What is Load Average and how is it calculated?

Load Average — is the average system load over specific time intervals (usually 1, 5, and 15 minutes). It shows the average number of processes that are either running on the CPU or waiting for it.

The calculation is done by counting processes in the "runnable" state and those waiting for I/O completion (uninterruptible sleep). These values are averaged using exponential moving average.

For example, if the Load Average is 2 on a system with 2 cores, it means the CPUs are fully loaded. If the value exceeds the number of cores, processes are waiting in the queue.

This metric helps to understand how loaded the system is and whether there are bottlenecks in task processing.

Load average — is a measure of the average system load over specific time intervals (usually 1, 5, and 15 minutes). It reflects the average number of processes that are either executing on the CPU or waiting in the queue to be executed (in a runnable or uninterruptible sleep state).

For example, a load average of 1 on a single-core system means the CPU is 100% loaded. If the load average exceeds the number of cores, it indicates processes are waiting in line, and the system is overloaded.

The uptime or top command in Linux displays these values, for example:

load average: 0.75, 1.20, 0.90

This means that over the last 1, 5, and 15 minutes, the average load was 0.75, 1.20, and 0.90 respectively.

It is important to understand that load average is not a percentage of CPU utilization, but an average number of processes in the queue for execution.