Table of Contents
The idea behind this repository is to benchmark different languages implementation of HTTP server.
The application i tested is minimal: the HTTP version of the Hello World example.
This approach allows including languages i barely know, since it is pretty easy to find such implementation online.
If you're looking for more complex examples, you will have better luck with the TechEmpower benchmarks.
Please do take the following numbers with a grain of salt: it is not my intention to promote one language over another basing on micro-benchmarks.
Indeed you should never pick a language just basing on its presumed performance.
I have filtered the languages by single runtime (i.e. Java on JVM): this way i can focus on a specific stack, keeping it updated to the last available version/APIs. Where possible i just relied on the standard library, but when it is not production-ready (i.e. Ruby, Python) or where the language footprint is deliberately minimal (i.e. Rust).
Ruby 2.5.1 is installed via rbenv.
Ruby is a general-purpose, interpreted, dynamic programming language, focused on simplicity and productivity According to his creator Ruby is inspired by Lisp, Perl, Python and Smalltalk.
Python 3.7.0 is installed via homebrew.
Python is a widely used high-level, general-purpose, interpreted, dynamic programming language.
It supports several programming paradigms and can count on a broad standard library.
Node.js version 10.1.0 is installed by official OSX package.
Dart version 1.24.3 is installed via homebrew.
Elixir 1.6.4 is installed via homebrew.
Elixir is a purely functional language that runs on the Erlang VM.
While preserving Erlang key-features, Elixir is strongly influenced by Ruby syntax and supports compile-time metaprogramming with macros and polymorphism.
Java JDK 10.0.1 is installed by official OSX package.
Java is a general-purpose language that is concurrent, class-based, object-oriented and, from version 8, supports functional paradigms. It is based on a battle-tested virtual machine (JVM) that kept the promise "write once, run anywhere".
C-sharp (C#) 7.0 language is part of the [.NET Core]((https://www.microsoft.com/net/core#macos) 2.1.3 framework.
C# is a simple, powerful, type-safe, object-oriented language. It inherited many features from Java, but recently added some desirable paradigms such as futures, pattern matching and deconstructions.
Crystal 0.25.0 is installed via homebrew.
Crystal has a syntax very close to Ruby, but brings some desirable features such as statically typing and raw metal speed, thanks to ahead of time (AOT) compilation.
Nim 0.18.0 is installed via homebrew.
GO language version 1.10.3 is installed by official OSX package.
GO focuses on simplicity by intentionally lacking features considered redundant. GO takes a straight approach to parallelism, coming with built in CSP and green threads (go-routines).
Rust language version 1.27.0 is installed by official package.
According to the official site Rust is a systems programming language that runs blazingly fast, prevents segfaults, and guarantees thread safety.
These bounds are granted courtesy of Rust's pretty unique ownership model enforced by the compiler.
I used wrk as the loading tool.
I measured each application server six times, picking the best lap (but for VM based languages demanding longer warm-up).
wrk -t 4 -c 100 -d30s --timeout 2000 http://0.0.0.0:9292
These benchmarks are recorded on a MacBook PRO 15 mid 2015 having these specs:
- macOS High Sierra
- 2.2 GHz Intel Core i7 (4 cores)
- 16 GB 1600 MHz DDR3
RAM and CPU
I measured RAM and CPU consumption by using macOS Activity Monitor dashboard and recording max consumption peak.
For the languages relying on pre-forking parallelism i reported the average consumption by taking a snapshot during the stress period.
|Language||App Server||Req./sec (local)||RAM (MB)||CPU (%)|
|Elixir||Plug with Cowboy||43100.86||42.4||530.8|
|Ruby||Puma||58710.33||> 100||> 390|
|Python||Gunicorn with Meinheld||100932.26||> 40||> 350|
puma -w 8 -t 8 --preload servers/puma_server.ru
Gunicorn with Meinheld
cd servers gunicorn -w 4 -k meinheld.gmeinheld.MeinheldWorker -b :9292 gunicorn_server:app
I used the cluster module included into Node's standard library.
I used the async HTTP server embedded into the Dart standard library.
Plug with Cowboy
cd servers/plug_server MIX_ENV=prod mix compile MIX_ENV=prod mix run --no-halt
I tested Java by using Jetty with the non blocking IO (NIO) APIs.
cd servers/jetty_server javac -cp jetty-all-uber.jar HelloWorld.java java -server -cp .:jetty-all-uber.jar HelloWorld
To test C# i opted for Kestrel: a cross-platform web server based on the libuv asynchronous I/O library.
cd servers/kestrel_server dotnet restore dotnet run
I used Crystal HTTP server standard library, not supporting parallelism yet.
crystal build --release servers/crystal_server.cr ./crystal_server
I used the asynchttpserver standard module of Nim, not supporting parallelism yet.
nim c -d:release servers/nim_server.nim ./servers/nim_server
I used the HTTP ServeMux GO standard library.
go run servers/go_server.go
cd servers/hyper_server cargo run --release