Workshop: Event-sourced system through Reactive Streams
- 1. Event-sourced system through Reactive Streams Workshop
- 2. Agenda • 18:30 presentation introduction • 19:00 exercise project reactor intro • 19:30 short break • 19:40 presentation event-sourcing • 19:50 workshop • 21:00 Closing & Networking
- 4. Reactive Programming Belgium Your non-blocking resource in a reactive world “We believe that RP is a good fit for the challenges that current applications/architectures entail”
- 5. Reactive Programming Belgium Your non-blocking resource in a reactive world • Research • Share • Blog http://www.reactiveprogramming.be • Meetups https://www.meetup.com/Reactive-Programming-Belgium/ • Presentations • ...
- 6. Reactive Programming Belgium Your non-blocking resource in a reactive world • Want to help? • Writing articles (blog) • Ideas/speakers/presenting… • Prepare workshops • Ad-hoc experiments • ...
- 7. Reactive Programming Belgium Your non-blocking resource in a reactive world • Workshop : ‘Event-sourced system through Reactive Streams’
- 9. Reactive programming Applications today • Applications today • Increased magnitude of scale • Big data (IoT, devices, …) • Realtime • Ex. Flight recorder 1 sec of flight = 1.5 TB of data • Re-think the way we create applications
- 10. Reactive programming Origin • Origin @ Microsoft • Erik Meijer • Reactive programming framework (Reactive Extensions) for .NET
- 11. Reactive programming Yet another programming model? • Reactive Something that responds to stimuli and actions • Reactive Programming • Rehashed programming model that emphasizes on scalability & response • A way to build a scalable architecture that’s resilient and quick to react to stimuli • Ex. spreadsheet
- 12. Reactive programming Data pipeline • CRUD • database -> function/process -> database • Data pipeline • data -> function -> data -> function -> data • Data transformation instead of data manipulation
- 13. Reactive programming Programming model – Reactive Streams • Publisher (emit data) • Subscriber (receive data) • Subscription (~ session / context info) • Processor Flux.range(1, 10) .filter(i -> isEven(i)) .map(i -> i * 2) .subscribe(System.out::println); Project reactor
- 14. Reactive programming Programming model – Reactive Streams • Publisher onComplete() onError() next()
- 15. Reactive programming Programming model – Reactive Streams • Backpressure
- 16. Reactive programming Programming model – Reactive Streams • Java 9 Reactive Streams • API (Publisher, Subscriber, Subscription, Processor) = Reactive streams spec • Frameworks/implementations • RxJava (Netflix), project reactor, Akka-Streams, … • RxJava 2 • Observable vs. Flowable >> both Publishers • Project reactor • Mono (0 - 1), flux (0 ..) >> both Publishers • Reactive toolkit • Vert.x
- 17. Reactive system
- 18. Reactive System Reactive manifesto • Elastic • Message-driven • Responsive • Resilience • Failure = first-class citizen
- 19. Reactive System 1 JVM JVM data -> function -> data -> function -> data JVM
- 20. Reactive System Distributed system JVM data -> data -> function -> data -> data-> function -> data JVM JVM JVM JVM JVM
- 21. Reactive System Distributed system • Between jvm's • http (spring webclient) • gRPC, RSocket • Messaging • Between jvm & database • non-sql vs sql • Between jvm & UI • RxJS
- 22. Project reactor
- 23. Project reactor Flux & mono • Flux = Publisher <0 .. N> • Mono = Publisher <0 – 1>
- 24. Project reactor Getting started • Exercises • Publisher / subscriber • onComplete • Error handling • Backpressure
- 25. Project reactor Getting started • https://github.com/vanseverk/projectreactorintro • IDE • IntelliJ : mark ‘test’ folder as ‘test sources root’ • TODO 01 – 06 • Extra (backpressure) • TODO 07 - 08
- 26. Event sourcing
- 27. Event Sourcing Intro • Persist events instead of domain objects • Immutable event-log • Replaying events can help development, debugging • No need to lock DB tables • Can be a perfect fit in event-driven architectures • Can be designed to handle temporary partitioning • Microservices handle (or replay) events and store what they require • Eventual consistency • Reactive Programming + Events -> Reactive Event Streaming
- 28. Event Sourcing Intro • Introducting: NATS • https://nats.io/ • High performance messaging system for cloud native applications, IoT messaging, and microservices architectures • High performance, lightweight, many extensions
- 29. Event Sourcing Intro • Introducing: NATS Streaming • Event Sourcing + Streaming including replay from time or sequence number • Durable subscriptions for transient clients • Persistent/guaranteed (at least once) message delivery • Pub/Sub ; Request/Reply; Queueing to share work
- 30. Workshop
- 32. Reactive system: measurements Practical • https://github.com/vanseverk/anomalydetection-nats • 2 branches • master = exercise (TODO) • solution • workshop.pdf