Navigating Green Waters: A Journey with Cassandra
summary
Maritime IoT platform that runs Cassandra on both ships and cloud handles 3,000–5,000 datapoints per 30s from ~450–1,600+ vessels, drove 230,000 tCO2 and $38M saved, and evolves from open-source Cassandra to DataStax Enterprise and moved to DataStax Astra to cut ops. Problems included poor schema, unstable clusters, and heavy maintenance; solutions were refactoring schemas, using Astra serverless for autoscaling and security, and storing sensor maps as JSON-string sensor payloads for best insert/read performance.
AI-written summary · about
chapters (19)
- 0:00Cassandra journey introduction
- 3:58Why we chose Cassandra
- 7:10Cassandra’s masterless architecture
- 8:14Early struggles with Cassandra
- 9:41Transition to DataStax Enterprise
- 11:06Challenges as new users
- 13:12DataStax Enterprise improvements
- 14:37Moving to Astra from DSE
- 16:25Astra’s serverless benefits
- 17:30Security and compliance in Astra
- 18:55Impact after switching to Astra
- 20:21Astra versus other services
- 21:06Astra innovations and features
- 22:10Astra tips and lessons
- 22:32Time-series data models
- 27:58Performance testing results
- 28:40Performance tips for Astra
- 29:44Cost efficiency with Astra
- 30:29Data migration strategies
related talks
35mMission Critical Data, and Evaluating the Community in OSS Tech
35mReplacing Cassandra: A Digital Transformation for the World’s Largest Digital Identity Network
24mStaying Ahead of the Curve with Spring and Cassandra 4
19mBuild Real time, AI powered Apps
52mStargate, an OSS Api Layer for Apache Cassandra.
16mBuild for real time on AWS: Thrive in an unpredictable market
46mCassandra Serving Netflix @ Scale
26mStargate.io, An OSS Api Layer for your Cassandra