Let's begin by looking at the various elements that can significantly affect database performance. Throughput best encapsulates how we think of database performance. A simple view of a database is a file residing on a hard drive containing information that we would like to read or modify. In a computer system, long-term storage, the hard drive, or disc is the slowest part, so improvements made to throughput speed from the storage system to the CPU will deliver the most significant performance gains.

Generally speaking, database servers aren't computationally intensive. Their primary role is to write data to disk and read it from disk, sending it to the appropriate client application. It turns out that much of a database server's CPU time is spent waiting for data to be retrieved from the disk. 

Managing the multiple requests to read and write data from and to disk is another critical aspect of the database server tuning. With data throughput and multiple requests management in mind, we'll look at storage and file configuration, along with resource configuration. When workloads and storage requirements outstrip the current server setup, we'll look at ways to upgrade as and when required.