Non blocking requests

67 views
Skip to first unread message

Janek Lasocki-Biczysko

unread,
May 29, 2015, 5:45:34 AM5/29/15
to gat...@googlegroups.com
Hi folks,

First off, Gatling is a great tool - I'm really enjoying using it - thank you Gatling Team!

I was wondering if there's a way to configure my scenario to fire off (some) http requests in a non blocking fashion. Meaning that if the following scenario is run:

  var scn = scenario("Test Blocking")
   
.exec(http("1").get("/?b=1"))
   
.execNonBlocking(http("2").get("/?b=1").check(jsonPath("$..message").exists))
   
.exec(http("3").get("/?b=2"))
   
.exec(http("4").get("/?b=1"))
   
.execNonBlocking(http("5").get("/?b=3"))
   
.exec(http("6").get("/?b=4"))
   
.exec(http("7").get("/?b=1"))
   
.exec(http("8").get("/?b=5"))

(using this imagined DSL "execNonBlocking" addition) when request "2" gets sent, request "3" gets sent immediately after, without waiting for a response from "2", but "2" still gets a check (jsonPath) eventually.
Is anything like this possible?

Thanks!

Stéphane LANDELLE

unread,
May 29, 2015, 7:16:39 AM5/29/15
to gat...@googlegroups.com
Thanks for your kind words.

The thing was asynchronous flow, be it resources, websockets, sse and polling, is that they live in different state spaces, and you need to reconciliate state at some point, ie you have to define your "eventually". And more than that, you probably want to be notified of when "eventually" happen.

So no, it's not currently possible.

This is more of an API design issue. Ideas welcome.

Stéphane Landelle
Lead developer


--
You received this message because you are subscribed to the Google Groups "Gatling User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gatling+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Janek Lasocki-Biczysko

unread,
May 29, 2015, 9:06:58 AM5/29/15
to gat...@googlegroups.com
Thanks for the reply - I see what you mean by the "eventually" and how it becomes a non trivial problem. 
Reply all
Reply to author
Forward
0 new messages