Profiling and performance #

This page is for application creators who make code faster. Measure, change one thing, measure again (Benchmarking, Benchmark reference).

Contents #

1. Measuring pages
2. Inside a request
3. Comparing fairly
4. The usual causes

Measuring pages #

php bin/php/console exp:benchmark --save=var/benchmark/base.json --allow-root-user
# change something
php bin/php/console exp:benchmark --baseline=var/benchmark/base.json --threshold=15 --allow-root-user
php bin/php/console exp:benchmark --cold --requests=50 --allow-root-user   # past the response caches

It requests the front page, menu pages, search and the admin login, 100 times each after warm-up, and prints median, p90, p99 and more.

Inside a request #

• exp:benchmark kernel --repeat=30 times boot, render (cold and warm), INI load, a content fetch, a database round trip and cache writes and reads in-process.
• exp:benchmark micro times pure-PHP hot paths without HTTP or a database.
• The debug bar shows each request's SQL queries, templates and timings (Exp debug bar).
• Under Velocity, the opcode cache profile shows why a page costs what it does (Velocity and the opcode cache).

Comparing fairly #

Same machine, nothing else running, OPcache as in production, Xdebug off, and judge two servers only by runs made back to back. Every run records its environment, and a comparison with a baseline made under another setup says so.

The usual causes #

Related pages #

• Scaling applications
• Caching in applications
• Performance and caching for websites
• Free documentation: this page is free to copy, translate and improve under the GNU FDL 1.3.