# Why Is My Linux CPU at 100%? (Problem → Solution for Beginners)

CPU usage stuck at 100% is one of the most stressful Linux problems. Suddenly:

• Your SSH session lags

• Pages load slowly

• Commands take forever

• Your server feels frozen

Relax. CPU spikes usually come from something simple: a process looping, a web server overload, a runaway script, or a misconfiguration.

Let’s fix it using easy steps anyone can follow.

---

## Step 1 — Identify What Process Is Eating the CPU

Use top

Run:

top

Look at the column %CPU.

If you see something like:

1234 php-fpm   250% 

or

mysqld   180%

That’s your culprit.

Use htop (if installed)

Much easier to read:

htop

Press:

• F6 → sort by CPU

• F3 → search a process

If you don’t have it:

sudo apt install htop -y     # Debian/Ubuntu
sudo yum install htop -y     # CentOS/RHEL

---

## Step 2 — Check If a Single Script Is Stuck in a Loop

This is common for PHP, Python, Node, or cron jobs.

Example: PHP script using 100% CPU

If top shows:

php-fpm  100%
php-fpm  99%
php-fpm  98%

A script might be looping.

Check logs:

tail -f /var/log/php*/error.log

Common causes:

• Infinite loops

• Bad SQL queries

• Bots hammering your site

• Missing index causing slow DB queries

• 
  

---

## Step 3 — Check for High MySQL CPU Usage

If you see:

mysqld  180%

Then MySQL is the problem.

Check slow queries:

mysqladmin processlist

Or:

SHOW PROCESSLIST;

Look for:

• “Locked”

• “Copying to tmp table”

• Very long “Query”

• Too many “Sleep”

Fixing slow MySQL queries often instantly fixes CPU.

---

## Step 4 — Check If Your Server Is Under Attack

A spike of CPU might be caused by:

• Bot traffic

• DDoS

• Brute-force login attempts

• Crawlers overloading your site

• A sudden increase of visitors

Check number of active connections:

ss -tuna

Check connections to your web server:

netstat -an | grep :80 | wc -l

Check if one IP is abusing you:

netstat -ntu | awk '{print $5}' | cut -d: -f1 | sort | uniq -c | sort -n

If an IP shows thousands of connections → block it:

sudo ufw deny from <IP>

or using firewall-cmd:

sudo firewall-cmd --permanent --add-rich-rule='rule family="ipv4" source address="<IP>" reject'
sudo firewall-cmd --reload

---

## Step 5 — Check Background Jobs (Cron Jobs)

Sometimes a cron job runs too often or gets stuck.

List cron jobs:

crontab -l

Check running jobs:

ps aux | grep cron

Look in cron logs:

grep CRON /var/log/syslog

Examples of CPU-killer cron jobs:

• Backup script that never finishes

• Log rotation loop

• Database dump running every minute

• Script running thousands of times

• 
  

---

## Step 6 — Check If a System Service Is Misbehaving

Run:

systemctl status

Look for services constantly restarting.

Examples:

• Apache restarting repeatedly

• Node application crashing and respawning

• Docker containers looping

To restart a suspicious service:

sudo systemctl restart service-name

---

## Step 7 — Check Disk Full (Yes, It Can Cause 100% CPU)

Run:

df -h

If root partition is 100% full, many processes break and CPU spikes.

Clear logs:

sudo journalctl --vacuum-size=100M

Clear apt cache:

sudo apt clean

Remove big log files:

sudo truncate -s 0 /var/log/*.log

---

## Step 8 — Check Load Average

In top, look at this line:

load average: 8.32, 7.10, 6.80

If your server has:

• 1 CPU core → load > 1 = overloaded

• 2 CPU cores → load > 2 = overloaded

• 4 cores → load > 4 = overloaded

Check CPU cores:

nproc

Small servers (1 vCPU) easily hit 100% usage.

---

## Step 9 — Restart the Offending Service (Temporary Fix)

If CPU is stuck at 100% and you want quick breathing room:

Examples:

sudo systemctl restart php7.4-fpm
sudo systemctl restart apache2
sudo systemctl restart nginx
sudo systemctl restart mysql
sudo systemctl restart docker

But remember:
A restart is temporary — not a real fix.

---

## Step 10 — Final Fixes Based on What You Found

If MySQL is high CPU

→ Optimize queries, add indexes, increase buffer pools.

If PHP or Python is high CPU

→ Fix infinite loops or errors in app code.

If Apache/Nginx is overloaded

→ Reduce worker processes
→ Add rate limiting
→ Use caching (Redis, FastCGI cache)

If server is under attack

→ Block IPs
→ Enable rate limiting
→ Install fail2ban

If CPU is too small

→ Upgrade VPS (from 1 vCPU → 2 vCPU)

---

## Beginner-Friendly Checklist

---

## Conclusion

A Linux server at 100% CPU can look scary, but the root cause is almost always something simple:

• A script looping

• A slow MySQL query

• Web traffic spike

• Cron job running too often

• Misbehaving service

• Server under attack

With the clear step-by-step process in this tutorial, even beginners can confidently diagnose and fix CPU overload problems.