Compose's first graph database: JanusGraph


At Compose we've always looked to ensure you can get the databases you need. Today, we are proud to announce that JanusGraph is coming to Compose and will bring with it the power of fully open source graph databases.

JanusGraph is a new player in databases with a deep heritage. It builds on a fork of the Titan graph database, a previous leader in open source graph databases. That code is capable of being plugged into a number of different database backends. It's all then integrated with the database-agnostic Apache Tinkerpop graph framework. The JanusGraph project itself is organized under the Linux Foundation and led by developers from Expero, Google, GRAKN.AI and IBM. And it's all open source with new companies joining the community to enhance JanusGraph.

At Compose, we've worked with IBM's JanusGraph developers to combine Compose's one-click deployment, high-availability, managed database platform with JanusGraph. A great graph database demands a great backend and we've teamed it with Scylla, the high-performance Cassandra compatible database for best reliability. Then we added our automated backup system, private VLAN configuration and HAProxy managed access to give peace of mind.

That means that from today, Compose users can deploy the industry leading graph database from their Compose account.

Why a graph database?

Graph databases model the world as nodes and directed connections, vertices, and edges as graph theory calls them. Both can have properties associated with them and the connections. Both are equal elements in how the database is managed and queried and a query on a graph database can start at a point and explore the connections around it so you can say "I'm looking for any person who likes brand X who has a friends or friends of friends who buy brand Y and Z".

Relational databases typically treat relations as a simple connection between one row and another or demand that you add another table to associate data with the relationship. That means that when you want to query across relationships and examine the network that exists, you have to do a lot of expensive queries.

A Graph database as part of your datalayer allows you to understand and explore relationships and networks within your data without compromising the performance of your production relational and document stores.

JanusGraph on Compose

We're launching JanusGraph on Compose as a beta as we build the functionality around it. You'll find it ready to deploy in the beta section of the Create Deployment view of Compose. If you haven't discovered Compose yet, you can sign up for a free 30-day trial below. Already use Compose? You can deploy JanusGraph and we'll add a credit to your account so the first month and 5GB are free. If you want to learn more about JanusGraph on Compose, check out our JanusGraph documentation.

Interested in learning more? Join us July 25th, 2017 for a webinar led by Keith Lohnes, “Gremlin Traversals for the SQL User.“

Try Compose free for 30 days

If you have any feedback about this or any other Compose article, drop the Compose Articles team a line at We're happy to hear from you.

attribution Nayuki

Josh Mintz
Josh Mintz is an Offering Manager at IBM Watson Data Platform. He has an enthusiasm for homemade hummus, foreign policy, and the English Premier League. Love this article? Head over to Josh Mintz’s author page to keep reading.

Conquer the Data Layer

Spend your time developing apps, not managing databases.