array of Inexpensive Disks) disk array storage technology should be the mainstream technology.
SCSI technology
SCSI itself is for the minicomputer (as distinct from computer terms) custom storage interface, SCSI Protocol Version 1 version is also only provides 5MB / s transfer rate of SCSI-1 bus types, interface definition of technical standards for cable specifications. As technology advances, SCSI Protocol Version 2 revised versions were larger, follow the SCSI-2 protocol 16-bit data bandwidth, high frequency of the SCSI storage devices start to appear and become the market mainstream products, but also makes fast SCSI technology firmly occupy the server storage market. SCSI-3 protocol is increased to meet the special equipment needed for command set protocol, so both SCSI protocol to adapt to traditional parallel transmission equipment, but also to adapt to the emergence of some new serial device communication needs, such as Fibre Channel Protocol (FCP) Serial Storage Protocol (SSP), serial bus protocols. Gradually, the \
Typically, users care about on the SCSI bus hardware, the work of the different SCSI model means there are different maximum transmission speed. Such as 40MB / s for Ultra SCSI, 160MB / s for Ultra 3 SCSI and so on. But the maximum transfer speed does not mean that devices can be achieved during normal operation, the average access speed, does not mean the different between the SCSI mode access speed must exist \SCSI controller and SCSI hard disk access speed of the actual models, technical data, and the transmission cable length, the anti-interference ability and other factors are closely related. SCSI bus must focus on improving the efficiency of the configuration of SCSI devices and the transmission cable terminal specifications and quality. It can be seen, Ultra 3 mode access speed is less than the actual visit to Ultra Wide mode 2 times the speed of the actual visit.
general, use high speed SCSI hard drives, SCSI channel on the appropriate increase in the number of connected hard drive, optimize application access to data on the disk mode, etc., can greatly improve the SCSI bus, the actual transmission speed. In particular, be noted that, under the same conditions, different disk access mode to obtain the SCSI bus, the actual transfer rate can vary by several times, on the application of the optimization is to obtain high-speed memory access must be the focus of attention, which is often Some users are ignored. 4KB block of data by random access six SCSI hard disk, SCSI bus speed of the actual visit to 2.74MB / s, SCSI bus, the efficiency of the bus bandwidth is only 1.7%; in full under the same conditions, according to 256KB of data blocks the order of the hard disk read and write, SCSI bus, the actual access speed 141.2MB / s, SCSI bus, the efficiency of the bus bandwidth of up to 88%.
as the transmission speed increases, the signal transmission and interference in the signal attenuation becomes more and more prominent, to some extent the end device may help reduce the signal wave reflection, the role of improving the signal quality. At the same time, LVD (Low-Voltage Differential) technology is also increasing. Mode is LVD and SE (Single-Ended) corresponding to the mode, it can be very good resistance to transmission interference, extended signal transmission distance. Meanwhile, Ultra 2 SCSI and Ultra 3 SCSI model is also a dedicated twisted pair through the use of SCSI cables to improve signal transmission quality.
the concept of the disk array, high-capacity hard drive does not refer to a single hard drive capacity, but to a single hard drive through the RAID technology, according to RAID level to form a larger capacity hard drive. Therefore, disk array technology, RAID technology is more critical, while the selected RAID level according to different, get the \
RAID is a very mature technology, but because of its relatively expensive price, configuration is not convenient, the relative lack of professional and technical personnel, the application is not very popular. According to statistics, 75% of worldwide server systems are not configured RAID. Because the server storage needs of data security, scalability and so on have become increasingly demanding, RAID market development potential. RAID technology is an industry standard, the manufacturers of the definition of RAID levels are not the same. The current definition of RAID levels is widely recognized by the industry's only 4, RAID 0, RAID 1, RAID 0 +1 and RAID 5.
RAID 0 is no data redundancy storage space striping, with a low-cost, high read and write performance, high utilization of the RAID-level storage for Video / Audio signal storage, dumping and other temporary files extremely strict requirements on the speed of the particular application. However, no data redundancy, significantly reducing its security to constitute a disk array of any damage will bring about catastrophic loss of data. So, in RAID 0 configuration four or more in the hard disk, for general applications is unwise.
RAID 1 mirror is two hard drives full of data, security is good, simple technology, easy management, read and write performance are good. But it can not be extended (single block of hard disk capacity), data space wastage, strictly speaking, should not be called \
RAID 0 +1 combines RAID 0 and RAID 1 the characteristics of independent disks configured as RAID 0, two complete sets of RAID 0 mirror each other. It is good to read and write performance, high security, but the cost of building into a large array of data space utilization is low, can not be called cost-effective solution.
RAID 5 is the most widely used RAID technology. Independent of each block striped partitioning the hard disk, the same parity stripe area (XOR), parity data distributed evenly in each hard disk. Construction of the n-drive RAID 5 array can have n-1 hard drive capacity, storage space utilization is very high (see Figure 6). Any piece of data loss on the hard disk, can be derived through the calibration data. It is the biggest difference between RAID 3 parity data is distributed evenly to each disk drive. RAID 5 with data security, read and write speed, space utilization, advantages, is widely used, but the downside is that a hard disk failure, the whole system performance greatly reduced.
for RAID 1, RAID 0 +1, RAID 5 array with hot-plug (also known as heat can be replaced) technology, the online data recovery, that is, when a RAID disk array for any damage, without the user off or stop the application services to replace the failed hard disk can repair the system and recover data to the realization of HA (High Availability) high availability systems is important.
the manufacturers are constantly introduced various RAID levels and standards. Such as higher security, from the start mirrored RAID controller RAID; faster read and write speeds, as constituting the hard disk RAID configuration of each block of the RAID Cache CPU and so on, but not universal. IDE hard drives with RAID technology is the building of a new and emerging technology direction, have greater impact on the market, its outstanding advantage is very cheap build RAID arrays. Can support the current IDE RAID RAID 0, RAID 1 and RAID 0 +1 are three levels, up to four IDE hard disk. IDE devices due to scalability limitations, but also because lack of IDE devices also support hot-replaceable reasons, IDE RAID applications not more.
In short, the development is the eternal theme in the server storage technology is no exception. On the one hand, a number of giant companies try to introduce new concepts or standards, to lead the server and storage technology development, more representative, such as Intel's IA-64 architecture pushing and storage concepts; on the other hand, is committed to storage professional firms with existing technology and industry standards, promote SCSI, RAID, Fibre Channel and other storage technologies and solutions based on existing rapid updating and development. In a market economy conditions, the only criterion for testing technology is the market recognition. Market calls for good technology, and new technology must play a role to promote the market forward in order to be widely accepted and recognized. With the development of high-performance computer market, high-performance ratio, high reliability, high security storage of new technologies will continue to emerge.
disk array on the market today there are many products, the user in the selection process of the disk array products, but also according to their needs for choice, and now to a few disk array products, but also for the needs of users of disk array products provide some options. Table 2 lists the disk array of several key technical indicators.
----------------------------------------------- ---------------------------------
Tips: disk array reliability and availability of reliable
, referring to the hard disk under the given conditions the probability of failure. Availability, refers to the use of the hard disk may be used in some time. Hard disk arrays can improve system reliability. It can be seen from Table 3 RAID disk subsystem with a single hard disk subsystem,
lebron james jerseys, the reliability of the comparison.
In addition, system availability, system availability than a single hard drive data redundancy in the disk array is not better, and the availability of redundant disk array is much better than a single hard drive. This is because the redundant disk array allows a single hard disk errors,
minnesota twins jerseys, and continue to work; a hard drive failure recovery time after the system has also greatly reduced (compared with recovering data from tape); redundant disk array fails, the data on the hard disk is the failure time data, replaced the hard drive will also include failure data. However, to be fully fault-tolerant performance, other components of computer hard disk subsystem must also be redundant.
Brand New 12.1″ WXGA Glossy Laptop LCD Screen For Acer Travelmate ...
Linear Mode
