Common RAID levels and features

Common RAID levels and features

Raid0: it is called a strip. It has no fault tolerance capability and a high failure rate, but has good read/write performance. The available disk space is N * disk (single disk capacity) and requires at least two hard disks to form a RAID 0.

Raid1 is called an image and has fault tolerance. Because data is split into chunks before being stored to a disk, the performance decreases slightly. The read performance is similar to that of raid0, and the performance is improved, the space usage is N * Disk/2. At least two hard disks are required to form RAID 1.

Raid4: Fault Tolerance and improved read/write performance. A maximum of one disk can be broken. The available space is (n-1) * disk. When a disk fails, the other two disks use the exclusive or operation to obtain and recover the data on the faulty disk. If there is no better disk in the case of a disk failure, the required data will be combined and output after calculation, at this time, the performance will be significantly reduced. At least three disk groups, raid 4, are required.

RAID5: Fault Tolerance and improved read/write performance. Each group of bad disks can use (n-1) * disks. To reduce the I/O pressure on the disk, the verification code is scattered on each disk. Other features are the same as raid4. At least three disk groups, raid5.

Raid10: Fault Tolerance and improved read/write performance. Each group can have a bad disk. The available space is N * Disk/2, and at least four disk groups are required for raid 10.

Raid01: Fault Tolerance and improved read/write performance. The available space is N * Disk/2. However, if a disk in each group fails, the entire raid array will fail.

Raid6: There are two verification disks, two disk failures are allowed, at least four disk groups raid6, each disk available space is (n-2) * disk.