Il 29/08/2013 09.58, Manish Sapariya ha scritto:
Chaityna/Kirk,
For a customer, I wrote a script that allows use of Jmeter scripts to Monitor
the performances of web applications.
It is generic, so you just have to configure the JMX file you want to test.
The internal logic is a little bit tricky, because you have to take in account of response timing, HTTP response and content
assertions to set a PASS/ WARN /FAIL response.
Timing can be monitored either or specific pages AND for the overall script.
I found this approach is quite effective on the timings side, but it has weaknesses on the maintenance, as you have to update the
script while software updates.
Also, it is not that easy to cope with heavy Javascript-automated pages.
For that case, I integrate a browser based approach (I use Watir/ webdriver)
that can better fulfil the purpose.
In that case, the architecture is similar to WebPageTest.
I'm planning to release my script as OS..
Please contact me privately if you are interested.
best regards
Sergio Boso
Thanks for the inputs.
As far as possible I wanted to keep the assertions only on the client
side to keep the test setup simple. I had originally thought of, as Kirk
suggested, to assert based on the access log timings, but the test setup
will be complicated to get up running and maintaining.
I though it would have been possible to do some kind of assertion,
based on the response time for each request in jMeter, which I
believe is reflection of the server introduced latencies as jMeter not
attempting to render any of it.
Thanks again for your inputs.
By the way I am also investigating the offline webpagetest.
https://sites.google.com/a/webpagetest.org/docs/private-instances/releases/webpagetest-2-12
Thanks and Regards,
Manish
kPoint wins *Global eLearning
Award<http://www.kpoint.com/kpoint-wins-learntech-global-e-learning-award/>
* in “Learning Technologies Solution”!
On Thu, Aug 29, 2013 at 12:46 PM, Kirk Pepperdine <[email protected]
wrote:
I think this thread is good in pointing out the limits of our tools... we
need to know them so we can use them more effectively.
Regards,
Kirk
On 2013-08-29, at 9:14 AM, chaitanya bhatt <[email protected]>
wrote:
OK kindly ignore my previous post.
@Kirk: Thanks for asking that question!
On Wed, Aug 28, 2013 at 11:58 PM, Manish Sapariya <[email protected]
wrote:
One clarification, I indeed want to measure the time server takes to
serve
pages
and not how much time browser take to render them.
Thanks and Regards,
Manish
kPoint wins *Global eLearning
Award<
http://www.kpoint.com/kpoint-wins-learntech-global-e-learning-award/
* in “Learning Technologies Solution”!
On Thu, Aug 29, 2013 at 12:17 PM, Kirk Pepperdine <
[email protected]
wrote:
Jmeter cannot be used for determining page load times.
Indeed, and page load/rendering times is a separate investigation IME
and
a number of issues tend to be browser specific.
Regards,
Kirk
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
--
Ing. Sergio Boso
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
