There is a reminder about RAID array levels below the calculator. For more info - check this.
*RAID 0 - (also known as a stripe set or striped volume) splits ("stripes") data evenly across two or more disks, without parity information, redundancy, or fault tolerance.
*RAID 1 - consists of an exact copy (or mirror) of a set of data on two or more disks; a classic RAID 1 mirrored pair contains two disks. With software support, it's possible to increase read performance as data may be read from any of the drives. Data is saved as long as at least one drive of the array is working.
*RAID 2 - stripes data at the bit (rather than block) level, and uses a Hamming code for error correction. Drives are divided into two groups: one for data and another is for the error correction. The number of drives in array is . The data is distributed among the data drives as in RAID 0, but in the case of any drive fault, it's possible to restore the data on the fly. This array requires a lot of drives for optimal use and was rarely used in practice.
*RAID 3 - uses byte-level striping. i.e Instead of striping the blocks across the disks, it stripes the bytes across the disks. In an array of n drives, the data is split into pieces smaller than a sector (split into bytes or blocks) and distributed to n-1 drives.Another drive is used to store parity blocks. Unlike RAID 2 it does not allow to recover errors on the fly. It provides high-performance only when single-tasking with large files due to the need to synchronize spindle drives. Also, there is a high load on a control drive in this array level. It wasn't widespread and was displaced by RAID 5.
*RAID 4 - an attempt to fix RAID 3 limitations. RAID 4 is similar to RAID 3, but uses block-level striping, thus have an increased read and write performances for small transfers. Although, write performance is slow due to the need to write all parity data to a single disk. It wasn't widespread and was displaced by RAID 5.
*RAID 5 - unlike RAID 4, parity information is distributed among the drives, requiring all drives but one to be present to operate. Thus, this array starts to operate in RAID 0 mode. You should also consider, that the RAID Reconstruction process (RAID data reconstruction based on redundancy) after the drive failure causes an intensive drive read stress for many continuous hours, thus, it can cause the failure of any remaining drives in this least safe RAID work period and can reveal previously undetected read failures in cold data arrays (data, which isn't referred to during the normal operations, like archive data) which increases the risk of failure during the data recovery.
*RAID 5E/5EE - non-standard RAID level (with the added E standing for Enhanced), generally refer to variants of RAID 5 or 6 with an integrated hot-spare drive, where the spare drive is an active part of the block rotation scheme. The difference between EE and E is in the method of space for hot-spare drive allocation, which allows higher data reading speed.
*RAID 6 — is similar to RAID 5 as consists of block-level striping but with double distributed parity. Double parity and Reed–Solomon error correction provides fault tolerance up to two failed drives.
In addition to standard RAID 0 - RAID 6 levels, described in «Common RAID Disk Drive Format (DEF) standard», there are hybrid levels with the titles like «RAID α+β» or «RAID αβ», which usually means «RAID β made out of several RAID α». Hybrid levels inherit both advantages and disadvantages of their "parents": striping in RAID 5+0 won't add parity, but it has a positive performance impact. RAID 1+5 is probably very reliable, but not very fast, and besides, it is uneconomic - usable space is less than the half of the total drive capacities.
*RAID 10 (or 1+0) — is a RAID 0 made out of several RAID 1 (mirrored pairs). RAID 10 array will be disabled only after all drives in the same RAID 1 array failed, in contrast to, for example, RAID 0+1. Thus, RAID 10 combines high parity and performance and it is one of the most common types of hybrid levels.