The reason why there are so many different RAID levels is that there are many different
ways to configure a bunch of hard disks, and many different needs of RAID users.
Distinguishing between different levels isn't easy at times, because many RAID levels are
similar to others in various ways. Worse, sometimes the differences between levels seem
subtle, but these small discrepancies can have a huge impact on the characteristics of the
array and the applications that make sense for it.
To completely and accurately portray each single and multiple RAID level, I describe each in terms of its most important
characteristics, including those related to fault tolerance, capacity, performance, cost
and other attributes. To avoid duplication, I have provided this section that describes
what each of these technical factors or attributes are about. For each one I explain
briefly what the attribute means, how it is defined, and how and why it helps
differentiate between various RAID levels. In the last sub-section, I also discuss the
reasons why some implementers may wish to consider creating more than one array for a
system if the needs of that system can't be met by one array type.
Note that in addition to the descriptions in this section, you may also want to take a
look at the more general discussions of performance and
reliability issues in the
section covering general concepts and issues. Some of the pages that follow in this
area will refer back to those or other pages in the site's coverage of RAID.
Next: Redundancy Technique: Mirroring vs. Parity
