ScyllaDB

ScyllaDB is a NoSQL database with a close-to-the-hardware, shared-nothing approach that optimizes raw performance, fully utilizes modern multi-core servers, and minimizes the overhead to DevOps. ScyllaDB is API-compatible with both Cassandra and DynamoDB, yet is much faster, more consistent, and with a lower TCO.

— ScyllaDB Homepage

ScyllaDB Setup and Connection

ScyllaDB is fully compatible with Cassandra. To use it as the data storage layer:

1. Spin up a Scylla cluster. You can do this using Scylla Cloud, using Docker, running Scylla in the cloud or on-prem. You can see a step-by-step guide on how to spin up a three node Scylla cluster using Docker in this lesson.
2. Run JanusGraph with “cql” as the storage.backend.
3. Specify the IP address of one of the Scylla nodes in your cluster as the storage.hostname.

Step by Step Tutorial

This Scylla University lesson provides step-by-step instructions for using JanusGraph with ScyllaDB as the data storage layer. The main steps in the lesson are:

• Spinning up a virtual machine
• Installing the prerequisites
• Running the JanusGraph server (using Docker)
• Running a Gremlin Console to connect to the new server (also in Docker)
• Spinning up a three-node Scylla Cluster and setting it as the data storage for the JanusGraph server
• Performing some basic graph operations