Value of item system.cpu.load[percpu,avg1] > 2.4

[wyang]

Hello,

From some Internet resources on understanding Linux CPU load, "the '100% utilization' mark is 1.00 on a single-core system." Therefore, I expect that the max value of the item system.cpu.load[percpu,avg1] <=1.

However, monitoring history shows that the maximum values of the item system.cpu.load[percpu,avg1] on two servers display > 2.4

Any idea on why? Thanks very much!

[Atsushi]

That understanding is incorrect.

The load average shows how many processes the server can run at that time.
As the load increases, the number of processes waiting to be processed increases,
so it will frequently happen more than 1 frequently.

see.
https://en.wikipedia.org/wiki/Load_(computing)

[wyang]

Thanks very much for the reply. It really helps.

Zabbix OS Linux template has a trigger
{Template OS Linux:system.cpu.load[percpu,avg1].avg(5m)}>5

How is this corresponding to % of CPU utilization?

Is there a way to evaluate % of CPU utilization so that triggers can be defined as 'CPU utilization on {HOST.NAME} > (75|90)%'?

[Atsushi]

The load average and the CPU utilization are not linked.

For example, even if 100% of the CPU is used up, if the number of processes waiting to be processed increases, the load average value may continue to increase further.