This is an supporting application for Gradoop, allows importing FinBench datasets as TPGM in Gradoop for further graph analyzing and processing.
This project is built using Apache Flink 1.9.3 and Gradoop 0.7.0-SNAPSHOT
This project was developed during a Bachelor's thesis at Leipzig University.
- Maven.
- Java 8.
- FinBench Datasets provided by LDBC.
Bulding should be done using Maven. Simply run this in the project directory:
mvn package
This will generate the JAR package FinBenchGradoopImporter-1.0.jar inside target folder.
This application can be executed using Apache Flink. An example Command is:
/bin/flink run -p 128 -c org.gradoop.importer.finbench.FinBenchImporter FinBenchGradoopImporter-1.0.jar -i
hdfs:///finbench/sf10 -o hdfs:///finbench/gradoop-parquet-protobuf -f protobuf
| Parameter | Argument | Description | Required |
|---|---|---|---|
-i |
/path/to/finbench |
The input path to a directory containing all Finbench's csv files. | Yes |
-o |
/path/out |
The output path for the Gradoop graph to be written. | Yes |
-f |
csv or indexed or parquet or protobuf |
The output format. CSV is default, parquet or protobuf are fastest. | Yes |
Apache®, Apache Accumulo®, Apache Flink, Flink® are either registered trademarks or trademarks of the Apache Software Foundation in the United States and/or other countries.