CODEWITHSUNDEEP
scalaasynchronousplayframework-2.0
Frank
Frank

Reputation: 13

Play 2.0 forkjoin thread count high

I'm currently evaluating play 2.0 framework in async scenario. I did an experiment as below:

def test = Action {
  val futureResponse: Future[play.libs.WS.Response] = scala.concurrent.Future {
    val remoteCall = WS.url("http://127.0.0.1:8080/slowweb/SlowServlet1").get();
    remoteCall.get(); 
  }
  Async {
    futureResponse.map(response => Ok("Got result: " + response.getBody()))
  }
}

The ":8080/slowweb/SlowServlet1" is a simulation of a slow web service call, it takes 2s to return.

Then I run jmeter against the endpoint ":9000/test". I observed the following behaviour: If I run 50 threads in jmeter, the forkjoin pool size is 50; if I run 200 threads in jmeter, the forkjoin pool size is 200.

It looks to me the thread pool size is the same as the concurrent request size. What I expect is the thread pool size should be much smaller if using async approach.

Have I done something wrong?

Thanks

Upvotes: 1

Views: 152

Answers (1)

ndeverge
ndeverge

Reputation: 21564

I don't know if it is related to your error, but your code could be simplified to this:

def test = Action {
   Async {
     val futureResponse: Future[play.libs.WS.Response] = 
        WS.url("http://127.0.0.1:8080/slowweb/SlowServlet1").get();

     futureResponse.map(response => Ok("Got result: " + response.getBody()))
   }
}

Upvotes: 1

Related Questions