Monitoring Server Resources with htop and top

Introduction

When a VPS or server becomes slow, the first question is usually: what is using the resources right now? Tools like top and htop help answer that question by showing CPU usage, memory usage, load average, running processes and high-resource tasks in real time.

These tools are useful during live troubleshooting, but they are not a full monitoring system by themselves. They show what is happening now. For long-term reliability, you should also use uptime monitoring, logs, alerts and historical resource metrics.

Quick answer

top and htop

top and htop are command-line tools for viewing live server resource usage.

Both tools help you see:

• current CPU usage
• memory usage
• swap usage
• load average
• running processes
• process IDs
• users running processes
• commands consuming resources
• stuck or long-running tasks

What to check first

When opening top or htop, start with the main server health indicators.

Check:

• load average
• CPU usage
• memory usage
• swap usage
• top processes
• process user
• command name
• process age
• disk activity if available
• number of running tasks
• zombie processes

Load average

Load average shows how many processes are running or waiting for CPU or I/O. It is usually shown for 1, 5 and 15 minutes.

load average: 2.50, 1.80, 1.20

Meaning:

• 1-minute average: recent load
• 5-minute average: medium trend
• 15-minute average: longer trend

Example on a 2-core VPS:

• load 1.0 may be fine
• load 2.0 may be near full use
• load 6.0 may indicate overload or waiting tasks

CPU usage

CPU usage shows how busy the processor is.

Common signs:

• one process using very high CPU
• many PHP or web server workers active
• database process consuming CPU
• backup/compression process running
• malware or unknown process using CPU
• high system CPU usage
• high I/O wait

Important values:

• user CPU: application workload
• system CPU: kernel/system work
• idle: unused CPU
• wait/I/O wait: waiting for disk operations

Memory and swap

Memory usage shows how much RAM is being used. Linux uses free RAM for cache, so low “free” memory is not always a problem.

Watch for:

• very low available memory
• high swap usage
• processes using too much RAM
• database memory pressure
• PHP workers consuming memory
• repeated out-of-memory errors
• server slowing when swap increases

Processes and users

Each running task appears as a process. top and htop show the process ID, user, CPU usage, memory usage and command.

Useful columns:

• PID: process ID
• USER: account running the process
• CPU%: processor usage
• MEM%: memory usage
• TIME+: total CPU time
• COMMAND: what process is running

Using top

top is usually installed by default.

Open top:
top

Sort by memory:
Press M

Sort by CPU:
Press P

Show full command:
Press c

Quit:
Press q

Watch a specific user:
top -u username

Using htop

htop is easier to read and navigate than top, but it may need installation.

Install htop on Ubuntu/Debian:
sudo apt update
sudo apt install htop -y

Install htop on AlmaLinux/Rocky:
sudo dnf install htop -y

Open htop:
htop

Sort by CPU:
Press F6 and select CPU%

Sort by memory:
Press F6 and select MEM%

Search process:
Press F3

Filter:
Press F4

Tree view:
Press F5

Kill process:
Press F9

Quit:
Press F10

Why this matters

Live tools like top and htop help you identify the current bottleneck before guessing or restarting services blindly.

How to check live server resources

Use htop or top during a slow website incident, then compare findings with logs and website checks.

1. Load average — Compare load with CPU core count.
2. CPU usage — Find whether CPU is saturated.
3. Memory available — Check whether RAM is low.
4. Swap usage — Look for heavy swap activity.
5. Top processes — Sort by CPU and memory.
6. Process user — Identify which account or service owns the load.
7. Command — Check whether load comes from web server, PHP, database, backup, cron or unknown process.
8. Logs — Compare process activity with web server, PHP and database logs.

Common problems

How to troubleshoot server load with top and htop

1. Step 1: Open htop or top

   Check live CPU, memory, swap and process usage.

2. Step 2: Check load average

   Compare 1, 5 and 15 minute load with CPU core count.

3. Step 3: Sort by CPU

   Identify processes using the most processor time.

4. Step 4: Sort by memory

   Find processes consuming large amounts of RAM.

5. Step 5: Identify the user

   Check whether load belongs to a website account, database, system service or unknown user.

6. Step 6: Check command

   Read the process command to understand what is running.

7. Step 7: Compare with logs

   Check web server, PHP, database and system logs at the same time.

8. Step 8: Take safe action

   Optimize, limit, reschedule or restart only after identifying the likely cause.

9. Step 9: Add monitoring

   Set alerts and historical metrics so the issue can be tracked over time.

Useful related commands

Check uptime and load:
uptime

Check memory:
free -m

Check disk usage:
df -h

Check largest directories:
du -h --max-depth=1 /var 2>/dev/null | sort -h

Check listening ports:
sudo ss -tulpn

Check recent system logs:
sudo journalctl -xe

Check Apache status:
systemctl status apache2

Check Nginx status:
systemctl status nginx

Check MySQL status:
systemctl status mysql

Check recent Apache errors:
sudo tail -100 /var/log/apache2/error.log

Check recent Nginx errors:
sudo tail -100 /var/log/nginx/error.log

Common processes you may see

The process list depends on your server stack.

Common processes:

• apache2 or httpd: Apache web server
• nginx: Nginx web server
• php-fpm: PHP worker process
• mysqld or mariadbd: MySQL/MariaDB database
• named: DNS service
• exim or postfix: mail server
• dovecot: IMAP/POP mail service
• redis-server: Redis cache
• cron: scheduled tasks
• rsync: file sync or backup
• tar/gzip: compression or backup
• certbot: SSL certificate tasks

What top and htop cannot show

top and htop show live process usage, but they do not explain everything.

They do not fully show:

• historical trends
• exact slow SQL queries
• full request URLs
• external network failures
• application-level errors
• CDN issues
• DNS issues
• user experience in browser
• long-term uptime
• exact root cause of every spike

Monitoring strategy

For production servers, combine live tools with continuous monitoring.

Monitor:

• uptime
• website HTTP status
• response time
• CPU usage
• memory usage
• swap usage
• disk space
• disk I/O if available
• database health
• web server health
• SSL expiry
• backup success
• security updates

Server resource examples

Example 1: High CPU

Symptom:
Website is slow.

htop:
php-fpm processes using 95% CPU.

Likely cause:
Heavy PHP requests, plugin issue, traffic spike or slow application code.

Next step:
Check access logs, PHP logs and recent website changes.

Example 2: Heavy swap

Symptom:
Server responds slowly.

free -m:
Swap usage is high.

htop:
Database and PHP workers consume most RAM.

Likely cause:
Not enough memory or too many concurrent workers.

Next step:
Tune services, reduce concurrency or upgrade RAM.

Example 3: Backup overload

Symptom:
Site slows every night.

htop:
tar and gzip processes use high CPU.

Likely cause:
Backup job running during active traffic.

Next step:
Move backup to low-traffic window and limit resource usage.

Frequently asked questions