LoadRunner and Jmeter suffer from similar problems (as far as I know) -
though LoadRunner has some support
http://www.wilsonmar.com/ajax_rec.htm.
However you are usually trying to simulate two things
a. The behavior for a single user (perhaps to figure out deadlocks or
parallel access in the same session ) - In this case a tool that drives the
browser like Selenium is a better choice than JMeter/LoadRunner. This covers
most functional tests too
b. The behavior under load
b.1 The application uses AJAX but usually in response to a user
action like the click of a button. - Even though this is AJAX , the requests
are actually serial. The page loaded . The User clicked a button. Some part
of the page refreshed. The user did something else. In this case
JMeter/LoadRunner should work as usual.
b.2 AJAX calls can actually be made in parallel (e.g. Auto Suggest
search terms based on what the user has currently typed) - However in most
of these cases the Server response time depends on the total number of
concurrent requests it receives and you can usually approximate this that if
you had 10 users making 3 parallel calls each , its almost the same as 30
users making 1 call. You cant accurately simulate this anyway - My typing
speed isnt the same as your typing speed so you have to make a guess as to
how many requests would have been made in parallel for a user (using a
browser tool) for the above example. So long as requests are independent you
can still use Jmeter by adjusting the number of threads. You might see a
slightly worse performance because you have increased your session
footprint.
b.3 Cases where you application does make multiple AJAX calls in
parallel where the requests are not independent(Ive yet to come across such
a scenario). - Jmeter probably wont be able to test these things without
writing some fancy code
Id probably take a look at Selenium Grid (or equivalent)
regards
deepak
On Thu, Oct 14, 2010 at 1:24 AM, Sangeetha M <[email protected]>wrote:
> So what can I say compared to loadrunner is JMeter a better choice to go
> for Ajax applications or not ?
> Does both have the same strategy to handle the Ajax requests? Or the
> JMeter effort and constraints are more?
>
>
> Regards
> Sangeetha M
>
> -----Original Message-----
> From: Felix Frank [mailto:[email protected]]
> Sent: Wednesday, October 13, 2010 5:19 PM
> To: JMeter Users List
> Subject: Re: Jmeter with ajax applications
>
> > I have seen in the archive lists it says JMeter can record Ajax
> > application using http proxy server. But my doubt are
> >
> > 1. " whether JMeter is able to simulate the asynchronous requests
> which
> > inherently meant browser just sends the request in pipeline/parallel
> > without waiting for the response from the server?"
>
> Jmeter will make all requests serially. Depending on your application,
> this may not be correct at all.
>
> > 2. Jmeter can simulate clientside "JavaScript's"?
>
> A recorded plan will behave just as the client scripts behaved *during
> recording*.
> In many cases, you will want to rebuild the client side logic in the
> Jmeter test plan, but the recorded plan cannot do this for you.
> You have to do this yourself, using PostProcessors mainly, like
> RegexExtractor.
>
> > 3. How it will simulate the frequency at which the Ajax request is
> sent
> > ?
>
> You choose the frequency at which Jmeter operates. If you don't insert
> timers, Jmeter will use the highest frequency it can come up with (i.e.,
> send each request as soon as the previous one is complete).
>
> HTH,
> Felix
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>
> Information transmitted by this e-mail is proprietary to MphasiS, its
> associated companies and/ or its customers and is intended
> for use only by the individual or entity to which it is addressed, and may
> contain information that is privileged, confidential or
> exempt from disclosure under applicable law. If you are not the intended
> recipient or it appears that this mail has been forwarded
> to you without proper authority, you are notified that any use or
> dissemination of this information in any manner is strictly
> prohibited. In such cases, please notify us immediately at
> [email protected] and delete this mail from your records.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>
