Massively Scaled High Performance Web Services with PHP
summary
Design Home runs PHP 7 behind nginx in Docker images on Amazon ECS serving over 1 million DAU and ~100,000 requests per minute with median responses around 56 ms. The stack uses Composer, New Relic, Redis and Couchbase for hot storage, Docker base images pinned to PHP versions, supervisorD to run nginx + php-fpm, and auto-scaling EC2 instances. Performance wins come from background work via fastcgi_finish_request, tuned gzip and proxy settings, Redis pipelining and persistent PHP Redis, and disciplined Docker images in ECS and Couchbase and Redis usage.
AI-written summary · about
chapters (10)
- 0:00Start
- 7:141.1. Decouple Microservices
- 10:011.2. Dockerized Microservices (cont.)
- 13:301.3. Development Tools Integration
- 16:122.1. Background Processing (cont.)
- 21:2822. HTTP Compression
- 22:3622 HTTP Compression (cont.)
- 31:273.3. Faster PHP Drivers (Redis)
- 36:184.1. Deployment with Amazon ECS (cont.)
- 39:5542. Hardware Upgrade
related talks
36mApache, nginx, Docker and the Swear Words We Love
45mPHP is the king ...
37mLessons learned from dockerizing complex Web Apps
37mScaling LAMP doesn't have to suck
49mDocker & PHP development and deployment
25mFrom PHP and on-prem servers to NodeJS/React and AWS
18mMy Take on PHP
31mFrom Runtime to Resilience: Scaling PHP