GE Predix: Industrial-Strength IoT at Scale

Scalability and performance are paramount for building an industrial IoT platform. At General Electric we have been using multiple data platforms to support our customers. Our time-series platform collects information from edge devices to offer data scientists, developers and controls engineers with a secured platform to analyze their workload. We used Cassandra as our system of record for several years. In the last 12 months we have shifted to ScyllaDB. ScyllaDB's ability to consolidate nodes has enabled us to reduce our cluster footprint by over 60%. The cost savings on operation and stability of ScyllaDB enables us to support more customers, increase our up-time and lower the cost of operations. In this session, we describe the migration process we followed when moving from Cassandra to ScyllaDB, while maintaining our operations and onboarding new customers.