Over the years, people designed and built servers and paid more money for server storage. These physical servers used local direct DAS, with the infrequent raid into two node clusters, with limited performance. When businesses were in need of greater performance, they scaled out with multiple servers, or some businesses bought expensive, dedicated Fiber Channel SAN with influential array technologies. This has changed a lot with the advent of virtualization; storage is now much more than just its own capacity. Arguably, the number of terabytes that our new storage array can provide is a relatively minor interest when we are investigating requirements.

Now let's explore the factors that need to be considered, such as hard disk, performance over capacity, IOPS, and spindle considerations.