Benchmarks for Tarantool PHP connectors

April 15, 2020 · View on GitHub

This repository contains code to benchmark PHP connectors for Tarantool. Currently, there are two connectors available:

A more detailed explanation of the benchmarks represented in this repository can be found in this article.

Requirements

First, make sure you have the bench.lua instance running.

If you want to run it on Docker, execute:

docker run -d --network host --name=tarantool-bench \
   -v $PWD/bench.lua:/bench.lua tarantool/tarantool:2 tarantool /bench.lua

Then run

make

to see the usage text and a list of all available benchmarks. For example, to (re)run all benchmarks, execute

make clean bench-all

Alternatively, you may run benchmarks on Docker using Docker Compose:
docker-compose run --rm benchmarks make clean bench-all

You may change default benchmark settings by defining the following environment variables (in parentheses are the default values):

For example:

make clean bench-all TNT_BENCH_REVOLUTIONS=20000 TNT_BENCH_RETRY_THRESHOLD=5

The below results were made by running benchmarks on Apple MacBook Pro (2015) on the following environment: