You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by Philip Peake <ph...@vogon.net> on 2006/10/16 19:26:35 UTC
Parallel processing of embeded URLs
In any modern browser, the browser normally supports multiple,
concurrent, HTTP requests (TCP connections).
This is typically used when processing a page requested by the user --
embedded links will be processed as they are received, and will be
processed in parallel up to the maximum concurrent connection limit of
the browser.
My question is, is it possible to emulate this behavior in Jmeter?
It would appear not. After fetching a page, each of the embedded URLs is
fetched by the same thread, serially. The reported time for the top
level page is reported as the sum of the times for loading the top level
page and each of the embedded URLs. This has no relation at all to the
time that will be taken by by a browser. the closes seems to be to
inhibit fetching embeded URLs, but this does not really emulate a real
world user load where the user will be fetching these URLs.
What is needed is that a thread fetches a page, discovers embedded URLs,
then spawns up to N more threads to fetch these URLs. The reported time
should be timed from the commencement of fetching the top level page, to
the end of the last fetch of an embedded URL. This will correctly
emulate a browser, and will report times in line with user experience.
Is there any way to do this?
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: Parallel processing of embeded URLs
Posted by Philip Peake <ph...@vogon.net>.
I thought that might be the answer :-)
Thanks.
Philip
---------------
sebb wrote:
> This is not possible at present; sorry.
>
> Best to create a Bugzilla enhancement request ...
>
> S
> On 16/10/06, Philip Peake <ph...@vogon.net> wrote:
>> In any modern browser, the browser normally supports multiple,
>> concurrent, HTTP requests (TCP connections).
>> This is typically used when processing a page requested by the user --
>> embedded links will be processed as they are received, and will be
>> processed in parallel up to the maximum concurrent connection limit of
>> the browser.
>>
>> My question is, is it possible to emulate this behavior in Jmeter?
>>
>> It would appear not. After fetching a page, each of the embedded URLs is
>> fetched by the same thread, serially. The reported time for the top
>> level page is reported as the sum of the times for loading the top level
>> page and each of the embedded URLs. This has no relation at all to the
>> time that will be taken by by a browser. the closes seems to be to
>> inhibit fetching embeded URLs, but this does not really emulate a real
>> world user load where the user will be fetching these URLs.
>>
>> What is needed is that a thread fetches a page, discovers embedded URLs,
>> then spawns up to N more threads to fetch these URLs. The reported time
>> should be timed from the commencement of fetching the top level page, to
>> the end of the last fetch of an embedded URL. This will correctly
>> emulate a browser, and will report times in line with user experience.
>>
>> Is there any way to do this?
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
>> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: Parallel processing of embeded URLs
Posted by sebb <se...@gmail.com>.
This is not possible at present; sorry.
Best to create a Bugzilla enhancement request ...
S
On 16/10/06, Philip Peake <ph...@vogon.net> wrote:
> In any modern browser, the browser normally supports multiple,
> concurrent, HTTP requests (TCP connections).
> This is typically used when processing a page requested by the user --
> embedded links will be processed as they are received, and will be
> processed in parallel up to the maximum concurrent connection limit of
> the browser.
>
> My question is, is it possible to emulate this behavior in Jmeter?
>
> It would appear not. After fetching a page, each of the embedded URLs is
> fetched by the same thread, serially. The reported time for the top
> level page is reported as the sum of the times for loading the top level
> page and each of the embedded URLs. This has no relation at all to the
> time that will be taken by by a browser. the closes seems to be to
> inhibit fetching embeded URLs, but this does not really emulate a real
> world user load where the user will be fetching these URLs.
>
> What is needed is that a thread fetches a page, discovers embedded URLs,
> then spawns up to N more threads to fetch these URLs. The reported time
> should be timed from the commencement of fetching the top level page, to
> the end of the last fetch of an embedded URL. This will correctly
> emulate a browser, and will report times in line with user experience.
>
> Is there any way to do this?
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org