Fault Tolerance and Availability

Of course, RAID levels vary a great deal in terms of their ability to handle faults (usually meaning drive failures) and keep data availability high. The need for fault tolerance dictates the choice of RAID level probably more than any other requirement. In "boiled down" terms, the more redundancy in the array, the more fault tolerant it is and the higher the availability of its data, but the way the drives are connected is also very important. Each level has the particulars of its ability to deal with these important reliability-related issues discussed in detail in its own section. For a more thorough, more general look at these matters, check out the section on reliability issues.

Note: While RAID levels obviously vary in the degree to which they can tolerate faults, availability is often almost as much a function of implementation as it is design. High availability in particular is dependent upon the use of advanced features such as drive swapping and hot spares.

Next: Degradation and Rebuilding