theoretically you could achieve what you mentioned by opening a 1000 browser
windows with some manual effort, but it would be hard to generate some good
statistics and concurrency out of it. Can't be termed as a controlled
environment. What you would need is a tool which can simulate such a
scenario and allow you to control the params -

http://www.eviware.com/blogs/oleblog/?p=861

http://blog.rubyamf.org/2008/01/20/amf-stress-testing-flex-ruby-jmeter/

http://www.adobe.com/devnet/flex/articles/amf_perftest.html

~Peeyush

http://www.mds.asia


On Fri, Feb 26, 2010 at 9:10 PM, as sass <apar...@yahoo.com> wrote:

>
>
> Yes, we want to implement a "load test" in fact. We want to test our system
> for concurrent sessions/users. For example, in FlexMonkey for one user it
> does the functional test. But i want to do a load test for 1000 or 10000
> concurent users that sending amf request to the server and trace the results
> about time.
>
> As a conclusion, the thing that i want is,  a kind of merged version:
> FlexMonkey functionality + load test for Flex application that sending amf
> requests from different buttons, actions to server.
>
>
>
> --- On *Fri, 2/26/10, Peeyush Tuli <peeyus...@gmail.com>* wrote:
>
>
> From: Peeyush Tuli <peeyus...@gmail.com>
> Subject: Re: [flexcoders] Flex stress Testing
> To: flexcoders@yahoogroups.com
> Date: Friday, February 26, 2010, 3:30 AM
>
>
>
>
> it seems like you want to test your system for concurrent sessions/users. I
> am not sure how does that relate to flex which is a client technology. Do
> you intend to stress test your back end/server infrastructure like
> webservices, LCDS/Blazeds infrastructure?
>
> On Fri, Feb 26, 2010 at 3:36 PM, as sass <apar...@yahoo. 
> com<http://mc/compose?to=apar...@yahoo.com>
> > wrote:
>
>>
>>
>> Hi, everybody,, i want to learn how to flex user-stress testing. For
>> example, i want to test how my system shows feedback after 10000 users enter
>> the system. How can i trace the request and response?Which tool can accept
>> my needs?
>>
>> Thank you...
>>
>>
>
>  
>

Reply via email to