(too) good results?

[Rob Brandt <rb...@csd-bes.net>]

I've now been through my first round of testing, and the results are 
better than we expected.  Rather than celebrate, I am suspicious.  Are 
there elements to this process that aren't realistic?  I'm thinking in 
terms of images, for example.  Our site isn't particularly image heavy, 
but it is a CMS and many images are retrieved via php functions rather 
than linked to directly on the file system.  Do the tests reflect all 
these functions being called?

Rob

---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org

Re: (too) good results?

[sebb <se...@gmail.com>]

On 10 December 2010 23:25, Rob Brandt <rb...@csd-bes.net> wrote:
> I've now been through my first round of testing, and the results are better
> than we expected.  Rather than celebrate, I am suspicious.  Are there
> elements to this process that aren't realistic?  I'm thinking in terms of
> images, for example.  Our site isn't particularly image heavy, but it is a
> CMS and many images are retrieved via php functions rather than linked to
> directly on the file system.  Do the tests reflect all these functions being
> called?

Cannot say for sure without knowing the system and what you have put
in the tests.
However there are some general checks you can make.

Can you get access logs for the server?

If so, then compare to see whether JMeter accesses files in the
approx. same proportions as browsers.

Are the JMeter sample times consistent with browser page load times?

Have you experienced poor (or reduced) performance on the system (not
using JMeter)?
If so, what happens if you simulate the load with JMeter?

JMeter will never produce exactly the same load as a browser, but then
nor will two different browsers produce exactly the same load.
But you can ensure the the JMeter sample rate and mix of samples are
typical for the system.

>
> Rob
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org