Mailing-List: contact user-help@jmeter.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "JMeter Users List" <user@jmeter.apache.org>
Received-SPF: pass (athena.apache.org: domain of shmulikk@gmail.com designates
 209.85.214.171 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <941709348625F140BE484DC832C50718205D39B5@BY2PRD0510MB388.namprd05.prod.outlook.com>
References: 
 <941709348625F140BE484DC832C50718205D37D5@BY2PRD0510MB388.namprd05.prod.outlook.com>
	<CAOGo0VZ4BJwSL-FWbzv0n23Xgtx9jMfm5D_-ctThBtC_8zZZHQ@mail.gmail.com>
	<941709348625F140BE484DC832C50718205D39B5@BY2PRD0510MB388.namprd05.prod.outlook.com>
Date: Fri, 14 Sep 2012 23:10:21 +0300
Message-ID: 
 <CAHnoMfpFchAfn-hCEnsr6Cxb8o_g2me9W2=Z0oUcC1GQN+aV-A@mail.gmail.com>
Subject: RE: GUI capture question
From: Shmuel Krakower <shmulikk@gmail.com>
To: JMeter Users List <user@jmeter.apache.org>
Content-Type: multipart/alternative; boundary=e89a8f647bf57e83ae04c9af03ae

--e89a8f647bf57e83ae04c9af03ae
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi
If you dont want to create many samplers to simulate all required requests
try googling for selenium sampler for jmeter.

There was someone in this mailing list which build it and with selenium HQ
you can do the thing you are looking for; creating a script which interacts
with the GUI instead with the underlying http traffic.

I never used it but worth a try.
On the other hand you should also read about selenium limitations.

Shmuel.
=D7=91=D7=AA=D7=90=D7=A8=D7=99=D7=9A 2012 9 13 23:32, =D7=9E=D7=90=D7=AA "B=
OLB (Bohdan L Bodnar)" <BOLB@panduit.com>:

> I must not have been clear in the request.  The problem is that the login
> procedure consists of almost 15 calls from the GUI to the server and when
> the web page opens there are about 30 calls to the server BEFORE the logi=
n
> procedure starts.  So, buried among all these calls is the start and end
> time AS SEEN BY THE USER.  After the login there are about a dozen other
> calls (the number varies depending on the user's credentials) before
> another activity I want to flag occurs.  I want the proxy to capture the
> user's activities and have jmeter mark the start and end time of each
> predetermined activity.  Summarizing:
>
> 1). GUI opens with numerous calls to the server.
> 2). User logs in - this generates numerous other calls.  The number of
> calls depends on the credentials
> 3). Login procedure terminates; this generates several other calls.
> 4). I want to mark when login starts (from user's perspective) and when i=
t
> finishes (from user's perspective).
>
> This is NOT a trivial GUI; hence, I want to know whether certain activity
> locations can be marked.  I know commercial load generator tools have thi=
s
> capability (e.g., some of IBM's Rational tools).  The alternative is to d=
ig
> through the calls to understand the GUI/server interaction.
>
> Bo
>
>
> -----Original Message-----
> From: sebb [mailto:sebbaz@gmail.com]
> Sent: Thursday, September 13, 2012 3:20 PM
> To: JMeter Users List
> Subject: Re: GUI capture question
>
> On 13 September 2012 14:32, BOLB (Bohdan L Bodnar) <BOLB@panduit.com>
> wrote:
> > I've a problem I'm hoping someone on the list may assist me with.
>  Here's the situation:
> >
> > I'm running jmeter's http proxy to capture user activity.  The activity
> consists of opening a web page, a login, some manipulations, and then a
> logout.  The data is exported to a .csv file.  The file includes each
> method called, latency time, and elapsed time and some other parameters. =
 I
> want to be able to mark the start and end of each activity, as seen by th=
e
> user.  For example, during a login, the user finishes entering the name a=
nd
> password into the GUI and then hits the "enter" key.  I want to have a
> "mark" showing the moment the "enter" key is hit to the moment
> authentication is completed, as seen by the user.  Examining the .csv fil=
e
> is difficult, as the start and end times require an in-depth understandin=
g
> of the methods that are called during these procedures (I don't have this
> knowledge).
>
> The start time is when "enter" is pressed, i.e. just before the request i=
s
> sent to the server.
> Latency finishes when the first part of the response is received.
> Elapsed time finishes when the complete response has been received.
>
> > The best analogy I can think of is the following:  in embedded systems,
> this is equivalent to having a logic analyzer capture the start and end
> time of a function call.
> >
> > Any assistance is greatly appreciated.
> >
> > Best regards,
> >
> > Bo
> >
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> > For additional commands, e-mail: user-help@jmeter.apache.org
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
>

--e89a8f647bf57e83ae04c9af03ae--