Introduction

A sudden spike in CPU usage or server load on a web server can slow down websites, trigger 504 Gateway Timeout errors, or cause complete downtime. Whether you are running Nginx, Apache, or a reverse-proxy combination, high resource usage usually stems from unoptimized configurations, heavy database/PHP queries, bot scraping, or traffic spikes.

In this guide, we will walk through a step-by-step process to diagnose the root cause of high CPU load on Linux servers and execute the exact fixes needed to keep your server running fast and stable.

---

Step 1: Diagnose the Overloaded Resource

Before tweaking web server settings, identify which process or user is consuming the CPU.

1. Check Real-Time Process Activity

SSH into your server and run:
top

Or, if installed, use htop for a better visual interface:
htop

Look at the %CPU column to identify whether the heavy processes belong to nginx, httpd/apache2, php-fpm, or mysqld.

2. Sort Heavy Processes by CPU Usage

Run the following command to sort the top 10 CPU-hogging processes:
ps aux --sort=-%cpu | head -n 10

---

Step 2: Fix High CPU Usage in Nginx

If Nginx processes are spiking, it is typically due to worker process misconfigurations, lack of caching, or worker connection limits.

Fix 1: Optimize Nginx Worker Processes

Open your Nginx configuration file:
nano /etc/nginx/nginx.conf

Ensure your worker_processes matches your physical CPU cores (or set it to auto so Nginx handles it dynamically):

worker_processes auto;
worker_connections 2048;

Fix 2: Enable Static File Caching & Compression

Serving uncompressed static files forces Nginx to work harder on every request. Enable GZIP compression and browser caching in your server block:

Enable Gzip Compression

gzip on; gzip_comp_level 5; gzip_types text/plain text/css application/json application/javascript text/xml;

Fix 2: Lower KeepAlive Timeout

After modifying settings, test and reload Nginx: `nginx -t` `systemctl reload nginx`

Step 3: Fix High CPU Usage in Apache

Apache high load is frequently caused by process-spawning loops in old MPM configurations or excessive KeepAlive connection holding times.

### Fix 1: Switch to Event MPM
If Apache is using the prefork module, every connection creates a full process, destroying CPU and RAM efficiency. Switch to the Event MPM module paired with PHP-FPM.

Check your active MPM module:
`apachectl -M | grep mpm`

Adjust your event configuration inside `/etc/httpd/conf.modules.d/00-mpm.conf` or `/etc/apache2/mods-available/mpm_event.conf`:

Long KeepAlive timeouts keep Apache worker slots busy waiting for idle clients. Reduce KeepAliveTimeout to 2 or 3 seconds in your Apache config:

KeepAlive On
MaxKeepAliveRequests 100
KeepAliveTimeout 3

Restart Apache to apply: systemctl restart apache2

Step 4: Address Underlying PHP & Database Bottlenecks

In many cases, the web server (Nginx/Apache) appears to take high CPU only because it is waiting on slow PHP executions or unindexed MySQL queries.

1. Enable PHP OPcache: OPcache stores precompiled script bytecode in memory so PHP doesn't have to load and parse scripts on every request.
2. Enable Slow Query Logging in MySQL: Find queries that block CPU threads by adding this to /etc/my.cnf:

slow_query_log = 1
slow_query_log_file = /var/log/mysql/slow.log
long_query_time = 2

Need Proactive 24/7 Server Management?

Tuning web servers and troubleshooting intermittent CPU spikes requires constant monitoring and expertise.

At CoreServerSupport, our team of Linux & Cloud infrastructure experts provides 24/7 server monitoring, performance tuning, hardening, and emergency troubleshooting to guarantee 99.99% uptime for your business.

Contact CoreServerSupport Today for a free server performance audit!