Catalogue
Monitoring System Load and Investigating Its Causes

Monitoring System Load and Investigating Its Causes

🌐 日本語で読む

Overview

A simple set of notes intended as an explanation for new graduates.

-Item- -Explain-
%user CPU usage in user space
%system CPU usage in kernel space
%iowait Percentage of time spent waiting for I/O
%idle Percentage of time the CPU is idle and not waiting for I/O

Investigating the cause from CPU-related graphs in Zabbix + Grafana on a given day

(1) %iowait is exceptionally high

  • %iowait high
  • %user low
  • %system low

Cause

Heavy swapping may be occurring.

Investigation steps

1. Check SwapIn & SwapOut

1
$ sar -W

2. System-wide memory usage

1
$ free

3. Sort by memory usage and identify the processes consuming memory

1
$ top
  • Shift+p: sort by CPU usage
  • Shift+m: sort by memory usage
The actual cause

Since it occurred periodically at the same time, I checked the cron configuration with crontab -l and found a batch job running that nobody knew about. sweat

(2) %user is exceptionally high

  • %iowait low
  • %user high
  • %system low

Cause

CPU usage is high.

Investigation steps

1. Sort by CPU usage in descending order to identify the process

1
$ top
  • Shift+p: sort by CPU usage

This is only a small part, but I hope it helps.
That’s all.

kenzo0107

kenzo0107