UNIX monitoring tools

The simple performance tools on a Unix-derived system are straightforward to use, and it's easy to show examples of good and bad behavior—the best way to teach how those tools are useful for monitoring. Note that the general background here, and the examples of what to look for, are still relevant even on a Windows system. The underlying hardware, the way the operating systems work, and the resulting performance concepts are no different. There's a table in a later section of this chapter that shows how to translate between the Unix and Windows monitoring terminology.

Sample setup

The server used here is the same one described in the pgbench chapter. For these examples, initially a small...