From Spring + Java to Spring + Akka - A Journey of Discovery
summary
Takes a Spring Travel app and replaces the search layer with Akka actors and Scala: an in-memory indexed cache, 9 million actors and ~3M messages/sec on a 12‑core testbed, linear scaling toward 64 cores, and a round‑robin router to turn one actor into five parallel workers. Builds a scatter‑and‑gather country/city shard tree, per‑request interceptor actors, an LRU query cache, a throttler using context.become, and supervision via let it crash (one‑for‑one/all‑for‑one) — all working, with remoting/config-only deployment and console monitoring.
AI-written summary · about
chapters (14)
- 0:00Introduction and speakers
- 2:20Hotel business scalability problem
- 10:40Why choose Akka framework
- 12:27What is an actor model
- 21:48Integrating Akka with Spring
- 33:49Scaling with actor routers
- 37:45Scaling out with Akka remoting
- 45:35Scatter gather pattern
- 54:04Fault tolerance with let it crash
- 1:06:05Dynamic behavior with become
- 1:08:53Live demo throttling system
- 1:15:50Query caching with actors
- 1:21:05Going fully non-blocking
- 1:23:13Q&A and resources
