CPU Monitoring — Monitoring Processor Performance in Linux¶

CPU Monitoring is the process of tracking processor utilization, system load, CPU wait times, process execution, and overall processor performance. High CPU utilization can lead to slow applications, increased response times, and poor system performance. By monitoring CPU usage, administrators can identify resource-intensive processes, detect bottlenecks, and optimize workloads before they impact production systems. Every Linux administrator, DevOps engineer, Cloud Architect, Platform Engineer, and Site Reliability Engineer (SRE) should understand how to monitor CPU performance in Linux.

Learning Path¶

Linux Mastery → Module 12: Monitoring & Logs → Lesson 7

Difficulty: Beginner → Intermediate

Reading Time: 95 Minutes

Course Progress

Course: Linux Mastery

Module: Monitoring & Logs

Lesson: 7 of 10

What You'll Learn¶

After completing this lesson, you'll be able to:

Understand CPU utilization
Interpret load average
Monitor processor performance
Identify CPU-intensive processes
Analyze CPU bottlenecks
Monitor CPU statistics
Troubleshoot CPU performance issues
Apply production monitoring best practices

Prerequisites¶

Complete:

Modules 1–11
Module 12 Lessons 1–6

Why Learn CPU Monitoring?¶

Imagine a production API server.

Without monitoring:

CPU Usage Reaches 100%

↓

Application Response Time Increases

↓

Users Experience Timeouts

With monitoring:

CPU Usage

↓

Alert at High Utilization

↓

Identify Problem Process

↓

Optimize

↓

Stable Performance

Monitoring CPU usage helps prevent performance degradation.

What is CPU Monitoring?¶

CPU monitoring involves observing:

CPU utilization
System load
CPU idle time
User and system CPU time
I/O wait
Process CPU usage
CPU bottlenecks

CPU Performance Overview¶

Processes

↓

CPU Scheduler

↓

Processor

↓

Execution

↓

Performance Metrics

Linux continuously schedules processes across available CPU cores.

CPU States¶

CPU time is divided into several categories.

State	Description
User (`us`)	Time spent running user processes
System (`sy`)	Time spent running kernel code
Idle (`id`)	CPU not performing work
I/O Wait (`wa`)	Waiting for storage operations
Nice (`ni`)	Time spent on lower-priority processes
IRQ (`hi`)	Hardware interrupt handling
SoftIRQ (`si`)	Software interrupt handling
Steal (`st`)	Time taken by the hypervisor (virtual machines)

Monitor CPU with top¶

Launch:

top

Example:

%Cpu(s):

us

sy

id

wa

Press:

Shift + P

Sort processes by CPU usage.

Monitor CPU with htop¶

htop

Features:

Per-core CPU graphs
Interactive process management
Colorized display
Easy sorting

Check System Load¶

Use:

uptime

Example:

load average:

0.45

0.60

0.70

The three values represent the average system load over:

1 minute
5 minutes
15 minutes

Understanding Load Average¶

Example:

4 CPU Cores

↓

Load Average

4.00

↓

System Fully Utilized

General guideline:

CPU Cores	Healthy Load
2	Around 2
4	Around 4
8	Around 8
16	Around 16

A sustained load average significantly higher than the number of CPU cores may indicate CPU contention or processes waiting for CPU time.

CPU Statistics¶

Use:

vmstat

Example:

vmstat 2

Useful CPU columns:

us

sy

id

wa

Detailed CPU Statistics¶

Install sysstat if necessary.

mpstat

Per-core statistics.

mpstat -P ALL

Displays utilization for every CPU core.

CPU Usage by Process¶

Display the highest CPU consumers.

ps aux --sort=-%cpu

Top 10:

ps aux --sort=-%cpu | head

View CPU Information¶

Display processor details.

lscpu

Example output:

Architecture

CPU(s)

Model Name

Thread(s)

Core(s)

Number of CPUs¶

nproc

Example:

CPU Frequency¶

cat /proc/cpuinfo

Search frequency.

grep "cpu MHz" /proc/cpuinfo

Real-Time CPU Monitoring¶

Refresh every second.

watch -n 1 grep "cpu " /proc/stat

CPU Information from proc¶

Detailed CPU statistics.

cat /proc/stat

Useful for low-level monitoring.

Common Commands¶

Monitor CPU.

top

System load.

uptime

CPU statistics.

vmstat

Per-core monitoring.

mpstat -P ALL

CPU information.

lscpu

Real Production Examples¶

Monitor CPU usage.

top

Check system load.

uptime

View per-core usage.

mpstat -P ALL

Find CPU-intensive processes.

ps aux --sort=-%cpu | head

Production Perspective¶

CPU monitoring is essential for:

Kubernetes worker nodes
Database servers
Java applications
API servers
Web servers
CI/CD servers
Virtual machines
Cloud infrastructure

High CPU utilization often indicates application inefficiencies, increased workload, or insufficient resources.

Hands-on Lab¶

Task 1¶

Open top.

top

Press:

Shift + P

Sort by CPU usage.

Task 2¶

Display system load.

uptime

Task 3¶

Monitor CPU statistics.

vmstat 2

Task 4¶

Display processor information.

lscpu

Task 5¶

Display the number of CPU cores.

nproc

Task 6¶

Monitor per-core utilization.

mpstat -P ALL

Task 7¶

Display the top CPU-consuming processes.

ps aux --sort=-%cpu | head

Task 8¶

Review /proc/stat.

cat /proc/stat

Command Deep Dive¶

Command	Purpose	Production Example
`top`	Real-time CPU monitoring	Performance analysis
`uptime`	Display system load	Capacity planning
`vmstat`	CPU statistics	Resource monitoring
`mpstat -P ALL`	Per-core CPU utilization	Multi-core analysis
`lscpu`	CPU hardware information	Infrastructure inventory
`ps aux --sort=-%cpu`	Top CPU consumers	Process troubleshooting

Common CPU Monitoring Mistakes¶

Mistake	Solution
Looking only at CPU percentage	Monitor load average as well
Ignoring I/O wait (`wa`)	High wait time may indicate storage bottlenecks
Assuming 100% CPU always indicates a problem	Verify workload characteristics first
Never checking per-core utilization	Monitor all CPU cores
Ignoring long-term CPU trends	Use historical monitoring tools

Production Troubleshooting Scenario¶

Scenario

Users report that an API service is responding slowly.

Investigation:

top

Result:

CPU

↓

95%

↓

Java Process

ps aux --sort=-%cpu | head

The Java application is consuming nearly all available CPU resources.

Further investigation reveals an inefficient query causing excessive processing.

After optimizing the application:

CPU Usage

↓

Normal

↓

Response Time Improved

Best Practices¶

Monitor CPU utilization continuously.
Track system load average.
Monitor per-core CPU usage.
Investigate sustained high CPU utilization.
Monitor I/O wait separately from CPU usage.
Configure CPU alerts before systems become overloaded.
Collect historical CPU metrics using monitoring platforms such as Prometheus and Grafana.
Correlate CPU metrics with memory, disk, and application performance.

Common Mistakes¶

❌ Monitoring only total CPU utilization.

✅ Avoid this mistake: monitoring only total CPU utilization.

❌ Ignoring load average.

✅ Always review load average.

❌ Confusing CPU utilization with I/O wait.

✅ Distinguish clearly between CPU utilization with I/O wait.

❌ Never identifying CPU-intensive processes.

✅ Always identifying CPU-intensive processes.

❌ Ignoring gradual increases in processor utilization.

✅ Always review gradual increases in processor utilization.

Interview Questions¶

Beginner¶

What does the top command display?
What is CPU utilization?
What does the uptime command show?
Which command displays CPU hardware information?

Intermediate¶

What is load average?
What does high I/O wait indicate?
How do you identify CPU-intensive processes?
Why is per-core monitoring important?

Architect Level¶

How would you monitor CPU utilization across hundreds of Linux servers?
How would you troubleshoot a Kubernetes node with consistently high CPU usage?
How would you determine whether a performance issue is caused by CPU, memory, or storage?

Summary¶

In this lesson, you learned:

CPU monitoring fundamentals
CPU utilization
Load average
Process CPU analysis
Per-core monitoring
CPU statistics
Performance troubleshooting
Production monitoring best practices

CPU monitoring is essential for maintaining high-performance Linux systems. By tracking processor utilization, load averages, CPU-intensive processes, and per-core performance, administrators can detect bottlenecks early, optimize workloads, and ensure reliable operation of production applications.

Key Takeaways¶

Use top and htop for real-time CPU monitoring.
Monitor system load using uptime.
Analyze CPU statistics with vmstat and mpstat.
Identify CPU-intensive processes using ps.
Investigate sustained high CPU utilization and I/O wait.
Combine CPU monitoring with memory, disk, and application metrics for comprehensive performance analysis.

What's Next?¶

Performance Troubleshooting — Diagnosing and Resolving Linux Performance Issues

You'll explore:

Identifying performance bottlenecks
CPU, memory, disk, and network analysis
System performance methodology
Troubleshooting tools
Root cause analysis
Performance optimization
Production troubleshooting best practices

By the end of the lesson, you'll be able to systematically diagnose Linux performance issues, identify bottlenecks, determine root causes, and optimize production systems for reliability and efficiency.