Said Olano

Posted on Aug 16

Play Framework: Reactive Web Development for the Modern JVM (2026-08-16 18:34)

#playframework #scala #reactive #webdev

Play Framework: Reactive Web Development

The Play Framework is a high-productivity web framework for the JVM that embraces reactive principles from the ground up. Built on top of Akka and Netty, Play enables developers to build scalable, non-blocking web applications in both Java and Scala. In this post, we'll explore what makes Play reactive and how to leverage its capabilities.

Why Reactive?

Traditional web frameworks assign one thread per request. Under heavy load, this model consumes memory and thread resources quickly. Reactive frameworks like Play use a small pool of threads to handle many concurrent connections through non-blocking I/O.

The four principles of reactive systems apply directly to Play:

Responsive: The system responds in a timely manner.
Resilient: The system stays responsive in the face of failure.
Elastic: The system stays responsive under varying workload.
Message-driven: Components communicate via asynchronous messages.

Getting Started

Create a new Play project using sbt:

sbt new playframework/play-scala-seed.g8

The generated project follows a clear structure with app/controllers, app/models, app/views, and a conf/routes file that maps URLs to controller actions.

Defining Routes

Routes are declared in conf/routes using a simple DSL:

GET     /               controllers.HomeController.index
GET     /users/:id      controllers.UserController.show(id: Long)
POST    /users          controllers.UserController.create

Play generates a type-safe reverse router, so you can reference routes in your code without hardcoding URLs.

Writing Asynchronous Actions

The heart of Play's reactive model is the Action. Instead of blocking, controllers return a Future[Result], allowing the framework to free threads while awaiting results.

import javax.inject._
import scala.concurrent.{ExecutionContext, Future}
import play.api.mvc._

@Singleton
class UserController @Inject()(
  cc: ControllerComponents,
  userService: UserService
)(implicit ec: ExecutionContext) extends AbstractController(cc) {

  def show(id: Long) = Action.async { implicit request =>
    userService.findById(id).map {
      case Some(user) => Ok(views.html.user(user))
      case None       => NotFound("User not found")
    }
  }
}

By using Action.async, the request thread is released immediately and only re-engaged when the Future completes.

Non-Blocking HTTP Calls

Play ships with an asynchronous HTTP client (Play WS) that integrates naturally with the reactive model:

import play.api.libs.ws._

class WeatherController @Inject()(
  ws: WSClient,
  cc: ControllerComponents
)(implicit ec: ExecutionContext) extends AbstractController(cc) {

  def forecast(city: String) = Action.async {
    ws.url(s"https://api.weather.com/v1/$city")
      .withRequestTimeout(5.seconds)
      .get()
      .map(response => Ok(response.json))
  }
}

This avoids blocking a thread while waiting for the remote service to respond.

Streaming with Akka Streams

For handling large payloads or real-time data, Play integrates with Akka Streams to process data incrementally rather than loading everything into memory.

import akka.stream.scaladsl.Source
import akka.util.ByteString

def stream = Action {
  val source = Source(1 to 1000)
    .map(i => ByteString(s"line $i\n"))
  Ok.chunked(source)
}

This is ideal for exporting reports, proxying files, or feeding event streams to clients.

WebSockets

Real-time bidirectional communication is a first-class citizen. Play models WebSockets as Akka Streams flows:

import play.api.mvc.WebSocket
import akka.stream.scaladsl.Flow

def socket = WebSocket.accept[String, String] { request =>
  Flow[String].map(msg => s"echo: $msg")
}

Dependency Injection

Play uses Guice for dependency injection by default. Components are wired via constructor injection with the @Inject annotation, promoting testability and clean separation of concerns.

Testing

Play provides excellent testing support through ScalaTestPlus:

class UserControllerSpec extends PlaySpec with GuiceOneAppPerTest {
  "UserController" should {
    "return 404 for unknown user" in {
      val request = FakeRequest(GET, "/users/999")
      val result = route(app, request).get
      status(result) mustBe NOT_FOUND
    }
  }
}

Best Practices

Never block the default execution context. Use a dedicated dispatcher for blocking operations like JDBC calls.
Propagate ExecutionContext implicitly rather than using the global one.
Prefer streaming for large or unbounded data sets.
Leverage the type-safe router to avoid runtime URL errors.

Conclusion

Play Framework delivers a productive, fully reactive development experience on the JVM. By building on Akka and Netty, it lets you write applications that scale gracefully under load while keeping your code expressive and maintainable. Whether you're building REST APIs, real-time dashboards, or streaming services, Play's reactive foundation provides the tools you need.

DEV Community