The purpose of the Kinetica Spark connector is to perform massive parallel ingest into Kinetica via Dataset for Spark 2.x. All interactions with the connector are through the SparkKineticaLoader API which is exposed inside the connector jar gpudb-spark-2.2.0-uber.jar.
This project uses Spark 2.2.0.
For documentation, please visit https://www.kinetica.com/docs/6.1/connectors/spark_guide.html
Please consult the setup document located at https://s3.amazonaws.com/kinetica-se/Spark-Connector/v1.0/Spark+2.x+Connector.docx