Kafka Storage Adapter for JanusGraph

JanusGraph is an Apache TinkerPop enabled graph database that supports a variety of storage and indexing backends. This project adds Kafka to the supported list of backends.

Installing the adapter from a binary release

Binary releases can be found on GitHub.

This installation procedure will copy the necessary libraries, properties, and Gremlin Server configuration files into your JanusGraph installation.

1. Download the JanusGraph release.
2. Download the Kafka storage adapter release.
3. Unzip the storage adapter zip file and run ./install.sh $YOUR_JANUSGRAPH_INSTALL_DIRECTORY

Assuming you have a Kafka cluster up and running, you can connect from the Gremlin console by running:

gremlin> graph = JanusGraphFactory.open('conf/janusgraph-kafka.properties')

To start Gremlin Server run gremlin-server.sh directly or bin/janusgraph.sh start which will also start a local Elasticsearch instance.

Installing from source

Follow these steps if you'd like to use the latest version built from source.

1. Clone the repository. git clone http://github.com/rayokota/janusgraph-kafka
2. Build the distribution package. mvn package -DskipTests
3. Follow the binary installation steps starting at step 3.

Configuration Options

Property	Description	Default
storage.conf-file	Path to the configuration file for KCache.
storage.kafka.bootstrap-servers	List of Kafka brokers to connect to. Overrides the value in the KCache configuration file.
storage.kafka.topic-prefix	The prefix to be used for topic names. Defaults to the graph name.
storage.kafka.rocksdb-enable	Whether to enable RocksDB within KCache.	true
storage.kafka.rocksdb-rootdir	Root directory for RocksDB storage.	/tmp