Addressing Messaging Challenges Using Open Technologies
summary
A synchronous kiosk app is refactored into an asynchronous pipeline using Spring Integration and Spring AMQP with RabbitMQ as the broker: HTTP inbound adapters become fire‑and‑forget producers, AMQP outbound adapters queue orders, and a separate consumer JVM pulls and completes deliveries so orders persist while the backend is down. Demo shows queued orders processed when the backend starts, claims RabbitMQ handles ~10–25k msgs/sec (persistence optional), and a Hadoop MapReduce job on HDFS counts device user agents (iPhone wins).
AI-written summary · about
chapters (19)
- 0:00Session introduction
- 1:22Distributed systems
- 3:54Big data and modern app dev
- 8:18Synchronous vs asynchronous
- 12:40Middleware challenges
- 18:04Spring integration basics
- 26:42Lame coffee shop demo
- 32:27Refactoring for async
- 36:41Amqp messaging demo
- 43:50Audience questions
- 46:17Amqp protocol details
- 50:12Routing and bindings
- 51:52Exchange types
- 53:57Persistence and reliability
- 58:59JMS vs AMQP
- 1:04:01Backend configuration
- 1:07:06Hadoop integration demo
- 1:13:37Data analysis results
- 1:24:52Concluding Q&A
related talks
35mNo Rocket Scientist Required - Developing +14,000 Message p/second IOT Systems
50mBuilding reactive microservices with MicroProfile
27mRetrofit your Java App with a Reactive Flow
47mReactive Spring
30mRetrofit Your Java App With a Reactive Flow Pipeline
46mAsyncAPI + Spring Cloud Stream = Event-Driven Microservices Made Easy
01h 01mSupersonic, Subatomic Java
53mRefactor to Reactive with Spring 5 and Project Reactor