zsjoska/akka-http-rest

Example of reactive REST service written on akka-http with slick

Akka Slick REST service template

Goal of example is to show how create reactive REST services on Lightband stack with Akka and Slick.

Example contains complete REST service for entity interaction.

Features:

• CRUD operations
• Entity partial updates
• CORS support
• Authentication with JWT tokens
• Test coverage with ScalaTest
• Migrations with FlyWay
• Ready for Docker
• Testing with in-memory postgres instance that launch automatically
• HikaryCP as connection pool

Requirements

• JDK 8 (e.g. http://www.oracle.com/technetwork/java/javase/downloads/index.html);
• sbt (http://www.scala-sbt.org/release/docs/Getting-Started/Setup.html);

Development guide

This application is fully tested with Unit and IT tests. You don't need to launch server locally for development. My recommendation is to write a test before changes and work via TDD. To ensure that application working properly, you should run it: sbt test.

Structure

All business logic is located in core package, every package inside is related to some domain. Service classes contains high level logic that related to data manipulation, that means that service MUST NOT implement storaging and querying for the data. For storaging there are Storage classes that always have interface with two implementation, production one and in-memory one. That's needed to fasten tests of services and make it independent from each other.

Checking code coverage

To generate code coverage report, please run: sbt clean coverage test coverageReport. Then you will have HTML pages with reports in /target/scala-2.12/scoverage-report

Packaging

Application packaging implemented via sbt-native-packager plugin. Currently in build.sbt enabled two types: docker and universal.

Universal packager
To package application as a universal app, use: sbt universal:packageBin. Application zip archive will be generated in /target/universal/ folder.

Docker packager
To package application as docker image, use sbt docker:publishLocal. It will generate and push application image into your local docker store. For information about publishing to external store, please, read plugin documentation.

Running

If you want to launch application locally (its not recommended) you need to start Postgres instance locally and fulfill some env variables:

• JDBC_URL - url to your database
• JDBC_USER - database username
• JDBC_PASSWORD - database password

After that, just run sbt run and enjoy hacking. For better expirience you can use sbt reStart that will give you ability to restart application without restarting of sbt.

Deployment on production

Easiest way to deliver your application, is to do it with docker. Publish image into the store and then use docker-compose file with structure like in docker-compose.yml.

Live example

Application deployed on heroku and can be accessed by URL http://akka-http-rest.herokuapp.com/.
First request can take some time, because heroku launch up project.
You can see documentation for this example on Apiary.

Copyright

Copyright (C) 2017 Arthur Kushka.
Distributed under the MIT License.