When running PHP on Apache servers, you typically have two main approaches:

// mod_php (traditional way)
LoadModule php_module modules/libphp.so

// PHP-FPM (FastCGI Process Manager)
FastCgiServer /usr/bin/php-cgi -processes 4

mod_php loads the PHP interpreter directly into each Apache process, consuming memory even for static content. PHP-FPM maintains separate PHP processes:

# Memory usage comparison (typical scenario)
mod_php: ~50MB per Apache process (PHP always loaded)
PHP-FPM: ~30MB per PHP worker + ~5MB for Apache (lightweight proxy)

PHP-FPM's process isolation delivers better stability during traffic spikes. Here's a load test configuration example:

; php-fpm.conf tuning for high traffic
pm = dynamic
pm.max_children = 50
pm.start_servers = 5
pm.min_spare_servers = 5
pm.max_spare_servers = 35

Running PHP as separate FastCGI processes allows for:

• User isolation (running PHP processes under different system users)
• Better protection against memory-based attacks
• Ability to chroot PHP processes

PHP-FPM enables per-site PHP configurations:

# Virtual host example with custom PHP settings
<VirtualHost *:80>
    ServerName example.com
    <FilesMatch \.php$>
        SetHandler "proxy:fcgi://127.0.0.1:9000"
    </FilesMatch>
    # Custom PHP settings for this host
    <Location "/admin">
        php_admin_value memory_limit 256M
    </Location>
</VirtualHost>

Tests on a 4GB VPS running WordPress:

At its heart, FastCGI operates as a persistent external process while mod_php runs as an Apache module. This fundamental distinction creates several technical implications:

// Traditional mod_php execution flow
Apache HTTPD -> mod_php -> PHP interpreter (embedded)

// FastCGI execution model
Web Server (nginx/Apache) -> FastCGI Protocol -> PHP-FPM Process Manager

With mod_php, each Apache child process loads the PHP interpreter - meaning 10 Apache workers consume 10x PHP memory overhead. FastCGI's process isolation means:

• PHP processes can be recycled independently of web server workers
• Memory leaks don't accumulate in the web server
• php.ini changes don't require full web server restart

Benchmarks on a WordPress installation show:

Setting up PHP-FPM with Apache requires these directives:

<FilesMatch \.php$>
    SetHandler "proxy:fcgi://127.0.0.1:9000"
</FilesMatch>

# In php-fpm.conf
pm = dynamic
pm.max_children = 50
pm.start_servers = 5
pm.min_spare_servers = 2
pm.max_spare_servers = 8

PHP-FPM provides granular process management not possible with mod_php:

# Different pools for different sites
[www]
user = www-data
listen = 127.0.0.1:9000

[api]
user = api-user
listen = 127.0.0.1:9001
request_slowlog_timeout = 5s
slowlog = /var/log/php-fpm/api-slow.log

The process isolation in FastCGI provides security benefits:

• PHP runs under separate user context from web server
• SuEXEC-style privilege separation possible
• No PHP environment variables exposed to web server