Skip to content

Latest commit

 

History

History

flink_benchmark

Sample streaming application using flink

Introduction

This application implements our use case as a flink application.

Setup

The project is build using the scala build tool (SBT). Download and set up as described in https://www.scala-sbt.org/download.html. After installing, the sbt shell should be available by running the command sbt. In this dir, run sbt assembly to build a fat jar of ~70MB. Other usefull commands are sbt clean, sbt update and sbt reload.

Usage

Install flink and startup the cluster using the commands provided in setup_flink.sh located in the root diretory of this repo. Either submit the previously build jar via CLI or open the web frontend on 127.0.0.1:8080 or 127.0.0.1:8081. The build will be located in /target/scala-2.11. In the web frontend choose 'Submit New Job' click 'Add New' and select the jar.

Possible (and aslo necessary) command line options are:

  • --ip The host from which to pull the data to process, ususally 127.0.0.1
  • --checkoutPort The port on the provided host, which provides the endpoint for checkout data
  • --adPort Dto. for ad data
  • --outputPath The file sink for processed data. Attention: If the file is allready present, it will be overwritten without warning

Before submitting the application the data generator must be ready for transmitting data.