Who are you choose (next) RAID level
Generally used RAID classes, RAID 0, RAID1, RAID 3, RAID 4, and RAID 5, plus two-type RAID 0 1 ﹝ or RAID 10 ﹞. Let's compare these RAID levels, shortcomings:
Raid level relatively advantageous relative shortcomings
RAID 0 Access speed is the fastest tolerance
RAID 1 is completely fault tolerant
RAID 3 is best not available with multiple task features
RAID 4 has multi-tasks and fault tolerant features Parity disk drives resulting in performance bottlenecks
RAID 5 has Overhead when writing multitasking and fault tolerance
RAID 0 1 / RAID 10 fast, complete fault-tolerant cost
Next, we discussed in-depth discussions for RAID 3, RAID 5, and RAID 0 1 / RAID 10.
RAID 3 Features and Applications
RAID 3 is to make the data first, after the Parity Data is generated, and the data and Parity Data are written to a member disk drive in parallel access mode, therefore has the advantages and disadvantages of parallel access mode. Further, each data transmission of RAID 3 updates the entire Stripe ﹝, that is, the data of each member disk drive is updated together, so there is no need to read some disk drive existing data, and new The data is xor operation, and the case will be written. This situation occurs in RAID 4 and RAID 5, generally called Read, Modify, Write Process, and we translate to read, change, write process. Therefore, in all RAID levels, the write performance of RAID 3 is best.
RAID 3 Parity Data is generally stored in a exclusive Parity Disk, but since each data is updated throughout Stripe, the Parity Disk of RAID 3 will not be accessed by the Parity Disk of RAID 4.
The parallel access mode of RAID 3 requires the support of the RAID controller special feature to reach the disk drive synchronous control, and the advantage of the above write performance, can be replaced with the current Caching technology, so it is generally considered RAID 3 Application will gradually fade out the market.
RAID 3 is especially suitable for applications such as large, continuous archives, such as drawings, images, video editing, multimedia, data warehousing, high-speed data capture, etc.
RAID 4 Features & Application
RAID 4 is to take independent access mode while stock Parity Data in a single exclusive Parity Disk. Each transmission of RAID 4 is a longer, and Overlapped I / O can be executed, so its performance is very good.
However, due to the use of a single exclusive Parity Disk to store Parity Data, it will cause a large bottleneck when writing. Therefore, RAID 4 is not widely used.
RAID 5 Features & Application
RAID 5 is also a stand-alone access mode, but its Parity Data is dispersed to the respective member disk drives, so in addition to the Overlapped I / O multi-task performance, it is also separated from RAID 4 single exclusive Parity Disk. Enter the bottleneck. However, RAI? D 5 is still slightly dragged by "reading, change, write process" in the seating data.
Since RAID 5 can perform Overlapped I / O multitasking, the more the number of member disk drives of RAID 5, the higher its performance, because one disk drive can only perform a Thread, the more disk drives, the more The more Threads that can be Overlapped, the higher the performance. But in turn, the more disk drives, the higher the chance of disk drive failures, the higher the array, or the MEAN TIME TO DATA LOSS is reduced. Since RAID 5 disperse Parity Data, it is characterized by the characteristics of XOR technology. For example, when several writes are required, these data to be written, and Parity Data may be dispersed in different member disk drives, so the RAID controller can make full use of Overlapped I / O while making several disk drives Work separately, and this, the overall performance of the array will increase much.
Basically, many people have more tasks, frequent access, data volume is not very large, suitable for the use of RAID 5 architectures, such as corporate file servers, web servers, online trading systems, e-commerce, etc. Small, access to frequent applications.
RAID 0 1 ﹝ RAID 10 ﹞
RAID 0 1 / RAID 10, combined with the advantages of RAID 0 and RAID 1, suitable for use in speed demand, and is completely faulty, of course, there are many applications. RAID 0 and RAID 1 are very simple, it is still very simple after combining, we don't plan to introduce, but to talk about it, RAID 0 1 should be RAID 0 over RAID 1, or RAID 1 over RAID 0, that is Say, do multiple RAID 1 into RAID 0, or make multiple RAID 0 into RAID 1?
RAID 0 over RAID 1
Suppose we have four disk drives, each two disk drives are made into RAID 1, then make two RAID 1 into RAID 0, this is RAID 0 over RAID 1:
(RAID 1) a = drive a1 drive a2 (mirrored)
(RAID 1) B = Drive B1 Drive B2 (mirrored)
RAID 0 = (RAID 1) A (RAID 1) B (Striped)
RAID 1 OVER RAID 0
Suppose we have six disk drives, each two disk drives are made into RAID 0, then make two RAID 0 into RAID 1, this is RAID 0 over RAID 1:
(RAID 0) A = Drive A1 Drive A2 (Striped)
(RAID 0) B = Drive B1 Drive B2 (Striped)
RAID 1 = (RAID 1) A (RAID 1) B (mirrored)
Under this architecture, if (raid 0) A has a disk drive failure, (raid 0) a is destroyed, of course, RAID 1 can still work properly; if this time (raid 0) B also has a disk drive failure , (Raid 0) B is also ruined. At this time, the two disk drives of RAID 1 are faulty, and the entire RAID 1 is destroyed.