《计算机组织与体系结构课件07》由会员分享，可在线阅读，更多相关《计算机组织与体系结构课件07（24页珍藏版）》请在金锄头文库上搜索。
1、RAIDTongwei RenSoftware Institute, Nanjing UniversityMar. 18, 2014Lecture 07c0864 Computer Organization and ArchitectureNANJING UNIVERSITYReview Error correction Parity checking Hamming code Cyclic redundancy check23NANJING UNIVERSITYRAID Redundant Array of Independent Disks Basic idea With multiple
2、 disks, separate I/O requests, even a single I/O request, can be handled in parallel, as long as the data required reside on separate disks Characteristic RAID is a set of physical disk drives viewed as a single logical drive Data are distributed across the physical drives Redundant disk capacity is
3、 used to store parity information, which guarantees data recoverability in case of a disk failure4NANJING UNIVERSITYRAID Level5NANJING UNIVERSITYRAID Level 0 The data are striped across the available disks Not include redundancy to improve performance (not a true member of the RAID family) Usage Hig
4、h rate data transfer High I/O request rate6NANJING UNIVERSITYRAID Level 0 (cont.)7 Compared to a single large disk If two different I/O requests are pending for two different blocks of data, then there is a good chance that the requested blocks are on different disksNANJING UNIVERSITYRAID Level 1 Re
5、dundancy is achieved by the simple expedient of duplicating all the data Data striping is used8NANJING UNIVERSITYRAID Level 1 (cont.) Advantage A read request can be serviced by either of the two disks that contains the requested data, whichever one involves the minimum seek time plus rotational lat
6、ency A write request requires that both corresponding strips be updated, but this can be done in parallel Recovery from a failure is simple Disadvantage High cost9NANJING UNIVERSITYRAID Level 1 (cont.) Usage Limited to drives that store system software and data and other highly critical files Compar
7、ed to RAID 0 RAID 1 can achieve high I/O request rates if the bulk of the requests are reads, in which the performance of RAID 1 can approach double of that of RAID 0 If a substantial fraction of the I/O requests are write requests, then there may be no significant performance gain over RAID 010NANJ
8、ING UNIVERSITYRAID Level 2 Use of a parallel access technique All member disks participate in the execution of every I/O request the spindles of the individual drives are synchronized so that each disk head is in the same position on each disk at any given time Data striping is used The strips are v
9、ery small, often a single byte or word11NANJING UNIVERSITYRAID Level 2 (cont.) An error-correcting code is calculated across corresponding bits on each data disk, and the bits of the code are stored in the corresponding bit positions on multiple parity disks Hamming code is typically used Read The r
10、equested data and the error-correcting code are fetched Write All data disks and parity disks must be accessed12NANJING UNIVERSITYRAID Level 2 (cont.) Disadvantage Still rather costly Only be an effective choice in an environment in which many disk errors occur, which is overkill and is not implemen
11、ted to individual disks and disk drives with the high reliability13NANJING UNIVERSITYRAID Level 3 Use of a parallel access technique Use very small data striping A simple parity bit is computed for the set of individual bits in the same position on all of the data disks Can be used to reconstruct da
12、ta when a drive fails0 = ()1 2314NANJING UNIVERSITYRAID Level 3 (cont.) Performance Achieve very high data transfer rates For large transfers, the performance improvement is especially noticeable Only one I/O request can be executed at a time In a transaction-oriented environment, performance suffer
13、s15NANJING UNIVERSITYRAID Level 4 Use of an independent access technique Each member disk operates independently, so that separate I/O requests can be satisfied in parallel Data striping is used A bit-by-bit parity strip is calculated across corresponding strips on each data disk, and the parity bit
14、s are stored in the corresponding strip on the parity disk16NANJING UNIVERSITYRAID Level 4 (cont.) Performance RAID 4 involves a write penalty when an I/O write request of small size is performed Each time that a write occurs, the array management software must update not only the user data but also
15、 the corresponding parity bits = 0 0 In the case of a larger size I/O write that involves strips on all disk drives, parity is easily computed by calculation using only the new data bits Every write operation must involve the parity disk, which therefore can become a bottleneck17NANJING UNIVERSITYRA
16、ID Level 5 Similar to RAID level 4 Distributes the parity strips across all disks Avoid the potential I/O bottleneck18NANJING UNIVERSITYRAID Level 6 Two different parity calculations are carried out and stored in separate blocks on different disks Advantage Provide extremely high data availability: three disks w
