Vahid
August 6th, 2008, 16:19
Linux Server Load Averages Explained
New Linux users often get confused when talking about server load and the load averages. How to see the current load values on a Linux machines? SSH into your server and switch to the root user or an user account with admin privileges. Then just type "top" and hit enter to see the current processes, memory usage, and CPU related data - including the load averages. The term "load average" indicates a rough idea of the number of queued processes waiting for a resource to become available plus it shows active (currently running) processes. Here is a screenshot of a "top" command output:
In the upper part of the screenshot you can see the load averages of the server when the screenshot was taken. Here the load is very low and the server is just peacefully humming along. The 3 number values of the load average indicate the load average over 3 different periods (1 minute, 5 minutes, 15 minutes). Please note that the "load average" doesnt accurately depict how busy your system is in that very moment, or why your system is busy in general. A server can have 1 single process running, nothing queued, and still have 100% CPU usage, or you could be running 25 processes at the same time and the CPU load is at about 30%.
So, what the does the load average values really tell you? If the load average of your server is consistently high it means one or some of your resources is probably overloaded. You might have a bottleneck on your system that slows everything down. Not enough memory can often be such a bottleneck. A very generic idea of a good load average is <2.00. However it depends on many variables and server admins rather see this value at 1.00 or below.
New Linux users often get confused when talking about server load and the load averages. How to see the current load values on a Linux machines? SSH into your server and switch to the root user or an user account with admin privileges. Then just type "top" and hit enter to see the current processes, memory usage, and CPU related data - including the load averages. The term "load average" indicates a rough idea of the number of queued processes waiting for a resource to become available plus it shows active (currently running) processes. Here is a screenshot of a "top" command output:
In the upper part of the screenshot you can see the load averages of the server when the screenshot was taken. Here the load is very low and the server is just peacefully humming along. The 3 number values of the load average indicate the load average over 3 different periods (1 minute, 5 minutes, 15 minutes). Please note that the "load average" doesnt accurately depict how busy your system is in that very moment, or why your system is busy in general. A server can have 1 single process running, nothing queued, and still have 100% CPU usage, or you could be running 25 processes at the same time and the CPU load is at about 30%.
So, what the does the load average values really tell you? If the load average of your server is consistently high it means one or some of your resources is probably overloaded. You might have a bottleneck on your system that slows everything down. Not enough memory can often be such a bottleneck. A very generic idea of a good load average is <2.00. However it depends on many variables and server admins rather see this value at 1.00 or below.