distributed Order-Dependency Discovery optimization - An HPI seminar work at the Information Systems chair

Link to data sets: HPI Repeatability: FDs.

You can configure the algorithm before packing it into a jar in the application.conf-file. For all available options please consider the reference.conf-file.

After the algorithm was packaged, you can still change some configuration options via the CLI. Use the option --help to display the help page shown below:

Usage

 dODo [options] : Distributed Order dependency Discovery Optimization - discover order dependencies through order compatibility dependencies in CSV files

Options

   --host       : this machine's hostname or IP to bind against
   --input-file : path to the input CSV file. Required for the seed node! If not specified, this node tries to fetch the data from another node (potentially the seed node).
   --port       : port to bind against
   --seed-host  : hostname or IP address of the cluster seed node
   --seed-port  : port of the cluster seed node

There are different ways to run the algorithm. They are described in the following sections.

Please make sure that you have SBT version >= 1.2.8 installed and your JAVA_HOME is pointing to a JRE or JDK 1.8.x installation. Newer versions of Java are not supported!

Issue the following command from the project root directory:

sbt "run --host=localhost --port=7877"

The double quotes around the run command and the program arguments are necessary.

1. Create a new run configuration.

   • Add new configuration of type Application
   • Set as the main class.
   • You can add the above options as Program arguments. See screenshot below.

   

2. Execute the new run configuration.

1. First build and package the application by running sbt assembly. This will create a fat jar (uber jar) that can be run independently.
2. Run the application with java -jar <dodo-name>.jar --help. Only Java version 1.8.x is supported!

If you want to contribute to this project, you are more then welcome to do so. Please read the contribution guidlines before submitting new issues or pull requests.

You can change the settings specified in the packed configuration files by supplying them during start of the application with: java -jar -Dcom.github.codelionx.dodo.input-file=data/flight_1k.csv -Dakka.remote.artery.canonical.port=7878 dodo-assembly-0.0.1.jar