RE: Functional Mode Checkbox and quick and dirty Functional testi ng

["Sosnowski, Andrew P [IT]" <an...@citigroup.com>]

Hi Mike

Sorry about the delay.
A little more info about the incomplete html page returned in the jtl file
while using Functional Testing Mode:

I.e when I view the html Using the View Results, it displays the complete
page but when I look in the jtl file, only part of the html is there.
The html ends abruptly at the beginning of a table and this is consistent
when I try the same page several times.

Any ideas?

Thanks
Andy

-----Original Message-----
From: Sosnowski, Andrew P [IT] 
Sent: Tuesday, August 27, 2002 10:25 AM
To: 'mstover1@apache.org'; Sosnowski, Andrew P [IT]
Subject: RE: Functional Mode Checkbox and quick and dirty Functional
testi ng


Sounds good.
One more thing I've discovered in my testing; the html for one of my pages
is not being fully loaded into the log file; is there any limitation to the
storing of html that you know of?

(Another curious thing is that the html saved changes all the html angle
brackets to < and > Not a problem just curious)

The page in question does display properly in both Netscape and IE.

This would limit the usefulness of this new feature; there was no error
detected but I did not have an assertion for </html>

Thanks
Andy

-----Original Message-----
From: Mike Stover [mailto:mstover1@apache.org]
Sent: Tuesday, August 27, 2002 9:26 AM
To: Sosnowski, Andrew P [IT]
Subject: RE: Functional Mode Checkbox and quick and dirty Functional
testi ng


Here's kinda what I imagined:

The visualizer looks like the View Results Tree (ie, there's a tree that
holds results or some 
sort, you click on a node, you see the details below).

It also has 2 additional fields where the user can specify the name of two
requests in the test.  
The visualizer will then create diffs against these two requests, and from
the diff info, create 
the nodes in the tree.  Thus, each node in the tree isn't a normal result
but rather, the diff 
between the two specified results.

I don't know exactly what your needs are, so you make it whatever you want,
of course.  The 
visualizer could be dependent on a "diff" command being on the system, and
windows users 
can always install cygwin (it's the first thing I do on any windows box).
It might also be worth a 
few minutes to search the web for a diff Java implementation.

-Mike

On 27 Aug 2002 at 7:17, Sosnowski, Andrew P [IT] wrote:

> Hi Mike
> 
> This seems to be working for me pretty well; so far I am using perl to
parse
> out the html files, massage them and compare them but that could be done
> easily with Java regular expressions as well.
> 
> You mention creating a visualizer; I'd like to perhaps build this if you
> feel it fits into jMeter plans; what do you think it should look like? Any
> pointers?
> I think there should be an additional input field to specify the run that
is
> to be compared.
> The actual result would consist of the output of the diff command on the
> html files within the 2 logs. However I'm not sure if there is a
> cross-platform version of diff.
> 
> Thanks in advance
> Andy
> 
> -----Original Message-----
> From: Mike Stover [mailto:mstover1@apache.org]
> Sent: Tuesday, August 20, 2002 5:09 PM
> To: JMeter Users List
> Subject: Re: Functional Mode Checkbox and quick and dirty Functional
> testing
> 
> 
> First, you have noticed the one and only difference that choosing
> "functional testing" makes.
> 
> Second, that sounds like an interesting idea - to make a visualizer that
> compares the 
> functional differences between two test runs?  I think your proposed use
of
> such a tool is an 
> excellent idea.
> 
> -Mike
> 
> On 20 Aug 2002 at 15:21, Sosnowski, Andrew P [IT] wrote:
> 
> > Hi Mike,
> > 
> > Can you give me an idea of what is the intent behind the Functional Mode
> > testing checkbox on the front page?
> > 
> > 
> > This was my plan for a quick and dirty functional test:
> > We have a Staging environment and a production environment for our app.
> > When we make a new release of our app I would like to capture and
compare
> > the html responses of running a test script with the new version of the
> app
> > on the Staging Environment and the old version still in Production.
> > Then I would compare the 2 versions of html and make sure that nothing
> > changed that was not expected to change.
> > Of course I have to ignore Referrer Headers values and time-stamps and
> > various cookies that are different on the different servers.
> > 
> > When I tried using this mode I did find the html within <binary> tags in
> the
> > log file.
> > It looks like it might work as a quick and dirty way to compare.
> > 
> > Perhaps it might be worth building in some additional support for this
> kind
> > of testing into jMeter? For example store html cleaned of timestamps in
a
> > file separate from the log file (or jtl file)
> > I might be willing to do this if others would find it useful. Also any
> ideas
> > of how to make it more useful and general are appreciated
> > 
> > What do you think?
> > 
> > Andy
> > 
> > 
> > --
> > To unsubscribe, e-mail:
> <ma...@jakarta.apache.org>
> > For additional commands, e-mail:
> <ma...@jakarta.apache.org>
> > 
> 
> 
> 
> --
> Michael Stover
> mstover1@apache.org
> Yahoo IM: mstover_ya
> ICQ: 152975688
> 
> --
> To unsubscribe, e-mail:
> <ma...@jakarta.apache.org>
> For additional commands, e-mail:
> <ma...@jakarta.apache.org>



--
Michael Stover
mstover1@apache.org
Yahoo IM: mstover_ya
ICQ: 152975688

--
To unsubscribe, e-mail:   <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>