How to Calculate and Optimize PHP-FPM pm.max_children for High-Traffic WordPress on VPS

The Most Common Cause of 502 Bad Gateway Outages

When WordPress traffic surges unexpectedly, web servers often crash with 502 Bad Gateway: server reached pm.max_children setting logged in PHP error files. When pm.max_children is set too low, incoming visitor requests queue up and time out. When set too high, PHP child processes consume more physical memory than the server possesses, invoking the Linux kernel's OOM Killer to terminate MySQL.

The Exact Mathematical Formula for Sizing PHP-FPM Pools

To determine optimal pool sizing, execute this diagnostic command to find the average memory footprint of a single PHP-FPM worker process on your active WordPress site:

ps --no-headers -o "rss,cmd" -C php-fpm8.3 | awk '{ sum+=$1; n+=1 } END { if (n>0) print sum/n/1024 " MB" }'

On typical WordPress sites, an active worker consumes between 50 MB and 85 MB.

The Calculation Formula

pm.max_children = (Total Server RAM - RAM Reserved for OS & Database) / Average PHP Process Size

For an 8GB RAM VPS (e.g. Contabo VPS 1 or Hostinger KVM 2):

• Total RAM: 8,192 MB
• Reserved for OS & MySQL Buffer Pool: ~3,000 MB
• Available for PHP-FPM: ~5,192 MB
• Calculation: 5,192 MB / 65 MB = ~80 children

Production Pool Configuration

Edit your pool configuration in /etc/php/8.3/fpm/pool.d/www.conf:

pm = dynamic
pm.max_children = 80
pm.start_servers = 15
pm.min_spare_servers = 10
pm.max_spare_servers = 25
pm.max_requests = 1000

Restart PHP-FPM with sudo systemctl restart php8.3-fpm to activate your stable, high-concurrency worker pool.