Case study
Slow WooCommerce checkout caused by database and PHP-FPM pressure
Slow checkout, admin screens and product pages can directly affect sales on a revenue-generating WooCommerce store. This case study outlines how we approach a production performance issue that needs careful diagnosis and safe changes.
Context
A growing online store was running WordPress/WooCommerce on a Linux server with NGINX, PHP-FPM, MySQL/MariaDB, Redis caching and Cloudflare in front.
The owner did not need a new website. They needed the existing stack made more reliable without breaking checkout or causing downtime.
The problem
- Checkout became inconsistent during busy periods and admin pages felt slow.
- PHP-FPM workers were being exhausted, which made NGINX return intermittent 502/timeout behaviour.
- The database contained slow queries and high-memory behaviour caused by plugin, order and session activity.
- Caching helped some pages, but it did not solve logged-in, cart or checkout performance.
Our approach
- Reviewed PHP-FPM pool behaviour, worker limits, slow logs and memory pressure before changing live settings.
- Checked MySQL/MariaDB configuration, slow query patterns, table growth and WordPress/WooCommerce-specific database load.
- Reviewed Redis/object-cache behaviour and made sure cache configuration was helping rather than masking the issue.
- Applied controlled tuning, then monitored checkout, admin and server load after changes.
Practical outcomes
Relevant technologies and keywords
These are the main technologies, services and search terms connected to this case study.
Related services
Relevant services for similar infrastructure problems.
Want help with a similar issue?
Send the symptoms, affected service, recent changes and business impact. We will suggest the most appropriate route: emergency support, a fixed-scope technical fix, an infrastructure review or a wider project.