Introduction to Akka Stream
summary
This talk presents Akka Streams end-to-end, covering sources, flows, sinks, custom graphs and materialized values, and how to compose reusable graphs in code. It shows a production Big Data case at Lizeo (tens of millions of products/day) and demos using a 1,000,000-message Kafka topic with simulated 500 ms async processing and a 4100 MB JVM cap to demonstrate that a naive pipeline OOMs or drops messages while an Akka Streams pipeline uses back-pressure, the Kafka connector, and materialized Future[Int] results to remain bounded and keep working.
AI-written summary · about
chapters (16)
- 0:00Start
- 1:01What is it for?
- 2:39Processor
- 2:53Reactive Streams - Akka Streams
- 3:22SimpleStream Demo
- 5:31Materialization
- 11:54Composition / Reusability
- 15:58Other forms...
- 18:05Processing Topology
- 19:03CustomFlow Demo
- 24:23Asynchronous?
- 25:06Overproduction
- 28:17DropMessages Demo
- 30:35Overconsumption
- 31:14Instability
- 31:38Pull-based backpressure
related talks
28mStructure to Streams: Applying Full-Lifecycle API Management to ED Systems
26mWeek 1 Live Q&A
41mFixin the Hard Bits of Event Processing with Restate & Kafka
44mAn Introduction to Edge Extensions, Client API and Data Explorer
38mInteractive Session: Flink on Autopilot with Splunk
40mInteractive Session - Live Code Contribution to Apache Flink
27mHow to be data-driven when you aren’t Netflix (or even if you are)
41mAkka support