The more traffic your websites and applications generate, the higher the load your server will have to carry. Over the course of a given amount of time, your server may have a variation of high, medium, and low load. The average of that load over time is called the load average.
Programs like “top” allow you to get a glimpse of your system’s load average. The average may look something like this:
load average: 0.23, 0.18, 0.27
The above is an example of very low load averages. The “uptime” command will also display your system’s load average. Each number represents the average number of processes waiting to be scheduled on the run queue over the last minute.
In very intense situations, load averages can climb to 20 or higher. Imagine 5 lanes of traffic trying to squeeze onto a one-lane bridge, and you will have an idea of what a high load average can do to your system.
A high load average is not always an immediate cause for concern, especially if your CPUs themselves are not under a tremendous about of stress. One time when it should definitely be a concern is when you are not running anything requiring a lot of CPU attention either on your end or on any website. This could mean someone has exploited a vulnerability in your system and is running background processes.