Software Testing - Performance test a desktop client server application???

Author

Performance test a desktop client server application???

RaleighDan

2006-07-18, 8:03 am

I have done performance testing on http/web based applications, but
would now like to do perfor/load testing on a desktop client/server
based application. I have not been able to find any tools that do this.
Am I missing something here?

Ideally it would be nice to see what response times are like if 50
"users" login to the application.

The client is a C# .net application that connects to a application
server and DB server.

Any ideas?

2006-07-18, 7:06 pm

RaleighDan wrote:
> I have done performance testing on http/web based applications, but
> would now like to do perfor/load testing on a desktop client/server
> based application. I have not been able to find any tools that do this.
> Am I missing something here?
>

Hi there,

Can you tell me what tools did you use to test http/web based
applications?
I'm also testing http/web based applications, and i'm searching for
tools to that.

Thanks,
RC

info@e-valid.com

2006-07-18, 7:06 pm

> RaleighDan wrote:
>
> ...snip...
>
> Can you tell me what tools did you use to test http/web based
> applications?
> I'm also testing http/web based applications, and i'm searching for
> tools to that.
>
> Thanks,
> RC

For testing ANY kind of web enabled application the best
possible tests arise from testing with an actual browser as
the test driver.

eValid is a complete web testing and analysis solution suite
built into a full-featured browser.

Because eValid is not a plugin, not a proxy, not a wrapper,
and not a passive tap on the HTTP/S protocol, the tests you run
have full power and capability to do every kind of browser
based action.

You can download an evaluation copy of the eValid solution
from:

http://www.e-Valid.com/Products/Dow...tml?status=FORM

Complete details about the eValid web analysis and testing
suite can be found at:

http://www.e-Valid.com

Michael Bolton

2006-07-19, 7:05 pm

info@e-valid.com wrote:
>
> For testing ANY kind of web enabled application the best
> possible tests arise from testing with an actual browser as
> the test driver.

How can you demonstrate the truth of that claim?

Wouldn't it be more accurate to say that "for testing MANY kinds of
Web-enabled applications, many excellent tests depend upon testing with
an actual browser as the test driver"?

To me, this would be more persuasive marketing because it would
recognize the value of the other tests that one could perform, such as
unit tests, integration tests behind the browser, and so on. And I bet
eValid can help with those too, in some contexts--right?

When you say "best possible" anything, you lose credibility with
serious testers.

---Michael B.

info@e-valid.com

2006-07-20, 7:03 pm

Michael Bolton wrote:
> info@e-valid.com wrote:
>
> How can you demonstrate the truth of that claim?
>
> Wouldn't it be more accurate to say that "for testing MANY kinds of
> Web-enabled applications, many excellent tests depend upon testing with
> an actual browser as the test driver"?
>
> To me, this would be more persuasive marketing because it would
> recognize the value of the other tests that one could perform, such as
> unit tests, integration tests behind the browser, and so on. And I bet
> eValid can help with those too, in some contexts--right?
>
> When you say "best possible" anything, you lose credibility with
> serious testers.
>
> ---Michael B.

That's a fair comment, but we believe that eValid does address
ANY kind of web browser enabled application.

For example, we have a 100% success rate at getting reliable
playback on ANY accessible web browser enabled application.

We even have a "guarantee" of that -- subject to some obvious
technical limits. (All bets are off if the application has
pages that are are broken, if there are JavaScript loops,
or any other kinds of strange behavior!)

We believe the claim of "best possible tests" is justified
because: (1) eValid tests execute inside the eValid browser;
and, (2) a server cannot distinguish that a robot is driving
the activity. Such tests are therefore 100% real, natural,
fully functional, and highly repeatable.

Another way of looking at eValid's strength is that testing a
web browser enabled application WITH a test enabled full-
featured browser is the obvious way to go. It makes the job
simple, unambiguous, and very effective.

eValid started with this evidently simple notion in 2000
(eValid Ver 1) and has been expanding on it ever since.
Please see:
http://www.e-Valid.com/Technology/why.browser.html

Moreover, you're quite correct about eValid helping with other
kinds of testing than "direct from the browser face." For
example, eValid assists with unit and component functional
testing, server side loading, web services testing, page
quality analysis (timing, tuning and checking), and complete
site analysis. Our goal has been to make eValid a full-
featured suite of capabilities.

Download your evaluation copy of eValid from:

http://www.e-Valid.com/Products/Dow...tml?status=FORM

Details on the eValid web analysis and testing suite are at:

http://www.e-Valid.com

Michael Bolton

2006-07-21, 7:05 pm

> > > For testing ANY kind of web enabled application the best
[color=darkred]
[color=darkred]
> That's a fair comment, but we believe that eValid does address
> ANY kind of web browser enabled application.

That's a different claim from the "best possible" claim.

> For example, we have a 100% success rate at getting reliable
> playback on ANY accessible web browser enabled application.
>
> We even have a "guarantee" of that -- subject to some obvious
> technical limits. (All bets are off if the application has
> pages that are are broken, if there are JavaScript loops,
> or any other kinds of strange behavior!)

....in which case a simple validator or a human tester, operating the
product manually, may be the "best possible" test.

> We believe the claim of "best possible tests" is justified
> because: (1) eValid tests execute inside the eValid browser;
> and, (2) a server cannot distinguish that a robot is driving
> the activity. Such tests are therefore 100% real, natural,
> fully functional, and highly repeatable.

That means that they're really good in certain contexts. It doesn't
mean that they're the best possible in all contexts.

I'm sure eValid is a fine product; I would expect nothing less from
Edward and his team. As such, it wouldn't require unsupportable
claims, would it?

---Michael B.

Sponsored Links