Know the RAID disk array
In recent years, hard disk has been greatly improved in terms of capacity, access speed, and reliability, this increase is not to keep up with the development requirements of the processor, so that the hard disk will still become a bottleneck in the computer system. In order to solve the application system's high-speed access to disk, people have taken a variety of measures. In 1988, the inexpensive redundant disk array proposed by Professor D.A.Patterson, the University of California, Berkeley, is one of them.
RAID constitutes a normal hard disk into a disk array, and the RAID controller is decomposed into multiple data blocks to write data to be written, and then write into disk arrays; when the host reads the data, the RAID controller is read. Take data on each hard disk in the disk array, and then re-combined it to the host. The accessibility of the storage system is improved due to parallel read writing operations. In addition, RAID disk arrays can also use mirroring, parity, etc. to improve system fault tolerance and ensure the reliability of data.
Different from the method used in RAID, it can be divided into 0-5 six levels:
RAID 0: When the host requires writing data, the RAID controller divides the data into a plurality of blocks, and writes them on each hard disk in the disk array; when reading data, the RAID controller reads data from each hard disk. Restore these data to the main sequence. The advantage of this method is that the data block, the parallel transfer method can improve the host read / write speed, and there is no redundancy in the disk array. However, it does not increase anything about the reliability of the system, and the system cannot recover when any hard disk medium fails.
RAID 1: It divides the hard disk in the disk array into the same two groups, mirroring, when removing the disk medium, can use the data on its mirror, thereby increasing the fault tolerance capability of the system. The operation of the data is still used in parallel transmission mode after blocking. All RAID
1 Not only improves the read and write speed, but also strengthen the reliability of the system. However, its disadvantage is that the hard disk is low, and the redundancy is 50%.
RAID 2 / RAID 4: These two levels are rarely applied in practice, and most systems are not supported, and they are not introduced here.
RAID 3: Like RAID 0, RAID 3 also uses a method of data block parallel, but all of which calculates their parity after data block, and then blocks the block data and parity information. And write into the hard disk array. This method has improved the access speed and reliability of the data. When either hard drive in the array is corrupted, the information on other data discs and parity discs can be reconstructed. RAID in terms of hard disk utilization
3 is higher than RAID 1, such as an array consisting of 5 hard drives, and redundancy is only 20%. However, RAID 3 also has a disadvantage, because the parity information is fixed to a hard disk, resulting in a new bottleneck.
RAID 5: Similar to the data processing method used in RAID 3, it is different that it crosss the parity information to each of the hard disks in the array, thereby overcoming the bottleneck problem in RAID 3.
In addition to the above, we can also use software technology to implement RAID disk arrays. The disk space for Windows NT operating system, disk space with parity, disk mirror, and duplex, etc., is actually the software implementation of RAID technology. Among them, disk spaces correspond to RAID.
0, disk mirroring and duplex correspond to RAID 1, with parity tabs with parity corresponding to RAID 5. The biggest advantage of these methods is cheap, but the performance is much lower than the performance.
In the above levels, the advantages of RAID 5 are the most prominent, and for large important application environments, they can be used as preferred. However, for most pursuit of DIY enthusiasts, the RAID 0 disk array consisting of two IDE hard drives is more appropriate. At present, Ide RAID cards have been sold on the market, more common is the FastTrak 66 of Promise, with two UltraDMA66 IDE channels, support RAID 0, 1, 1 0, a price of more than 10,000 yuan. Let's take the FastTrak66 as an example to describe the installation process of the IDE RAID:
· Plug FastTrak 66 on the PCI slot of the motherboard;
• Two pieces of hard drives are used as the MASTER disk directly on the two IDE interfaces of FastTrak 66;
· Power on, press Ctrl F to enter the FastTrak 66 setting screen;
· Set the RAID MODE to Stripe, save and exit;
· Partition, formatted ... That's OK!
