Simulation blocked when more than 7 scenarios

[tpitz]

Hello,

I'm trying to execute a Simulation which contains 13 scenarios.
Firstly, I executed the test with 2 simulations, and everything worked as expected, but as soon as I tried to merge the 2 simulations into one, gatling won't start the test.
When I try to launch up to 7 scenarios in the setUp(), it works.
As soon as I put an 8th scenario in the setUp(), gatling stays pending and the simulation never starts.
No logs are written, even if DEBUG mode is enabled in logback.xml

I have to kill the process to make it end.

Below some informations about the code:

setUp(
    scnUpdateOMSSendToWorkingOrder.inject(constantUsersPerSec(1) during (60)),
    scnIncorrectOMSSendToOrcDeadletter.inject(constantUsersPerSec(1) during (60)),
    scnOMSWithoutPrepItemSendToWorkingOrder.inject(constantUsersPerSec(1) during (60)),
    scnOMSWithPrepItemSendToWorkingOrder.inject(constantUsersPerSec(1) during (60)),
    scnTryToSendMailWithoutFSID.inject(constantUsersPerSec(1) during (60)),
    scnReceivedSPLSendConfirmMail.inject(constantUsersPerSec(1) during (60)),
    scnSPLCancelMail.inject(constantUsersPerSec(1) during (60)),
    scnMetiSplClkOrderSendToMetiOMS.inject(constantUsersPerSec(1) during (60)),
    scnMetiSplClkOrderSendToMetiOMSWithDiscount.inject(constantUsersPerSec(1) during (60)),
    scnMetiRF604Cancel.inject(constantUsersPerSec(1) during (60)),
    scnMetiRF604Ready.inject(constantUsersPerSec(1) during (60)),
    scnMetiRF604Prepared.inject(constantUsersPerSec(1) during (60)),
    scnMetiRF604IncorrectToDLQ.inject(constantUsersPerSec(1) during (60))
  ).protocols(amqpConf)
    .assertions(
      forAll.successfulRequests.percent.gt(90),
      forAll.responseTime.percentile3.lte(200)
    )

All these scenarios are base on the function below:

def omsScenario(scenarioName: String, exchangeName: String, routeKey: String, replyQueue: String, action: String, filePathMessage: String): ScenarioBuilder = {
val scnMeti: ScenarioBuilder = scenario(scenarioName)
      .feed(correlationIds)
      .exec(
        amqp(scenarioName).requestReply
          .directExchange(exchangeName, routeKey)
          .replyExchange(replyQueue)
          .bytesMessage(RawFileBody(filePathMessage))
          .contentType("application/json")
          .headers("action" -> action, "correlationId" -> "${correlationId}")
          .priority(0)
      )
    scnMeti
  }

Here is the connection configuration:

val amqpConf: AmqpProtocolBuilder = amqp
    .connectionFactory(
      rabbitmq
        .host("my_endpoint")
        .port(5672)
        .username("user")
        .password("pw")
        .vhost("/my_vhost")
    )
    .usePersistentDeliveryMode
    .replyTimeout(900000)
    .matchByMessage(getMatcher)

Any help would be great. Thank you in advance.

Terry

[ehmkah]

Some thoughts:

• Could you please post your getMatcher method? maybe there is something strange.
• Is the order in which you add scenarios is relevant?
• Do you see something happening int the queue (RabbitMQ Managament GUI)?
• Are you able to reproduce this behavior with a small example project which you could share?

[tpitz]

1. Here is the method:

def getMatcher(message: AmqpProtocolMessage): String = {
    message.amqpProperties.getHeaders.get("correlationId").toString
  }

2. The order is not important. I notice it's not executed when it exceeds 7 scenarios only. When I split the scenarios into 2 separate simulation classes, it works ! The two simulations are executed sequentially, but this is not what I want. I want them to be grouped in the same report.
3. No activty at all in the queues !
4. You will find a repo here: https://github.com/tpitz/rabbit_one_simulation
   To make this work you will need to provide a local rabbitMQ environment.
   The image I use is: https://hub.docker.com/_/rabbitmq with tag: rabbitmq:3-management

Let me know if you encounter problems making it work locally.
Thank you in advance for your help.

[tpitz]

Hi, did you take a look for a possible solution ? :)

Thank you in advance