Do not confuse run queue length with the load average. The load average is available from v$osstat,4 the sar -q, and uptime commands, as well as in the /proc/loadavg file. The load average statistic is calculated differently on various flavors of Unix and Linux. On your system, it may represent the run queue length, but only your experience or analyzing the actual kernel source code will provide an acceptable answer. While the CPU run queue and utilization analyses are consistent across platforms, the load average is not.

Figure 4-14. A sar command providing average run queue information. This database server consists of four CPU cores, so processes are frequently waiting for CPU resources.

Compared to CPU, IO, and network analysis, memory bottleneck analysis is simple. The challenge is understanding the unique words used and not used on your particular platform. In fact, words like bottleneck and swapping can invoke a hostile and near-violent response from many operating system administrators. And don't even think about saying the word swap around an operating system vendor. It's just not worth the abuse you'll receive.