Reactive Vert.x
5 likes1,595 views
The document discusses a polyglot approach using event-driven architectures with technologies like Erlang, Node.js, and Vert.x. It highlights concepts such as non-blocking IO, scalable messaging, and callback management in reactive programming. Key points include the importance of not blocking the event loop and the integration of various programming languages in a comprehensive framework.
Reactive Vert.x
- 1. REACTIVE
- 2. = Erlang + Node.js + Polyglot ―――――――――――― JVM http://vertx.io
- 5. Scalable Distributed event-bus Location transparency Clustered in-memory cache
- 6. Resilient Module and verticle isolation Worker processes Automatic HA failover
- 7. Responsive “Don't block the event loop!” Time-outs Event-bus “back pressure”
- 13. Callback Hell eventBus.<JsonObject> sendWithTimeout(metricsAddress, METERS_BUS_REQUEST, 1000L, ar1 -> { JsonObject metersData = ar1.succeeded() ? ar1.result().body() : new JsonObject(); eventBus.<JsonObject> sendWithTimeout(metricsAddress, HISTOGRAMS_BUS_REQUEST, 1000L, ar2 -> { JsonObject histogramData = ar2.succeeded() ? ar2.result().body() : new JsonObject(); processData(metersData.mergeIn(histogramData)); }); });
- 14. Callback Help Observable<JsonObject> observeMeters = eventBus.<JsonObject, JsonObject> send(metricsAddress, METERS_BUS_REQUEST) .map(msg -> msg.body()) .timeout(1L, SECONDS, from(new JsonObject())); Observable<JsonObject> observeHistograms = eventBus.<JsonObject, JsonObject> send(metricsAddress, HISTOGRAMS_BUS_REQUEST) .map(msg -> msg.body()) .timeout(1L, SECONDS, from(new JsonObject())); zip(observeMeters, observeHistograms, (jo1, jo2) -> jo1.mergeIn(jo2)) .subscribe(data -> processData(data));