Applying Reactive Programming to Existing Applications
summary
Shows how to retrofit RxJava observables as an async facade over a blocking Java codebase, wrapping JDBC and HTTP calls with create/defer/from and scheduling them on a subscribeOn/io scheduler so results stream instead of returning lists, adding Reactive Streams backpressure and TestSubscriber/TestScheduler-based tests. Uses Hystrix for isolation, bulkheading and automatic request collapsing, reduces cross-service calls and end-to-end latency in production but creates thread-count tuning, mixed-sync pitfalls and harder step‑through debugging to manage.
AI-written summary · about
chapters (10)
- 0:00Start
- 10:03IPC via Apache HTTP Memcached Cassandra
- 10:40Many things could not realistically change
- 43:59easy to block async code
- 46:03green threads, native threads, fibers, actors, event loops, theoretically equivalent, OS kernels, etc, etc, etc...
- 52:22Event loop architecture ⚫ reduced thread migrations ⚫ cache warmth → memory locality → instructions per cycle → lower cpu consumption per request
- 53:15With current JVM and Linux, event loops have efficiency benefits.
- 53:31Simpler But Not Easier
- 56:36Concurrency via Async Service Composition
- 58:22Leverage RxJava Schedulers and Threads
related talks
27mRetrofit your Java App with a Reactive Flow
35m5 Reasons to Use Reactive Programming if You Are Not eBay
50mThe reactive landscape
44mReacting to the future of application architecture
54mReacting to an Event-Driven World
01h 42mWhat is a Reactive Application?
49mRethinking Reactive Architectures
54mHow I Learned to Stop Worrying & Love the Flow