This article discusses how to view and interpret system load averages.
If you have a reseller hosting account, WebHost Manager prominently displays system load average information at the top of the page:
cPanel does not display load average information. However, you can view load averages at any time on a server by typing the following command at the command prompt:
Here is sample output from the uptime command:
username@a2server [~]# uptime 21:19:12 up 388 days, 17:32, 0 users, load average: 5.85, 5.98, 6.18
So what do the numbers in load averages represent? Let's look again at the sample output from the uptime command in the previous section:
load average: 5.85, 5.98, 6.18
The three numbers are, from left to right, the system load averages for the past one, five, and fifteen minutes. In this case, system load has been slightly decreasing over the past 15 minutes.
The numbers themselves represent the number of processes waiting to use the CPU (or multiple CPU cores) on the computer. A completely idle computer, for example, would have a load average of 0. A computer with one CPU core that has a single process running at 100% all of the time would have a load average of 1.0. Similarly, on a computer with four CPU cores, a load average of 4.0 indicates the CPUs are being utilized 100%.
Most newer computers (including A2 Hosting's servers) have multiple CPU cores, so a seemingly “high” load average number is not necessarily a cause for concern. For example, consider the load average numbers of 5.85, 5.98, and 6.18 listed above. On a computer with a single-core CPU, these numbers would indicate severe system overloading, with the CPU running well over 100% of capacity. But on a computer with eight or more CPU cores, these numbers would be completely normal.
Therefore, to accurately interpret the load averages for your account, you need to know how many CPU cores are on the system. To do this, log in to your account using SSH, and then type the following command:
The nproc command displays the total number of CPU cores on the system, allowing you to interpret system load averages correctly.