State Management in Apache Flink (Notes)
What are key takeaways
- Managing application state.
- Strong consistency guarantees.
- Adaptivity to cluster reconfig.
What is meant by transparent and non-intrusive?
What is meant by application state declared explicitly?
How does this help efficient partitioning?
How are stream databases used in contrast to traditional databases?
- Traditional database will share access to a db instance, perform computation on that instance.
- Stream database on the other hand will perform computation, where n people can subscribe to various computations.