The full name of the disk array is: Redundan Array of In Expensive Disk, referred to as RAID technology. It is a disk redundancy technology proposed by Professor David Patterson and others in 1988 at the University of California, Berkeley. Since then, disk array technology has developed rapidly and has gradually matured. The following eight series are now generally recognized.
1.RAID0 (level 0 disk array)
RAID0, also known as data blocking, distributes data on multiple disks without fault tolerance. Its capacity and data transmission rate are N times the stand-alone capacity. N is the total number of disk drives that make up the disk array. The I/O transmission rate is high, but the average time between failures MTTF (Mean Time To Failure) is only N of a single disk drive. In one part, the reliability of the zero-order disk array is the worst.
2. RAID 1 (level 1 disk array)
RAID1, also known as mirror disk, uses image fault tolerance to improve reliability. That is, each working disk has a mirror disk. Each time data is written, the mirror disk must be written at the same time. When reading data, only the working disk is read. Once the working disk fails, it is immediately transferred to the mirror disk, data is read from the mirror disk, and then the correct data of the working disk is restored by the system. Therefore, data in this way can be reconstructed, but the working disk and the mirror disk must maintain a one-to-one correspondence. This type of disk array is highly reliable, but its effective capacity is reduced to less than half of the total capacity. Therefore, RAID1 is often used in applications where the error rate is very strict, such as financial and financial fields.
3.RAID 2 (level 2 disk array)
RAID 2 is also referred to as bit-interleaving. It uses Hamming code for discontinuity checking and does not require CRC (Cyclic Re Dundancy check) after each sector. Hamming code is an (n, k) linear block code, n is the length of the code word, k is the number of data bits, r is the number of bits used for the test, it is: n = 2r-1r = n-k
Therefore, bit-by-bit interleaving is most beneficial for Hamming code verification. This kind of disk is suitable for reading and writing big data. However, the overhead of redundant information is still too large to prevent the widespread use of such disks.
4. RAID 3 (level 3 disk array)
RAID 3 is a single-disc fault-tolerant parallel transmission array disk. It is characterized by reducing the number of checker disks to one (multiple RAID2 check-disks, 1 to 1 for DAID1 check-disks), and the data is stored on each disk as bits or bytes (distributed records in the same sector number in the group. On each disk drive). Its advantage is that the bandwidth of the entire array can be fully utilized and the batch data transmission time is reduced. The disadvantage is that each read and write has to affect the entire group, and only one I/O can be completed at a time.
5. RAID 4 (4-level disk array)
RAID 4 is an array that reads and writes independently from each disk in the group. There is only one check plate.
The difference between RAID 4 and RAID 3 is that RAID 3 is accessed by bit or by byte, while RAID 4 is accessed by blocks (sectors) and can be operated on a single disk. It does not need RAID 3 In that case, even if each small I/O operation involves the entire group, only two disk units in the group (one data disk, one test disk) can be involved. This improves the I/O rate for small amounts of data.
6.RAID 5 (level 5 disk array)
RAID 5 is an array of spin-parity independent accesses. It differs from the RAID 1, 2, 3, and 4 disk arrays in that it does not have a fixed parity disk. Instead, it evenly distributes redundant parity information to all disks to which the array belongs according to a certain rule. on. So there are both data and parity information on the same disk unit. This change resolves the issue of contention for the parity disk, so DAID5 allows multiple write operations concurrently within the same group. Therefore, RAID5 is suitable for the operation of large amounts of data and is also suitable for various transaction processing. It is a fast, large-capacity and fault-tolerant distribution disk array.
7. RAID 6 (Class 6 Disk Array)
RAID 6 is a double-even parity independent access disk array. Its redundant check and error correction information is evenly distributed on all disks, and data is still stored on the disks in a cross-like manner with variable-sized blocks. This type of disk array can tolerate double disk errors.
8. RAID 7 (Class 7 Disk Array)
RAID 7 is based on RAID 6, using cache technology, which makes the transmission rate and response speed have greatly improved. Cache is a kind of cache, that is, data is written into the cache before it is written to the disk array. Generally, the cache partition size is the same as the data partition size in the disk array, that is, one cache partition corresponds to one disk partition. Write data to two separate caches, so even if one of the caches fails, the data will not be lost. The write operation will respond directly at the cache level before going to the disk array. When the data is written from the cache to the disk array, the data of the same track will be completed in one operation, which avoids the problem of many blocks of data being written many times and improves the speed. In reading, the host is also directly read from the cache, rather than read from the array disk, reducing the number of read operations with the disk, so that more fully utilize the disk bandwidth.
In this way, the combination of the cache and the disk array technology can make up for the shortage of the disk array (such as the poor response of the block write request response), thereby providing the entire system with an efficient, fast, large-capacity, high-reliability, flexible, and convenient storage system. Users, thus meeting the needs of the current technological development, especially the needs of multimedia systems.
Key technologies for resolving disk arrays
Storage technology has received extensive attention in computer technology. Server storage technology is a hot topic in the industry. When it comes to server storage technology, people are almost immediately associated with the Small Computer Systems Interface (SCSI) technology. Although inexpensive IDE hard drives have been greatly improved in key technical indicators such as performance and capacity, they can meet or exceed the needs of existing server storage devices. However, due to the popularity and rapid development of the Internet, the size of network servers has also become larger and larger. At the same time, the Internet not only imposes stringent requirements on the network server itself, but also on the server storage technology. Unending market demand has prompted the rapid development of server storage technology. The disk array is a mature type of server storage technology, and it is also one of the large-capacity peripherals that are more common in the market.
At the high end, the traditional storage model cannot meet the expanding storage requirements of special applications in terms of scale, security, or performance. New technologies or applications such as storage area networks (SANs) continue to emerge, and new storage architectures and solutions are emerging. Server storage technology extends from direct-attached storage (DAS) to storage networking technology (NAS). At the mid-to-low end, with the continuous development of hardware technology and driven by strong market demand, localized, direct-attached disk array storage technology continues to take a new level in terms of speed, performance, and storage capacity. In addition, in order to meet the needs of users for storage data security, access speed, and large storage capacity, disk array storage technology also focuses on technological innovation and system optimization, and technology-driven technology promotion period has gradually entered the emphasis on industry. Standards, market size, mature product-led product popularity period.
Reviewing the development history of disk arrays has been closely related to the development of SCSI technology. The proprietary technologies introduced by some vendors, such as IBM's SSA (Serial Storage Architecture) technology, are not satisfactory due to the compatibility and upgrade capabilities. The impact is far less than the extensive SCSI technology. Due to the good compatibility of SCSI technology and strong market demand, SCSI technology has developed rapidly. From the original SCSI-1 with a 5 MB/s transfer rate, Ultra 160 SCSI has been developed to 160 MB/s for LVD interfaces, and the Ultra 320 SCSI interface for 320 MB/s is also available in 2001 (see table). 1). From the current market perspective, Ultra3SCSI technology and RAID (Redundant Array of In Expensive Disks) technology should also be the mainstream technology for disk array storage.
The SCSI technology SCSI itself is a customized storage interface for minicomputers (in contrast to microcomputers). The Version 1 SCSI protocol also specifies SCSI-1 bus types, interface definitions, and cable specifications for 5 MB/s transfer speeds. standard. With the development of technology, Version 2 of the SCSI protocol has been greatly revised. Following the 16-bit data bandwidth of the SCSI-2 protocol, high-frequency SCSI storage devices have emerged one after another and become mainstream products in the market, and also make the SCSI technology secure. Securely occupied the server's storage market. The SCSI-3 protocol adds a command set that can meet the special device protocols, making the SCSI protocol adaptable to both traditional parallel transmission devices and the communication needs of some of the latest serial devices, such as the Fibre Channel Protocol (FCP). , serial storage protocol (SSP), serial bus protocol, etc. Gradually, the concept of "minicomputers" began to weaken, and the concepts of "high-performance computers" and "servers" were strengthened in people's minds. SCSI was once a standard for users to distinguish "servers" and PCs from hardware.
In general, the user's interest in the SCSI bus is placed on the hardware. Different SCSI modes of operation mean different maximum transfer speeds. Such as 40 MB/s Ultra SCSI, 160 MB/s Ultra 3 SCSI and so on. However, the maximum transmission speed does not represent the average access speed that can be achieved when the device works normally, nor does it mean that there is an inevitable “multiplier†relationship between the access speeds of different SCSI operating modes. The actual access speed of the SCSI controller is closely related to the SCSI hard disk model, technical parameters, transmission cable length, and anti-jamming capability. To improve the efficiency of the SCSI bus, you must pay attention to the configuration and quality of the SCSI device and the transmission cable. It can be seen that the actual access speed obtained in the Ultra 3 mode is less than twice the actual access speed in the Ultra Wide mode.
In general, choosing a high-speed SCSI hard disk, appropriately increasing the number of hard disks connected to the SCSI channel, and optimizing the application access to disk data can greatly increase the actual transmission speed of the SCSI bus. In particular, it should be noted that, under the same conditions, the actual transmission speed of the SCSI bus obtained under different disk access modes can be several tens of times different. Optimization of the application is the key point that must be paid attention to when obtaining high-speed storage access, and this is often Some users ignore it. When accessing 6 SCSI HDDs randomly by 4 KB data blocks, the actual access speed of the SCSI bus is 2.74 MB/s. The SCSI bus's work efficiency is only 1.7% of the bus bandwidth. Under the condition of complete change, the data blocks are 256 KB. Sequential reading and writing of the hard disk, the actual access speed of the SCSI bus is 141.2 MB/s, and the work efficiency of the SCSI bus is as high as 88% of the bus bandwidth.
With the increase of transmission speed, the problem of signal attenuation and interference in the signal transmission process becomes more and more prominent. The terminator can play a role in reducing signal wave reflection and improving signal quality to some extent. At the same time, the application of LVD (Low-Voltage Differential) technology is also increasing. The LVD working mode corresponds to the SE (Single-Ended) mode, which can well resist transmission interference and extend the signal transmission distance. At the same time, the Ultra 2 SCSI and Ultra 3 SCSI modes also improve the quality of signal transmission by using dedicated twisted-pair SCSI cables.
In the concept of a disk array, a large-capacity hard disk does not mean that a single hard disk has a large capacity, but means that a single hard disk is combined into a larger-capacity hard disk according to RAID levels through RAID technology. Therefore, in the disk array technology, RAID technology is more critical. At the same time, depending on the selected RAID level, the function of the "big hard disk" is different.
RAID is a very mature technology, but its application is not very popular due to its relatively expensive and inconvenient configuration and lack of relatively professional technicians. According to statistics, 75% of the world's server systems currently do not have RAID configuration. As the storage requirements of servers increase the requirements for data security and scalability, the development potential of the RAID market is huge. RAID technology is an industry standard, and various vendors have different definitions of RAID levels. At present, the definition of RAID level can be widely recognized only four kinds of industry, RAID0, RAID1, RAID0+1 and RAID5.
RAID0 is a storage space stripe without data redundancy. It has a RAID level with low cost, high read/write performance, and high storage space utilization. It is suitable for very fast speed requirements such as Video/Audio signal storage and temporary file storage. Strict special applications. However, because there is no data redundancy, its security is greatly reduced, and any piece of hard disk that constitutes an array is damaged, causing catastrophic data loss. Therefore, configuring more than four hard disks in RAID 0 is unwise for general applications.
RAID1 is a complete mirroring of two hard disk data. It has good security, simple technology, convenient management, and good read/write performance. However, it cannot be expanded (a single hard disk capacity) and the data space is wasted. In a strict sense, it should not be called an "array."
RAID0+1 combines the characteristics of RAID0 and RAID1. The independent disks are configured as RAID0. The two complete RAID0 mirrors each other. It has excellent read and write performance and high security. However, the cost of constructing an array is high, and the utilization of data space is low. It cannot be called a cost-effective solution.
RAID5 is currently the most widely used RAID technology. Each piece of independent hard disk is stripped and divided, and the same stripe area is subjected to parity check (XOR operation). The parity data is evenly distributed on each hard disk. A RAID 5 array built with n hard disks can have n-1 hard disk capacity and very high storage space utilization (see Figure 6). Loss of data on any hard disk can be deduced from the verification data. The biggest difference between it and RAID3 is whether the parity data is evenly distributed to each hard disk. RAID5 has the advantages of data security, fast read and write speeds, high space utilization, and is widely used. However, the disadvantage of RAID5 is that the performance of the entire system is greatly reduced after a hard disk fails.
For RAID1, RAID0+1, and RAID5 arrays, hot-swappable (also called hot-replaceable) technology enables online data recovery. That is, when any hard disk in a RAID array is damaged, users do not need to shut down or stop the application service. It is of great significance to realize high availability (HA) systems by replacing failed hard disks, repairing systems, and recovering data.
Various vendors continue to introduce various RAID levels and standards. For example, higher security, starting from the RAID controller mirroring RAID; faster read and write speeds, configuring RAID for each disk and RAID CPU, and so on, but are not universal. The use of IDE hard drives to build RAID technology is a new emerging technology direction and has a greater impact on the market. Its outstanding advantage is that it is very cheap to build RAID arrays. At present, IDERAID can support three levels of RAID0, RAID1, and RAID0+1, and supports up to four IDE hard disks. Due to the limitation of the scalability of IDE devices, and because of the lack of hot and replaceable technical support for IDE devices, there are not many IDERAID applications.
In short, development is an eternal theme, and it is no exception in the area of ​​server storage technology. On the one hand, some giant manufacturers try to introduce new concepts or standards to lead the development direction of servers and storage technologies, such as the Intel-driven IA-64 architecture and storage concepts. On the other hand, they are dedicated to storage. Based on existing technologies and industry standards, professional vendors promote the rapid update and development of existing storage technologies and solutions based on SCSI, RAID, and Fibre Channel. Under market economy conditions, the only criterion for the inspection of technological development is market recognition. Markets call for good technology, and new technologies must be widely accepted and recognized when they must promote market development. With the development of the high-performance computer market, new high-performance, high-reliability, and high-security storage technologies will continue to emerge.
Currently, there are many disk array products on the market. Users must choose according to their own needs in the process of selecting disk array products. Now list several disk array products and provide users with options for disk array products. . Table 2 lists the main technical specifications of several disk arrays.
1.RAID0 (level 0 disk array)
RAID0, also known as data blocking, distributes data on multiple disks without fault tolerance. Its capacity and data transmission rate are N times the stand-alone capacity. N is the total number of disk drives that make up the disk array. The I/O transmission rate is high, but the average time between failures MTTF (Mean Time To Failure) is only N of a single disk drive. In one part, the reliability of the zero-order disk array is the worst.
2. RAID 1 (level 1 disk array)
RAID1, also known as mirror disk, uses image fault tolerance to improve reliability. That is, each working disk has a mirror disk. Each time data is written, the mirror disk must be written at the same time. When reading data, only the working disk is read. Once the working disk fails, it is immediately transferred to the mirror disk, data is read from the mirror disk, and then the correct data of the working disk is restored by the system. Therefore, data in this way can be reconstructed, but the working disk and the mirror disk must maintain a one-to-one correspondence. This type of disk array is highly reliable, but its effective capacity is reduced to less than half of the total capacity. Therefore, RAID1 is often used in applications where the error rate is very strict, such as financial and financial fields.
3.RAID 2 (level 2 disk array)
RAID 2 is also referred to as bit-interleaving. It uses Hamming code for discontinuity checking and does not require CRC (Cyclic Re Dundancy check) after each sector. Hamming code is an (n, k) linear block code, n is the length of the code word, k is the number of data bits, r is the number of bits used for the test, it is: n = 2r-1r = n-k
Therefore, bit-by-bit interleaving is most beneficial for Hamming code verification. This kind of disk is suitable for reading and writing big data. However, the overhead of redundant information is still too large to prevent the widespread use of such disks.
4. RAID 3 (level 3 disk array)
RAID 3 is a single-disc fault-tolerant parallel transmission array disk. It is characterized by reducing the number of checker disks to one (multiple RAID2 check-disks, 1 to 1 for DAID1 check-disks), and the data is stored on each disk as bits or bytes (distributed records in the same sector number in the group. On each disk drive). Its advantage is that the bandwidth of the entire array can be fully utilized and the batch data transmission time is reduced. The disadvantage is that each read and write has to affect the entire group, and only one I/O can be completed at a time.
5. RAID 4 (4-level disk array)
RAID 4 is an array that reads and writes independently from each disk in the group. There is only one check plate.
The difference between RAID 4 and RAID 3 is that RAID 3 is accessed by bit or by byte, while RAID 4 is accessed by blocks (sectors) and can be operated on a single disk. It does not need RAID 3 In that case, even if each small I/O operation involves the entire group, only two disk units in the group (one data disk, one test disk) can be involved. This improves the I/O rate for small amounts of data.
6.RAID 5 (level 5 disk array)
RAID 5 is an array of spin-parity independent accesses. It differs from the RAID 1, 2, 3, and 4 disk arrays in that it does not have a fixed parity disk. Instead, it evenly distributes redundant parity information to all disks to which the array belongs according to a certain rule. on. So there are both data and parity information on the same disk unit. This change resolves the issue of contention for the parity disk, so DAID5 allows multiple write operations concurrently within the same group. Therefore, RAID5 is suitable for the operation of large amounts of data and is also suitable for various transaction processing. It is a fast, large-capacity and fault-tolerant distribution disk array.
7. RAID 6 (Class 6 Disk Array)
RAID 6 is a double-even parity independent access disk array. Its redundant check and error correction information is evenly distributed on all disks, and data is still stored on the disks in a cross-like manner with variable-sized blocks. This type of disk array can tolerate double disk errors.
8. RAID 7 (Class 7 Disk Array)
RAID 7 is based on RAID 6, using cache technology, which makes the transmission rate and response speed have greatly improved. Cache is a kind of cache, that is, data is written into the cache before it is written to the disk array. Generally, the cache partition size is the same as the data partition size in the disk array, that is, one cache partition corresponds to one disk partition. Write data to two separate caches, so even if one of the caches fails, the data will not be lost. The write operation will respond directly at the cache level before going to the disk array. When the data is written from the cache to the disk array, the data of the same track will be completed in one operation, which avoids the problem of many blocks of data being written many times and improves the speed. In reading, the host is also directly read from the cache, rather than read from the array disk, reducing the number of read operations with the disk, so that more fully utilize the disk bandwidth.
In this way, the combination of the cache and the disk array technology can make up for the shortage of the disk array (such as the poor response of the block write request response), thereby providing the entire system with an efficient, fast, large-capacity, high-reliability, flexible, and convenient storage system. Users, thus meeting the needs of the current technological development, especially the needs of multimedia systems.
Key technologies for resolving disk arrays
Storage technology has received extensive attention in computer technology. Server storage technology is a hot topic in the industry. When it comes to server storage technology, people are almost immediately associated with the Small Computer Systems Interface (SCSI) technology. Although inexpensive IDE hard drives have been greatly improved in key technical indicators such as performance and capacity, they can meet or exceed the needs of existing server storage devices. However, due to the popularity and rapid development of the Internet, the size of network servers has also become larger and larger. At the same time, the Internet not only imposes stringent requirements on the network server itself, but also on the server storage technology. Unending market demand has prompted the rapid development of server storage technology. The disk array is a mature type of server storage technology, and it is also one of the large-capacity peripherals that are more common in the market.
At the high end, the traditional storage model cannot meet the expanding storage requirements of special applications in terms of scale, security, or performance. New technologies or applications such as storage area networks (SANs) continue to emerge, and new storage architectures and solutions are emerging. Server storage technology extends from direct-attached storage (DAS) to storage networking technology (NAS). At the mid-to-low end, with the continuous development of hardware technology and driven by strong market demand, localized, direct-attached disk array storage technology continues to take a new level in terms of speed, performance, and storage capacity. In addition, in order to meet the needs of users for storage data security, access speed, and large storage capacity, disk array storage technology also focuses on technological innovation and system optimization, and technology-driven technology promotion period has gradually entered the emphasis on industry. Standards, market size, mature product-led product popularity period.
Reviewing the development history of disk arrays has been closely related to the development of SCSI technology. The proprietary technologies introduced by some vendors, such as IBM's SSA (Serial Storage Architecture) technology, are not satisfactory due to the compatibility and upgrade capabilities. The impact is far less than the extensive SCSI technology. Due to the good compatibility of SCSI technology and strong market demand, SCSI technology has developed rapidly. From the original SCSI-1 with a 5 MB/s transfer rate, Ultra 160 SCSI has been developed to 160 MB/s for LVD interfaces, and the Ultra 320 SCSI interface for 320 MB/s is also available in 2001 (see table). 1). From the current market perspective, Ultra3SCSI technology and RAID (Redundant Array of In Expensive Disks) technology should also be the mainstream technology for disk array storage.
The SCSI technology SCSI itself is a customized storage interface for minicomputers (in contrast to microcomputers). The Version 1 SCSI protocol also specifies SCSI-1 bus types, interface definitions, and cable specifications for 5 MB/s transfer speeds. standard. With the development of technology, Version 2 of the SCSI protocol has been greatly revised. Following the 16-bit data bandwidth of the SCSI-2 protocol, high-frequency SCSI storage devices have emerged one after another and become mainstream products in the market, and also make the SCSI technology secure. Securely occupied the server's storage market. The SCSI-3 protocol adds a command set that can meet the special device protocols, making the SCSI protocol adaptable to both traditional parallel transmission devices and the communication needs of some of the latest serial devices, such as the Fibre Channel Protocol (FCP). , serial storage protocol (SSP), serial bus protocol, etc. Gradually, the concept of "minicomputers" began to weaken, and the concepts of "high-performance computers" and "servers" were strengthened in people's minds. SCSI was once a standard for users to distinguish "servers" and PCs from hardware.
In general, the user's interest in the SCSI bus is placed on the hardware. Different SCSI modes of operation mean different maximum transfer speeds. Such as 40 MB/s Ultra SCSI, 160 MB/s Ultra 3 SCSI and so on. However, the maximum transmission speed does not represent the average access speed that can be achieved when the device works normally, nor does it mean that there is an inevitable “multiplier†relationship between the access speeds of different SCSI operating modes. The actual access speed of the SCSI controller is closely related to the SCSI hard disk model, technical parameters, transmission cable length, and anti-jamming capability. To improve the efficiency of the SCSI bus, you must pay attention to the configuration and quality of the SCSI device and the transmission cable. It can be seen that the actual access speed obtained in the Ultra 3 mode is less than twice the actual access speed in the Ultra Wide mode.
In general, choosing a high-speed SCSI hard disk, appropriately increasing the number of hard disks connected to the SCSI channel, and optimizing the application access to disk data can greatly increase the actual transmission speed of the SCSI bus. In particular, it should be noted that, under the same conditions, the actual transmission speed of the SCSI bus obtained under different disk access modes can be several tens of times different. Optimization of the application is the key point that must be paid attention to when obtaining high-speed storage access, and this is often Some users ignore it. When accessing 6 SCSI HDDs randomly by 4 KB data blocks, the actual access speed of the SCSI bus is 2.74 MB/s. The SCSI bus's work efficiency is only 1.7% of the bus bandwidth. Under the condition of complete change, the data blocks are 256 KB. Sequential reading and writing of the hard disk, the actual access speed of the SCSI bus is 141.2 MB/s, and the work efficiency of the SCSI bus is as high as 88% of the bus bandwidth.
With the increase of transmission speed, the problem of signal attenuation and interference in the signal transmission process becomes more and more prominent. The terminator can play a role in reducing signal wave reflection and improving signal quality to some extent. At the same time, the application of LVD (Low-Voltage Differential) technology is also increasing. The LVD working mode corresponds to the SE (Single-Ended) mode, which can well resist transmission interference and extend the signal transmission distance. At the same time, the Ultra 2 SCSI and Ultra 3 SCSI modes also improve the quality of signal transmission by using dedicated twisted-pair SCSI cables.
In the concept of a disk array, a large-capacity hard disk does not mean that a single hard disk has a large capacity, but means that a single hard disk is combined into a larger-capacity hard disk according to RAID levels through RAID technology. Therefore, in the disk array technology, RAID technology is more critical. At the same time, depending on the selected RAID level, the function of the "big hard disk" is different.
RAID is a very mature technology, but its application is not very popular due to its relatively expensive and inconvenient configuration and lack of relatively professional technicians. According to statistics, 75% of the world's server systems currently do not have RAID configuration. As the storage requirements of servers increase the requirements for data security and scalability, the development potential of the RAID market is huge. RAID technology is an industry standard, and various vendors have different definitions of RAID levels. At present, the definition of RAID level can be widely recognized only four kinds of industry, RAID0, RAID1, RAID0+1 and RAID5.
RAID0 is a storage space stripe without data redundancy. It has a RAID level with low cost, high read/write performance, and high storage space utilization. It is suitable for very fast speed requirements such as Video/Audio signal storage and temporary file storage. Strict special applications. However, because there is no data redundancy, its security is greatly reduced, and any piece of hard disk that constitutes an array is damaged, causing catastrophic data loss. Therefore, configuring more than four hard disks in RAID 0 is unwise for general applications.
RAID1 is a complete mirroring of two hard disk data. It has good security, simple technology, convenient management, and good read/write performance. However, it cannot be expanded (a single hard disk capacity) and the data space is wasted. In a strict sense, it should not be called an "array."
RAID0+1 combines the characteristics of RAID0 and RAID1. The independent disks are configured as RAID0. The two complete RAID0 mirrors each other. It has excellent read and write performance and high security. However, the cost of constructing an array is high, and the utilization of data space is low. It cannot be called a cost-effective solution.
RAID5 is currently the most widely used RAID technology. Each piece of independent hard disk is stripped and divided, and the same stripe area is subjected to parity check (XOR operation). The parity data is evenly distributed on each hard disk. A RAID 5 array built with n hard disks can have n-1 hard disk capacity and very high storage space utilization (see Figure 6). Loss of data on any hard disk can be deduced from the verification data. The biggest difference between it and RAID3 is whether the parity data is evenly distributed to each hard disk. RAID5 has the advantages of data security, fast read and write speeds, high space utilization, and is widely used. However, the disadvantage of RAID5 is that the performance of the entire system is greatly reduced after a hard disk fails.
For RAID1, RAID0+1, and RAID5 arrays, hot-swappable (also called hot-replaceable) technology enables online data recovery. That is, when any hard disk in a RAID array is damaged, users do not need to shut down or stop the application service. It is of great significance to realize high availability (HA) systems by replacing failed hard disks, repairing systems, and recovering data.
Various vendors continue to introduce various RAID levels and standards. For example, higher security, starting from the RAID controller mirroring RAID; faster read and write speeds, configuring RAID for each disk and RAID CPU, and so on, but are not universal. The use of IDE hard drives to build RAID technology is a new emerging technology direction and has a greater impact on the market. Its outstanding advantage is that it is very cheap to build RAID arrays. At present, IDERAID can support three levels of RAID0, RAID1, and RAID0+1, and supports up to four IDE hard disks. Due to the limitation of the scalability of IDE devices, and because of the lack of hot and replaceable technical support for IDE devices, there are not many IDERAID applications.
In short, development is an eternal theme, and it is no exception in the area of ​​server storage technology. On the one hand, some giant manufacturers try to introduce new concepts or standards to lead the development direction of servers and storage technologies, such as the Intel-driven IA-64 architecture and storage concepts. On the other hand, they are dedicated to storage. Based on existing technologies and industry standards, professional vendors promote the rapid update and development of existing storage technologies and solutions based on SCSI, RAID, and Fibre Channel. Under market economy conditions, the only criterion for the inspection of technological development is market recognition. Markets call for good technology, and new technologies must be widely accepted and recognized when they must promote market development. With the development of the high-performance computer market, new high-performance, high-reliability, and high-security storage technologies will continue to emerge.
Currently, there are many disk array products on the market. Users must choose according to their own needs in the process of selecting disk array products. Now list several disk array products and provide users with options for disk array products. . Table 2 lists the main technical specifications of several disk arrays.
A sweetener is a Food Additive that confers a sweet taste in food or feed. Currently, there are many kinds of sweeteners, which can be divided into natural sweeteners and artificial sweeteners according to their sources. According to its nutritional value, it is divided into nutritive sweeteners and non-nutritive sweeteners; According to its chemical structure and nature, it is classified into sugar and non-carbohydrate sweeteners.
Sweetening Agent,Sweeteners Flavoring Agents,Natural Sweetening Agents,Food Sweetening Agent
SHANDONG ZHISHANG CHEMICAL CO.LTD , https://www.zhishangchemical.com