There are many different ways to organize data in a RAID array. These ways are called "RAID levels". Different RAID levels have different speed and fault tolerance properties. RAID level 0 is not fault tolerant. Levels 1, 1E, 5, 50, 6, 60, and 1+0 are fault tolerant to a different degree - should one of the hard drives in the array fail, the data is still reconstructed on the fly and no access interruption occurs.

RAID levels 2, 3, and 4 are theoretically defined but not used in practice.

There are some more complex layouts: RAID 5E/5EE (integrating some spare space) and RAID DP but they are beyond the scope of this reference.

Use RAID0 when you need performance but the data is not important.

In a RAID0, the data is divided into blocks, and blocks are written to disks in turn.

RAID0 provides the most speed improvement, especially for write speed, because read and write requests are evenly distributed across all the disks in the array. Note that RAID1, Mirror, can provide the same improvement with reads but not writes. So if the request comes for, say, blocks 1, 2, and 3, each block is read from its own disk. Thus, the data is read three times faster than from a single disk.

However, RAID0 provides no fault tolerance at all. Should any of the disks in the array fail, the entire array fails and all the data is lost.

RAID0 solutions are cheap, and RAID0 uses all the disk capacity.

If RAID0 controller fails, you can do a RAID0 recovery relatively easy using RAID recovery software. However you should keep in mind that if the disk failure happens, data is lost irreversibly.

RAID1E is a mirror made over odd number of disks. With RAID1E you still get 50% overhead because each data block is stored on two mirror copies. Unlike RAID1, RAID1E uses the striping technique which gives you an increase in read speed even for degraded configurations.

With RAID1, you are supposed to use only 2 drives or maximum three (3-way mirror) because to have more than 3 copies of the same data is really costly in terms of disk space. RAID1E allows you to stick to the mirror configuration while having more than two disks in the set.

Use RAID1E when you need to get a reliable storage, surviving a single disk failure, made over odd number of disks.

Use mirroring when you need reliable storage of relatively small capacity.

Mirroring (RAID1) stores two identical copies of data on two hard drives. Should one of the drives fail, all the data can be read from the other drive. Mirroring does not use blocks and stripes.

Read speed can be improved in certain implementations, because read requests are sent to two drives in turn. Similar to RAID0, this should increase speed by the factor of two. However, not all implementations take advantage of this technique.

Write speed on RAID1 is the same as the write speed of a single disk, because all the copies of the data must be updated.

RAID1 uses the capacity of one of its drives to maintain fault tolearnce. This amounts to 50% capacity loss for the array. E.g. if you combine two 500GB drives in RAID1, you'd only get 500GB of usable disk space.

If RAID1 controller fails you do not need to recover neither array configuration nor data from it. To get data you should just connect any of the drives to the known-good computer.

RAID5 fits as large, reliable, relatively cheap storage.

RAID5 writes data blocks evenly to all the disks, in a pattern similar to RAID0. However, one additional "parity" block is written in each row. This additional parity, derived from all the data blocks in the row, provides redundancy. If one of the drives fails and thus one block in the row is unreadable, the contents of this block can be reconstructed using parity data together with all the remaining data blocks.

If all drives are OK, read requests are distributed evenly across drives, providing read speed similar to that of RAID0. For N disks in the array, RAID0 provides N times faster reads and RAID5 provides (N-1) times faster reads. If one of the drives has failed, the read speed degrades to that of a single drive, because all blocks in a row are required to serve the request.

Write speed of a RAID5 is limited by the parity updates. For each written block, its corresponding parity block has to be read, updated, and then written back. Thus, there is no significant write speed improvement on RAID5, if any at all.

The capacity of one member drive is used to maintain fault tolerance. E.g. if you have 10 drives 1TB each, the resulting RAID5 capacity would be 9TB.

If RAID5 controller fails, you can still recover data from the array with RAID 5 recovery software. Unlike RAID0, RAID5 is redundant and it can survive one member disk failure.

While the diagram on the right might seem simple enough, there is a variety of different layouts in practical use. Left/right and synchronous/asynchronous produce four possible combinations (see here for diagrams). Further complicating the issue, certain controllers implement delayed parity.

RAID50 consists of several RAID5 arrays combined into a RAID0. As discussed above, to create a RAID0 array, you need at least two disks, while for a RAID5 you need to provide three disks minimum. Given the numbers, we may conclude that for a RAID50 we need at least six disks.

With a RAID50, you can get the performance increase in terms of read speed by (N-1)*K times, where N is the number of disks in each RAID5 group and K is the number of RAID5 groups forming a RAID0.

As far as fault tolerance goes, RAID 50, like a regular RAID5, is guaranteed to survive a single disk failure. However, sometimes more disks may fail without data loss - this is possible if the failed disks are from different RAID5 groups.

In practice, the configurations like RAID50 are typically used with a large number of disks, and with several disks reserved as hot spare.

RAID6 is a large, highly reliable, relatively expensive storage.

RAID6 uses a block pattern similar to RAID5, but utilizes two different parity functions to derive two different parity blocks per row. If one of the drives fails, its contents are reconstructed using one set of parity data. If another drive fails before the array is recovered, the contents of the two missing drives are reconstructed by combining the remaining data and two sets of parity.

Read speed of the N-disk RAID6 is (N-2) times faster than the speed of a single drive, similar to RAID levels 0 and 5. If one or two drives fail in RAID6, the read speed degrades significantly because a reconstruction of missing blocks requires an entire row to be read.

There is no significant write speed improvement in RAID6 layout. RAID6 parity updates require even more processing than that in RAID5.

The capacity of two member drives is used to maintain fault tolerance. For an array of 10 drives 1TB each, the resulting RAID6 capacity would be 8TB.

The recovery of a RAID6 from a controller failure is fairly complicated. The main approaches to RAID6 data recovery in particular and data recovery in general are covered in data recovery book.

Actually, RAID60 is similar to a RAID50 with the only difference that instead of RAID5 arrays, RAID6 arrays are combined into a RAID0. As in RAID50, minimum of two sets of RAID6 arrays are required. Since a regular RAID6 requires at least 4 disks, for a “minimal” RAID60 you need at least 8 disks spread equally over two RAID6 groups.

With a RAID60 you get an increase in read speed by (N-2)*K times, where N is the number of disks in each RAID6 group and K is the number of RAID6 groups.

Talking about fault tolerance, RAID 60 array, like a regular RAID6 can survive two member disk failures without data loss. In practice, if you are lucky enough, RAID60 can lose 2*K disks – 2 disks from each RAID6 set.

Like with a RAID50, to ensure stable work you need to have enough disks reserved as hot spare.

RAID10 is a large, fast, reliable, but expensive storage.

RAID10 uses two identical RAID0 arrays to hold two identical copies of the content.

Read speed of the N-drive RAID10 array is N times faster than that of a single drive. Each drive can read its block of data independently, same as in RAID0 of N disks.

Writes are two times slower than reads, because both copies have to be updated. As far as writes are concerned, RAID10 of N disks is the same as RAID0 of N/2 disks.

Half the array capacity is used to maintain fault tolerance. In RAID10, the overhead increases with the number of disks, contrary to RAID levels 5 and 6, where the overhead is the same for any number of disks. This makes RAID10 the most expensive RAID type when scaled to large capacity.

If there is a controller failure in a RAID10, any subset of the drives forming a complete RAID0 can be recovered in the same way the RAID0 is recovered.

Similarly to RAID 5, several variations of the layout are possible in implementation. For more diagrams, refer here.