Under the Hood of Reactive Data Access
summary
Spring Data provides reactive connectors for MongoDB, Redis (Lettuce), Cassandra (async wrapped into a reactive facade) and Couchbase, while no JPA/JDBC support is offered because JDBC is blocking. MongoDB uses an async driver under reactive layers, drives batch sizes from backpressure and adds MongoDB change streams in 3.6; filters can trigger many getMore calls unless Reactor’s limitRate or custom batching is used. Redis Lettuce is single-connection, Netty-based and reduces latency; Cassandra remains async for now. Command-level timeouts replace read timeouts.
AI-written summary · about
chapters (16)
- 0:00Introduction to reactive data access
- 1:15Why no JPA reactive support
- 3:01Asynchronous JDBC with Spring Data
- 5:09Reactive vs relational databases
- 6:14Synchronous vs reactive drivers
- 9:45Costs of reactive programming
- 12:57Reactive driver requirements
- 15:50Blocking calls and event loops
- 16:11Driver internals explained
- 18:40MongoDB reactive driver
- 20:27MongoDB batch size handling
- 24:04Fixing MongoDB request issues
- 26:19MongoDB change streams
- 26:59Redis reactive driver
- 29:50Cassandra reactive session
- 31:17Resources and closing
related talks
51mUnder the Hood of Reactive Data Access
46mReactive Meets Data Access
25mWhat’s New in Spring Data MongoDB
53mIntroducing Hibernate Reactive – The first reactive ORM framework
25mPolyglot Persistence Unleashed
01h 30mCassandra, Couchbase and Spring Data in the Enterprise
39mFramewars: the battle between NoSQL and Java in the cloud arena
50mNext Level Redis With Spring Data