Requests per second incorrect?

316 views
Skip to first unread message

Garrett

unread,
Apr 3, 2012, 10:12:10 AM4/3/12
to gat...@googlegroups.com
Doing some simple test (a scenario that makes one http request vs a scenario that makes 3 http requests) I've notice that requests per second seems to be counting the scenario start and scenario end events (the scenario with 1 http request comes in with 3 requests on the requests per second chart. The scenario with 3 http requests comes in with 5 requests.) Is this intentional? It seems to be really misleading when you are using short scenarios as it highly inflates the number of requests per second.

Romain Sertelon

unread,
Apr 3, 2012, 11:14:36 AM4/3/12
to gat...@googlegroups.com
Hi,

Which version of Gatling do you use?

If it actually counts the start and end events, it is not an expected behaviour, of course.

Thanks for your feedback.

Cheers,
BluePyth


On Tue, Apr 3, 2012 at 4:12 PM, Garrett <garrett...@gmail.com> wrote:
Doing some simple test (a scenario that makes one http request vs a scenario that makes 3 http requests) I've notice that requests per second seems to be counting the scenario start and scenario end events (the scenario with 1 http request comes in with 3 requests on the requests per second chart. The scenario with 3 http requests comes in with 5 requests.) Is this intentional? It seems to be really misleading when you are using short scenarios as it highly inflates the number of requests per second.



--
Romain SERTELON / BluePyth
Blog ~ @BluePyth ~ +BluePyth ~ Github ~ LinkedIn ~ DoYouBuzz

Garrett

unread,
Apr 3, 2012, 11:22:16 AM4/3/12
to gat...@googlegroups.com
I'm using 1.1.1

Romain Sertelon

unread,
Apr 3, 2012, 11:31:03 AM4/3/12
to gat...@googlegroups.com
Okay, we'll take a look at that, it will be fixed in 1.1.2 coming up on Thursday ;)

The issue is here : https://github.com/excilys/gatling/issues/455

Cheers,
BluePyth

On Tue, Apr 3, 2012 at 5:22 PM, Garrett <garrett...@gmail.com> wrote:
I'm using 1.1.1

Stéphane Landelle

unread,
Apr 3, 2012, 11:45:32 AM4/3/12
to gat...@googlegroups.com
Hi Garrett,

You might be right, let me check this.

Cheers,

Steph

2012/4/3 Garrett <garrett...@gmail.com>

Stéphane Landelle

unread,
Apr 3, 2012, 1:12:37 PM4/3/12
to gat...@googlegroups.com
Fixed in master (will be in 1.1.2).

Thanks for reporting!

Cheers,

Steph

2012/4/3 Stéphane Landelle <slan...@excilys.com>

Manoj KS

unread,
Apr 16, 2014, 8:25:31 AM4/16/14
to gat...@googlegroups.com
Hello Steph,

In version 1.5.5 I am still getting same behavior

Here it goes
  • When I perform test with only 1 request, the value 'Req/s' shows as 7
  • With 3 requests, the value is 25
Please do the needful

Regards,
Manoj Kapuganti

Stéphane Landelle

unread,
Apr 16, 2014, 11:15:56 AM4/16/14
to gat...@googlegroups.com
The rps is an average value there: number of requests / run duration.
As you have only 1 request that last less than 1 sec, here's what you get.
From my point of view (open to discussion of course), the result is expected: during your run you did have a burst of 7rps.


--
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.

Reply all
Reply to author
Forward
0 new messages