Skip to content

tuftjs/bench

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

37 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Tuft Benchmark Comparison

This repository contains a suite of basic example applications for a few different routing-enabled frameworks, including Tuft, that have been set up for benchmarking purposes. Each application has a single GET route that responds with the text 'Hello, world!\n'.

The frameworks to be tested include:

Methodology

The results below were obtained using two CPU-optimized instances in the same VPC on Digital Ocean. The application servers were executed on the first instance (the server), and the load testing was performed on the second instance (the client) using AutoCannon.

Results

  • Date: September 16, 2020
  • Hardware: Digital Ocean CPU-optimized droplet
  • OS: Ubuntu 20.04 (LTS) x64
  • Node version: 12.18.4
  • Client command: autocannon -c 100 -d 40 (two rounds: one for warm-up, one for results)
Framework Version Req/s Bytes/s (MB) Latency (ms)
none N/A 22,474 3.48 4.08
Tuft 2.3.1 20,872 3.23 4.17
Fastify 3.4.1 18,858 2.92 5.08
Koa 2.13.0 14,959 2.32 6.14
Hapi 20.0.0 12,583 2.53 7.28
Express 4.17.1 6,037 1.32 16.1

People

The creator and maintainer of Tuft is Stuart Kennedy.

License

MIT