Load average difference

[imarinkovic]

Hi friends...

today I was facing a problem with the performance of a web server. Looking zabbix information, load average was "perfectly normal"...but when I enter the server, I noted the load average given by "top" was too different!

Load Average with Zabbix:
1 min: 0.13 5 min: 0.15 15 min: 0.13

Load Average with top:
load average: 2.33, 2.41, 2.15

- Red Hat Enterprise Linux Server release 6.4 (Santiago)
- zabbix_agentd (daemon) (Zabbix) 3.0.8

[imarinkovic]

Example

[jan.garaj]

Template OS Linux uses normalized CPU load (percpu), while top shows total load. Normalized load = total load divided by online CPU count.

Normalized load 0.15 doesn't look like a problem.

[imarinkovic]

Seems ok to me, but I have to explain that

Great, thanks for your answer!

Template OS Linux uses normalized CPU load (percpu), while top shows total load. Normalized load = total load divided by online CPU count.

Normalized load 0.15 doesn't look like a problem.