Case Study: Netflix achieves personalized, always-on streaming at massive scale with DataStax

Netflix Gives Users Exactly What They Want Everytime

Netflix, the world’s leading streaming service with tens of millions of members and billions in digital revenue, faced a scaling and availability crisis as it moved vast amounts of customer behavior data to the cloud. Its legacy Oracle database became a single point of failure—causing frequent downtime for schema changes and limiting capacity—while international expansion and personalization demands drove traffic to billions of reads and writes per day.

Netflix evaluated alternatives and adopted Apache Cassandra and then DataStax Enterprise to replace its relational backend. The schemaless, multi–data-center architecture eliminated downtime for schema changes, enabled rapid cluster creation around the globe, and delivered over 10 million transactions per second; today about 95% of Netflix’s non-program data lives in DataStax, powering massive personalization and data-driven decisions such as commissioning new content.

Netflix

Adrian Cockcroft

Cloud Architect