You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by "PAULO LOURENCO (RE-ESI-Workflow)" <pa...@esi.pt> on 2008/12/16 20:52:29 UTC

completely different time responses

Greetings!

 

First of all, allow me to introduce myself. My name is Paul and I have been working with JMeter, roughly for the past 2 months. It has been an interesting ride, especially after discovering the BeanShell component :-)

 

Anyway, let me get down to the important stuff. Today I discovered something that is worrying me to a certain degree and I'll try to explain:

-          I have a JMeter script that is being used to test the responsiveness of a web site of mine. The web site basically: 1) controls access using a login page; 2) presents a dynamically built list of tickets; 3) the user opens a ticket (a form, but doesn't process anything in the form); 4) closes the ticket.

-          My objective is to determine the time spent in each of these 4 steps.

 

I added various methods (just out of curiosity to ensure that the values are always the same no matter which component I use) to determine the time in each sampler: an Aggregate Report Listener, Simple Data Writer and I'm also using BeanShell (using the SampleResult.getTime(); method).

 

I'm executing the tests using two different approaches: via JMeter GUI and non-GUI JMeter call in the command line.

 

Now here's the interesting part: if I run the test from within JMeter GUI I receive two different sets of results, depending on whether the Aggregate Report is enabled or not. Every test I run with Aggregate Report Listener enabled gives me response times on average 2 seconds faster then with the Aggregate Report Listener Disabled. 

 

Has anyone experienced such behaviour? Obtaining considerably different response times depending on whether a particular component is active or not...

 

I'm completely available to better explain this scenario in case I haven't made it clear.

 

Thanks in advance for any answer on this one!

 

Regards,

Paulo Lourenço