Reactive and Asynchronous Models

Reactive and asynchronous models are both attempts to deal with work that should not block a straightforward call stack, but they solve different problems. Async workflows focus on deferred completion and composition. Reactive models focus more heavily on event streams, demand, and backpressure.

That distinction matters in Java because teams often mix these tools without being clear about the reason. A CompletableFuture pipeline, a reactive stream, and a virtual-thread request flow may all be valid, but they express different assumptions about lifetime and load.

The child pages focus on those assumptions so you can choose the right model instead of layering multiple async abstractions on top of one another by habit.

In this section

• Introduction to Reactive Programming in Java
  Learn the core reactive programming model in Java and when event-driven, non-blocking flows are worth the extra abstraction.
• Asynchronous Data Streams in Java
  Use asynchronous data streams in Java when publishers, subscribers, and demand management fit the workload better than request-style concurrency.
• Composing Complex Asynchronous Workflows in Java
  Compose complex asynchronous Java workflows with clear merge, retry, and failure boundaries instead of letting reactive chains become opaque.