scala-datapipeline-dsl's Introduction

AWS DataPipeline DSL for Scala

A Scala domain-specific language and toolkit to help you build and maintain AWS DataPipeline definitions.

This tool aims to ease the burden of maintaining a large suite of AWS DataPipelines. At Shazam, we use this tool to define our data pipelines in Scala code and avoid the boilerplate and maintenance headache of managing 10s or 100s of JSON pipeline configuration files.

Benefits:-

Write and maintain Scala code instead of JSON configuration
Use the DSL's >> syntax to clearly express dependencies between your pipeline's activities
Share code/configuration between your pipeline definitions
Never write dependsOn or precondition again, this library manages all ids and object references for you
Add your own wrapper around this library to predefine most your most commonly-used data pipeline objects

Tutorial

Build the compiler using sbt:

$ sbt assembly

Create a "Hello World" AWS Data Pipeline definition Scala file:

object HelloWorldPipeline {

  import datapipeline.dsl._

  val pipeline =
    AwsDataPipeline(name = "HelloWorldPipeline")
      .withSchedule(
        frequency = Daily,
        startDateTimeIso = "2018-01-01T00:00:00"
      )
      .withActivities(
        ShellCommandActivity(
          name = "Echo Hello World",
          workerGroup = "my-task-runner",
          Command("echo 'Hello AWS Data Pipeline World!'")
        )
      )

}

Use the compiler to produce JSON from our Scala definition:

$ java -jar target/scala-2.12/datapipeline-compiler.jar HelloWorldPipeline HelloWorldPipeline.scala
Writing pipeline definition to: ./HelloWorldPipeline.json

The output JSON file contains your pipeline definition ready to deploy to AWS.