JMeter assumes a 4XX response code is a failure

[Noel O'Brien]

Hi All,

I'm using JMeter to do some negative testing on my webapp, but when I add a 
Response Assertion which (correctly) expects that the response code will be 
404, the result is always a fail.

I looked through the source code and it appears in HTTPSamplerBase, that 
JMeter will only check the response assertion for codes 200-399 inclusive; 
otherwise it assumes it's a failure.

To me this is a bug in JMeter as it prevents one from executing negative 
testing. Perhaps if a response assertion was omitted, then it would be 
logical to assume that a 4XX would be a failure, but when I've specified that 
I expect a 4XX shouldn't it be honoured?

Does any one know of a way to get around this problem? I'd appreciate any 
feedback on this issue.

Regards,
Noel

-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: JMeter assumes a 4XX response code is a failure

[Noel O'Brien]

I "Read the FM" but missed that bit ;) Thanks Sebb for pointing it out!

Regards,
Noel

On Monday 26 January 2009 17:51:46 sebb wrote:
> On 26/01/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  I'm using JMeter to do some negative testing on my webapp, but when I
> > add a Response Assertion which (correctly) expects that the response code
> > will be 404, the result is always a fail.
> >
> >  I looked through the source code and it appears in HTTPSamplerBase, that
> >  JMeter will only check the response assertion for codes 200-399
> > inclusive; otherwise it assumes it's a failure.
> >
> >  To me this is a bug in JMeter as it prevents one from executing negative
> >  testing. Perhaps if a response assertion was omitted, then it would be
> >  logical to assume that a 4XX would be a failure, but when I've specified
> > that I expect a 4XX shouldn't it be honoured?
>
> http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Respon
>se_Assertion
>
> says
>
> "HTTP Responses with statuses in the 4xx and 5xx ranges are normally
> regarded as unsuccessful. The "Ignore status" checkbox can be used to
> set the status successful before performing further checks. Note that
> this will have the effect of clearing any previous assertion failures,
> so make sure that this is only set on the first assertion. "
>
> >  Does any one know of a way to get around this problem? I'd appreciate
> > any feedback on this issue.
> >
> >  Regards,
> >  Noel
> >
> >  -
> >  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: RESTful API

[Noel O'Brien]

I've done some JMeter testing against restful API. Do you have any specific 
questions?

Regards,
Noel


On Tuesday 27 January 2009 20:30:33 Ray Trejo wrote:
> Hi,
>
> Has anyone out there done any load testing with Jmeter and restful API?
>
>
>
> Thank you,
>
>
>
> Ray Trejo
>
> Rackspace
>
>
>
> Quality Assurance Engineer
>
> Direct:  (210) 312 - 5326
>
> Email:  ray.tr...@rackspace.com 
>
>
>
>
>
> Confidentiality Notice: This e-mail message (including any attached or
> embedded documents) is intended for the exclusive and confidential use of
> the individual or entity to which this message is addressed, and unless
> otherwise expressly indicated, is confidential and privileged information
> of Rackspace. Any dissemination, distribution or copying of the enclosed
> material is prohibited. If you receive this transmission in error, please
> notify us immediately by e-mail at ab...@rackspace.com, and delete the
> original message.
> Your cooperation is appreciated.



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: RESTful API

[Noel O'Brien]

What are the objectives of your test? (What do you want to prove/disprove)? 
Figure out this first, then you can design your tests in JMeter accordingly.

I haven't load tested REST over HTTPS but I don't think it should be any 
different than HTTP. There's a good article here if you need to get going 
quickly:

http://jakarta.apache.org/jmeter/usermanual/boss.html

Is the authentication you describe done by providing the username/password 
through a form or through HTTP BASIC (a popup dialog)?

Regards,
Noel

On Wednesday 28 January 2009 14:46:01 Ray Trejo wrote:
> Noel,
> Thanks for responding.
> i need to create a load test using jmeter for our rest API.
>
> Do you have a simple example you can send me?
>
> We are authenticating with a username and password.  Also, the rest
> services are https.
>
> I need a little help getting started.
> Let me know.  Ray
>
> Thank you,
>
> Ray Trejo
> Rackspace
>
> Quality Assurance Engineer
> Direct:  (210) 312 - 5326
> Email:  ray.tr...@rackspace.com
> -Original Message-
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: Wednesday, January 28, 2009 3:21 AM
> To: jmeter-user@jakarta.apache.org
> Cc: Ray Trejo
> Subject: Re: RESTful API
>
> I've done some JMeter testing against restful API. Do you have any
> specific
> questions?
>
> Regards,
> Noel
>
> On Tuesday 27 January 2009 20:30:33 Ray Trejo wrote:
> > Hi,
> >
> > Has anyone out there done any load testing with Jmeter and restful
>
> API?
>
> > Thank you,
> >
> >
> >
> > Ray Trejo
> >
> > Rackspace
> >
> >
> >
> > Quality Assurance Engineer
> >
> > Direct:  (210) 312 - 5326
> >
> > Email:  ray.tr...@rackspace.com <mailto:ray.tr...@rackspace.com>
> >
> >
> >
> >
> >
> > Confidentiality Notice: This e-mail message (including any attached or
> > embedded documents) is intended for the exclusive and confidential use
>
> of
>
> > the individual or entity to which this message is addressed, and
>
> unless
>
> > otherwise expressly indicated, is confidential and privileged
>
> information
>
> > of Rackspace. Any dissemination, distribution or copying of the
>
> enclosed
>
> > material is prohibited. If you receive this transmission in error,
>
> please
>
> > notify us immediately by e-mail at ab...@rackspace.com, and delete the
> > original message.
> > Your cooperation is appreciated.
>
> Confidentiality Notice: This e-mail message (including any attached or
> embedded documents) is intended for the exclusive and confidential use of
> the individual or entity to which this message is addressed, and unless
> otherwise expressly indicated, is confidential and privileged information
> of Rackspace. Any dissemination, distribution or copying of the enclosed
> material is prohibited. If you receive this transmission in error, please
> notify us immediately by e-mail at ab...@rackspace.com, and delete the
> original message.
> Your cooperation is appreciated.



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: how to use "CSV Data Set Config" to take input from file

[Noel O'Brien]

What kind of testing (functional, load)?
Are you limited to a set of login data (e.g. 20 users)

Regards,
Noel

On Wednesday 28 January 2009 16:56:55 porobashi wrote:
> please help me out guys.
>
> porobashi wrote:
> > hi all,
> >
> > i am trying to test a sign-in(registration) form. hence, i need to use
> > different data for each submission. i think using "CSV Data Set Config"
> > would be the easiest route. i need your help to give me a step-by-step
> > example on "CSV Data Set Config" using. all fields are not mandatory for
> > the form, is there way to only entering mandatory fields???
> >
> > please help me out ASAP.
> >
> > thanks in advance.



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: how to use "CSV Data Set Config" to take input from file

[Noel O'Brien]

The number of concurrent users is controlled by he number of threads you 
configure in the Thread Group module.

CVS Data Set Config module will load one line of a CSV file for each thread, 
so create a CSV file with at least as many entries as threads.

Then you can add a HTTP Request sampler, specify the protocol as HTTPS and 
fill in the rest of the necessary data

Don't forget to add a listener (e.g. View Results Tree) so that you can see 
what's going on

Regards,
Noel

On Wednesday 28 January 2009 17:23:57 porobashi wrote:
> I am trying to load test our registration page. I am not limited on users,
> but will start with 10-20 users.
>
> I will like to enter only required fields value.
>
> My page is https, I will use BadBoy to record the page. Then I need to add
> "CSV Data Set Config" to the thread group, right? How to go from there?
>
> Thanks
>
> Noel O'Brien wrote:
> > What kind of testing (functional, load)?
> > Are you limited to a set of login data (e.g. 20 users)
> >
> > Regards,
> > Noel
> >
> > On Wednesday 28 January 2009 16:56:55 porobashi wrote:
> >> please help me out guys.
> >>
> >> porobashi wrote:
> >> > hi all,
> >> >
> >> > i am trying to test a sign-in(registration) form. hence, i need to use
> >> > different data for each submission. i think using "CSV Data Set
> >> > Config" would be the easiest route. i need your help to give me a
> >> > step-by-step example on "CSV Data Set Config" using. all fields are
> >> > not mandatory
> >>
> >> for
> >>
> >> > the form, is there way to only entering mandatory fields???
> >> >
> >> > please help me out ASAP.
> >> >
> >> > thanks in advance.
> >
> > -
> > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: how to use "CSV Data Set Config" to take input from file

[Noel O'Brien]

Please copy jmeter-user@jakarta.apache.org in all responses so that others may 
benefit and help contribute to the thread.

What happens when you run it?

Regards,
Noel

On Wednesday 28 January 2009 18:32:09 Porobashi M wrote:
> Hi Noel,
>
> Thanks for your help. As I said earlier I am testing one of your
> applications Registration Page.
>
> I am attaching a screenshot so that you can how my test plan is.
>
> What to do now???
>
> Thanks
>
>
>
>
> ____
> From: Noel O'Brien 
> To: jmeter-user@jakarta.apache.org
> Cc: porobashi 
> Sent: Wednesday, January 28, 2009 10:34:37 AM
> Subject: Re: how to use "CSV Data Set Config" to take input from file
>
> The number of concurrent users is controlled by he number of threads you
> configure in the Thread Group module.
>
> CVS Data Set Config module will load one line of a CSV file for each
> thread, so create a CSV file with at least as many entries as threads.
>
> Then you can add a HTTP Request sampler, specify the protocol as HTTPS and
> fill in the rest of the necessary data
>
> Don't forget to add a listener (e.g. View Results Tree) so that you can see
> what's going on
>
> Regards,
> Noel
>
> On Wednesday 28 January 2009 17:23:57 porobashi wrote:
> > I am trying to load test our registration page. I am not limited on
> > users, but will start with 10-20 users.
> >
> > I will like to enter only required fields value.
> >
> > My page is https, I will use BadBoy to record the page. Then I need to
> > add "CSV Data Set Config" to the thread group, right? How to go from
> > there?
> >
> > Thanks
> >
> > Noel O'Brien wrote:
> > > What kind of testing (functional, load)?
> > > Are you limited to a set of login data (e.g. 20 users)
> > >
> > > Regards,
> > > Noel
> > >
> > > On Wednesday 28 January 2009 16:56:55 porobashi wrote:
> > >> please help me out guys.
> > >>
> > >> porobashi wrote:
> > >> > hi all,
> > >> >
> > >> > i am trying to test a sign-in(registration) form. hence, i need to
> > >> > use different data for each submission. i think using "CSV Data Set
> > >> > Config" would be the easiest route. i need your help to give me a
> > >> > step-by-step example on "CSV Data Set Config" using. all fields are
> > >> > not mandatory
> > >>
> > >> for
> > >>
> > >> > the form, is there way to only entering mandatory fields???
> > >> >
> > >> > please help me out ASAP.
> > >> >
> > >> > thanks in advance.
> > >
> > > -
> > > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Jmeter is keeping the http response in memory

[Noel O'Brien]

Hi,

Make sure you have "Functional Mode" unchecked and if you have any listeners 
writing to file, configure them to only save the data you absolutely need.

Regards,
Noel

n Wednesday 04 February 2009 15:32:56 kretes wrote:
> Hello.
>
> I have a specific need of testing an app, that is responding with a file.
> currently it is about 33MB for response. And I want to stress test my app
> with let's say a 100 simultaneously requests. I would like JMeter to throw
> away this data - it's not needed.
> And What i get is a outofmemory of course.
>
> I do not have anything except sampler. is there a way to tell JMeter not to
> keep the response?
> Right now I have a 10 * 33 MB response, and JMeter eats 850MB of memory
> after test...
>
> Thanks in advance
> Tomasz Bartczak



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Additional double quotes after csvread?

[Noel O'Brien]

Are you importing the CSV data into JMeter using a CSV Data Set Config 
element?

I haven't used it myself, but I see from
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#CSV_Data_Set_Config
that there's a "Allow Quoted Data" option. Have you played around with that?

Regards,
Noel

On Friday 06 February 2009 12:14:32 stevens_san...@hotmail.com wrote:
> hmmm, looked at the beanshell pre-processor, but I am not that technical.
> can't figure out how to make it work. Isn't there a simple way of putting
> something in front of the double quote in my csv, in order to make sure
> that jmeter will handle the quotes as text during the parse?> From:
> nobr...@newbay.com> To: jmeter-user@jakarta.apache.org> Subject: Re:
> Additional double quotes after csvread?> Date: Fri, 6 Feb 2009 12:10:52
> +> CC: stevens_san...@hotmail.com> > One way to do it would be to have
> a beanshell pre-processor to substitute the > single quotes for double
> quotes. > > Regards,> Noel> > On Friday 06 February 2009 12:04:42
> stevens_san...@hotmail.com wrote:> > Hi, thanks fro the prompt reply.> >
> Yes, tried that. That works fine (the parsing part at least), but I need> >
> the double quotes in my post request. My http request with single quotes> >
> doesn't work.> >> > > From: nobr...@newbay.com> To:
> jmeter-user@jakarta.apache.org> Subject:> > > Re: Additional double quotes
> after csvread?> Date: Fri, 6 Feb 2009> > > 11:50:08 +> CC:
> stevens_san...@hotmail.com> > Have you tried it with> > > single quotes?> >
> tag='to be or not to be'> > On Friday 06 February 2009> > > 11:46:12
> stevens_san...@hotmail.com wrote:> > Hi all,> >> > I have a> > > problem
> with my data driven test. My testdata is in an external> > csv> > > file. I
> got one cell with the content tag="to be or not to be". Notice> >> > > the
> double quotes around the string! Now I do a csvread and fill a> > >
> variable> > with the content of that specific cell. In this process,> > >
> additional quotes> > are added to my string. It looks like this:> > >
> query="tag=""to be or not to> > be""". So around tag= are additional> > >
> quotes and around the string are> > additional quotes. Why? I can't post> >
> > this request looking like this.> >> > Hope somebody has seen this before>
> > > and got a solution for me.> >> > Thanks!> > Sander> >> > >
> _> >> > >
> Nieuw: nu ook chatten op je mobiel!> >> > >
> http://www.overaljevriendenbijje.nl/#superdeal> > > >> > >
> -> To>
> > > unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org> For> >
> > additional commands, e-mail: jmeter-user-h...@jakarta.apache.org>> >> >
> _> > Nieuw:
> nu ook chatten op je mobiel!> >
> http://www.overaljevriendenbijje.nl/#superdeal> > > >
> -> To
> unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org> For
> additional commands, e-mail: jmeter-user-h...@jakarta.apache.org>
> _
> Download de nieuwste emoticons voor in je Messenger
> http://www.msnmessengerexperience.nl/chuck/



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Additional double quotes after csvread?

[Noel O'Brien]

Have you tried it with single quotes?

tag='to be or not to be'

On Friday 06 February 2009 11:46:12 stevens_san...@hotmail.com wrote:
> Hi all,
>
> I have a problem with my data driven test. My testdata is in an external
> csv file. I got one cell with the content tag="to be or not to be". Notice
> the double quotes around the string! Now I do a csvread and fill a variable
> with the content of that specific cell. In this process, additional quotes
> are added to my string. It looks like this: query="tag=""to be or not to
> be""". So around tag= are additional quotes and around the string are
> additional quotes. Why? I can't post this request looking like this.
>
> Hope somebody has seen this before and got a solution for me.
>
> Thanks!
> Sander
> _
> Nieuw: nu ook chatten op je mobiel!
> http://www.overaljevriendenbijje.nl/#superdeal



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Additional double quotes after csvread?

[Noel O'Brien]

One way to do it would be to have a beanshell pre-processor to substitute the 
single quotes for double quotes. 

Regards,
Noel

On Friday 06 February 2009 12:04:42 stevens_san...@hotmail.com wrote:
> Hi, thanks fro the prompt reply.
> Yes, tried that. That works fine (the parsing part at least), but I need
> the double quotes in my post request. My http request with single quotes
> doesn't work.
>
> > From: nobr...@newbay.com> To: jmeter-user@jakarta.apache.org> Subject:
> > Re: Additional double quotes after csvread?> Date: Fri, 6 Feb 2009
> > 11:50:08 +> CC: stevens_san...@hotmail.com> > Have you tried it with
> > single quotes?> > tag='to be or not to be'> > On Friday 06 February 2009
> > 11:46:12 stevens_san...@hotmail.com wrote:> > Hi all,> >> > I have a
> > problem with my data driven test. My testdata is in an external> > csv
> > file. I got one cell with the content tag="to be or not to be". Notice> >
> > the double quotes around the string! Now I do a csvread and fill a
> > variable> > with the content of that specific cell. In this process,
> > additional quotes> > are added to my string. It looks like this:
> > query="tag=""to be or not to> > be""". So around tag= are additional
> > quotes and around the string are> > additional quotes. Why? I can't post
> > this request looking like this.> >> > Hope somebody has seen this before
> > and got a solution for me.> >> > Thanks!> > Sander> >
> > _> >
> > Nieuw: nu ook chatten op je mobiel!> >
> > http://www.overaljevriendenbijje.nl/#superdeal> > > >
> > -> To
> > unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org> For
> > additional commands, e-mail: jmeter-user-h...@jakarta.apache.org>
>
> _
> Nieuw: nu ook chatten op je mobiel!
> http://www.overaljevriendenbijje.nl/#superdeal



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: JMeter JMS Point-to-Point with Weblogic JMS

[Noel O'Brien]

Hi Peter,sing it with SwiftMQ and have not seen any problems like you 
describe. 

What are you putting in the "Content" field in JMeter?

Regards,
Noel

On Monday 09 February 2009 05:21:41 O'Neill, Peter (SAPOL) wrote:
> Hi There,
> I am trying to send a JMS Point-to-Point request to a Weblogic JMS
> queue. We are using jmeter 2.3.2 and Weblogic 10.
>
> We have the settings correct and our JMS queue is receiving messages,
> however the message 'content' is always empty.
>
> Our jmeter settings are as follows ...
>
> QueueConnection Factory : weblogic.jms.ConnectionFactory
> JNDINameRequestQueue : 
> InitialContext Factory : weblogic.jndi.WLInitialContextFactory
> Provider URL : t3://localhost:7001
>
> Does anyone have any experience with JMeter JMS Point To Point and
> Weblogic?
>
> Regards,
> Peter O'Neill



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Copying Test elements from one JMX file to another ()

[Noel O'Brien]

Hi All,

I recently introduced JMeter to my workplace and now there are several people 
working on different parts of the same test plan file (Acceptance Test 
suite.jmx). 

The only problem we're seeing is when we try to check it into our source code 
management system (Perforce). As files have been edited in multiple places 
they need to be resolved (as expected). However, any diff/merge tool I've 
tried to use doesn't correctly pick up the diffs, possibly due to the nested 
nature of XML, e.g. if I only add a parent controller to a file and check it 
in, the new  controller is seen as a change to the controller it now has 
beneath it, and this effect cascades down through the file.

I'm just curious how others deal with this problem or if they even experience 
it. How have people gotten around it? I've used the merge tool in P4V and 
I've tried Meld too, but both display the issue described above. Maybe I need 
to flatten (remove indentation) the file first, then diff/merge, then 
re-indent after merge?

Thanks for you help,
Noel

-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Copying Test elements from one JMX file to another

[Noel O'Brien]

Hi All,

I've figured out how to diff successfully using P4Merge: Select File > 
Comparison Method > Ignore Line Ending and All White Space Differences

I couldn't figure it out in Meld but hopefully what I've posted will help 
someone.

One the same topic, but slightly different question: How can I copy test 
elements from one test plan to another? If I open two test plans in their own 
instance of JMeter it is not possible to drag / copy / CTRL-C from one to 
another. Also, I've tried copying elements to the Workbench, then opened a 
new test plan, but the Workbench is cleared when a test plan is closed.

Maybe it would be a good feature to have the workbench store test elements 
regardless of whether there's a test plan currently loaded or not (But it 
would still be in-memory). this would enable users to copy elements to-from 
testplans.

Regards,
Noel

On Monday 09 February 2009 10:26:50 Noel O'Brien wrote:
> Hi All,
>
> I recently introduced JMeter to my workplace and now there are several
> people working on different parts of the same test plan file (Acceptance
> Test suite.jmx).
>
> The only problem we're seeing is when we try to check it into our source
> code management system (Perforce). As files have been edited in multiple
> places they need to be resolved (as expected). However, any diff/merge tool
> I've tried to use doesn't correctly pick up the diffs, possibly due to the
> nested nature of XML, e.g. if I only add a parent controller to a file and
> check it in, the new  controller is seen as a change to the controller it
> now has beneath it, and this effect cascades down through the file.
>
> I'm just curious how others deal with this problem or if they even
> experience it. How have people gotten around it? I've used the merge tool
> in P4V and I've tried Meld too, but both display the issue described above.
> Maybe I need to flatten (remove indentation) the file first, then
> diff/merge, then re-indent after merge?
>
> Thanks for you help,
> Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Copying Test elements from one JMX file to another

[Noel O'Brien]

Thanks for your help Axel/Sebb. Looks like saving  / merging test plan parts 
is the only option for now. 

I've logged an enhancement here:

https://issues.apache.org/bugzilla/show_bug.cgi?id=46677

suggesting a persistant workbench while JMeter is running and drag/drop and 
copy/paste between JMeter instances.

Regards,
Noel

On Monday 09 February 2009 11:44:17 sebb wrote:
> On 09/02/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  I've figured out how to diff successfully using P4Merge: Select File >
> >  Comparison Method > Ignore Line Ending and All White Space Differences
> >
> >  I couldn't figure it out in Meld but hopefully what I've posted will
> > help someone.
> >
> >  One the same topic, but slightly different question: How can I copy test
> >  elements from one test plan to another? If I open two test plans in
> > their own instance of JMeter it is not possible to drag / copy / CTRL-C
> > from one to another. Also, I've tried copying elements to the Workbench,
> > then opened a new test plan, but the Workbench is cleared when a test
> > plan is closed.
>
> One way is to select the element(s) in the source plan and use Save As...
>
> Then open the target plan and use Merge ... to add the snippet to the
> appropriate place in the tree.
>
> >  Maybe it would be a good feature to have the workbench store test
> > elements regardless of whether there's a test plan currently loaded or
> > not (But it would still be in-memory). this would enable users to copy
> > elements to-from testplans.
>
> Feel free to open a Bugzilla enhancement request.
>
> >  Regards,
> >  Noel
> >
> >  On Monday 09 February 2009 10:26:50 Noel O'Brien wrote:
> >  > Hi All,
> >  >
> >  > I recently introduced JMeter to my workplace and now there are several
> >  > people working on different parts of the same test plan file
> >  > (Acceptance Test suite.jmx).
> >  >
> >  > The only problem we're seeing is when we try to check it into our
> >  > source code management system (Perforce). As files have been edited in
> >  > multiple places they need to be resolved (as expected). However, any
> >  > diff/merge tool I've tried to use doesn't correctly pick up the diffs,
> >  > possibly due to the nested nature of XML, e.g. if I only add a parent
> >  > controller to a file and check it in, the new  controller is seen as a
> >  > change to the controller it now has beneath it, and this effect
> >  > cascades down through the file.
> >  >
> >  > I'm just curious how others deal with this problem or if they even
> >  > experience it. How have people gotten around it? I've used the merge
> >  > tool in P4V and I've tried Meld too, but both display the issue
> >  > described above. Maybe I need to flatten (remove indentation) the file
> >  > first, then diff/merge, then re-indent after merge?
> >  >
> >  > Thanks for you help,
> >  > Noel
> >  >
> >  > -
> >  > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
> >
> >  -
> >  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: JMeter JMS Point-to-Point with Weblogic JMS

[Noel O'Brien]

Hmmm, I don't have any other ideas. Is there any useful information printed to 
STDOUT or jmeter.log?

Regards,
Noel

On Tuesday 10 February 2009 03:11:05 O'Neill, Peter (SAPOL) wrote:
> Hi Nick,
> We have straight text in our content -
>
> content : test
>
> Regards,
> Peter
>
> -----Original Message-
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: Monday, 9 February 2009 7:54 PM
> To: jmeter-user@jakarta.apache.org
> Cc: O'Neill, Peter (SAPOL); Bryant, Kane (SAPOL)
> Subject: Re: JMeter JMS Point-to-Point with Weblogic JMS
>
> Hi Peter,sing it with SwiftMQ and have not seen any problems like you
> describe.
>
> What are you putting in the "Content" field in JMeter?
>
> Regards,
> Noel
>
> On Monday 09 February 2009 05:21:41 O'Neill, Peter (SAPOL) wrote:
> > Hi There,
> > I am trying to send a JMS Point-to-Point request to a Weblogic JMS
> > queue. We are using jmeter 2.3.2 and Weblogic 10.
> >
> > We have the settings correct and our JMS queue is receiving messages,
> > however the message 'content' is always empty.
> >
> > Our jmeter settings are as follows ...
> >
> > QueueConnection Factory : weblogic.jms.ConnectionFactory
> > JNDINameRequestQueue :  InitialContext Factory :
> > weblogic.jndi.WLInitialContextFactory
> > Provider URL : t3://localhost:7001
> >
> > Does anyone have any experience with JMeter JMS Point To Point and
> > Weblogic?
> >
> > Regards,
> > Peter O'Neill
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Resetting a "Counter" Preprocessor

[Noel O'Brien]

Hi All,

Is it possible to reset a counter preprocessor? I've been pulling my hair out 
about it all morning.

I've got a For Each controller which contain a Simple controller and a Loop 
controller respectively. The loop controller contains a Counter, which I 
would like to reset each time the For Each controller iterates. I've tried 
resetting the variable that the counter writes to ('count') but it appears 
the count index is maintained in the Counter element, and over-writes any 
value I've set in the 'count' variable. 

I've worked around the problem by implementing a "My Counter" Beanshell 
Preprocessor that I've put in the Loop controller and a "Reset Counter" 
Beanshell Postprocessor in the Simple controller. While my solution works, 
I'd prefer to stick to build-in modules for this kind of thing, otherwise 
maintaining test plans becomes more difficult and error-prone.

Does anyone know of a more elegant solution? Maybe the Counter element could 
be modified to allow it to be reset?

Regards,
Noel

-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

JMS Assertion??

[Noel O'Brien]

Hi All,

The web application I'm testing uses JMS queues for communicating externally 
when certain events occur. The JMS point to point sampler in JMeter only 
provides for Request and Request/Response tests, however almost all of the 
messaging I will be testing will be driven by a HTTP request, i.e. make a 
HTTP request and there should be a corresponding message on a JMS queue. 

What is the best way to (at least) functionally test this? I was thinking of 
writing a JMS Assertion that I could add to the HTTP Request sampler, does 
that sound reasonable? The JMS message occurs as a result of the HTTP request 
being made so to me it makes sense for it to be an assertion.

Is there any other way of having a "Response" only JMS sampler?

Regards,
Noel

-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: JMS Assertion??

[Noel O'Brien]

On Monday 16 February 2009 15:42:46 sebb wrote:
> On 16/02/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  The web application I'm testing uses JMS queues for communicating
> > externally when certain events occur. The JMS point to point sampler in
> > JMeter only provides for Request and Request/Response tests, however
> > almost all of the messaging I will be testing will be driven by a HTTP
> > request, i.e. make a HTTP request and there should be a corresponding
> > message on a JMS queue.
> >
> >  What is the best way to (at least) functionally test this? I was
> > thinking of writing a JMS Assertion that I could add to the HTTP Request
> > sampler, does that sound reasonable? The JMS message occurs as a result
> > of the HTTP request being made so to me it makes sense for it to be an
> > assertion.
>
> Is the JMS message sent before the HTTP response is returned? Can one
> guarantee this?
> I thought JMS was asynchronous.

JMS is indeed asynchronous. The HTTP response will always return first, and 
the logic that puts the message on the queue is run in a background thread 
which has a default pause interval or 30 seconds. So there would be a maximum 
30 second delay between the http request completing and the message appearing 
on the queue

> I would expect the HTTP Response Assertion to be used for checking
> that the HTTP response is OK, i.e. the page contains "message sent OK"
> or whatever.
>
> >  Is there any other way of having a "Response" only JMS sampler?
>
> I've never used JMS, but isn't that what JMS Subscriber does?

We are not using topic/subscriber model for JMS, but queues (FIFO). Will the 
JMS Subscriber sampler work with queues?

> i.e what does the intended JMS recipient do?

The intended JMS recipient parses the message and does further processing. In 
our example, it's an SMS gateway, but is outside the scope of our app, so I 
do not need to test it. I do need to test that the correct messages are 
coming off the queue (in quantity and content) and will need to performance 
test it in future

Regards,
Noel

> >  Regards,
> >  Noel
> >
> >  -
> >  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: JMS Assertion??

[Noel O'Brien]

On Monday 16 February 2009 16:35:56 sebb wrote:
> On 16/02/2009, Noel O'Brien  wrote:
> > On Monday 16 February 2009 15:42:46 sebb wrote:
> >  > On 16/02/2009, Noel O'Brien  wrote:
> >  > > Hi All,
> >  > >
> >  > >  The web application I'm testing uses JMS queues for communicating
> >  > > externally when certain events occur. The JMS point to point sampler
> >  > > in JMeter only provides for Request and Request/Response tests,
> >  > > however almost all of the messaging I will be testing will be driven
> >  > > by a HTTP request, i.e. make a HTTP request and there should be a
> >  > > corresponding message on a JMS queue.
> >  > >
> >  > >  What is the best way to (at least) functionally test this? I was
> >  > > thinking of writing a JMS Assertion that I could add to the HTTP
> >  > > Request sampler, does that sound reasonable? The JMS message occurs
> >  > > as a result of the HTTP request being made so to me it makes sense
> >  > > for it to be an assertion.
> >  >
> >  > Is the JMS message sent before the HTTP response is returned? Can one
> >  > guarantee this?
> >  > I thought JMS was asynchronous.
> >
> > JMS is indeed asynchronous. The HTTP response will always return first,
> > and the logic that puts the message on the queue is run in a background
> > thread which has a default pause interval or 30 seconds. So there would
> > be a maximum 30 second delay between the http request completing and the
> > message appearing on the queue
> >
> >  > I would expect the HTTP Response Assertion to be used for checking
> >  > that the HTTP response is OK, i.e. the page contains "message sent OK"
> >  > or whatever.
> >  >
> >  > >  Is there any other way of having a "Response" only JMS sampler?
> >  >
> >  > I've never used JMS, but isn't that what JMS Subscriber does?
> >
> > We are not using topic/subscriber model for JMS, but queues (FIFO). Will
> > the JMS Subscriber sampler work with queues?
>
> No idea, sorry.
>
> >  > i.e what does the intended JMS recipient do?
> >
> > The intended JMS recipient parses the message and does further
> > processing. In our example, it's an SMS gateway, but is outside the scope
> > of our app, so I do not need to test it. I do need to test that the
> > correct messages are coming off the queue (in quantity and content) and
> > will need to performance test it in future
>
> I meant - what JMS actions does the recipient do to acquire the
> message, not what it does with the message once received.

I have no idea how the SMS gateway interacts with the queue, but as it's 
written in Java I assume it would use the onMessage() function.

I will investigate further the JMS Subscriber sampler and report my findings.

Thanks for your help,
Noel

> >  Regards,
> >
> > Noel
> >
> >  > >  Regards,
> >  > >  Noel
> >  > >
> >  > > 
> >  > > 
> >  > >- To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  > > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
> >  >
> >  > -
> >  > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: JMS Assertion??

[Noel O'Brien]

On Monday 16 February 2009 17:02:32 sebb wrote:
> On 16/02/2009, Noel O'Brien  wrote:
> > On Monday 16 February 2009 16:35:56 sebb wrote:
> >  > On 16/02/2009, Noel O'Brien  wrote:
> >  > > On Monday 16 February 2009 15:42:46 sebb wrote:
> >  > >  > On 16/02/2009, Noel O'Brien  wrote:
> >  > >  > > Hi All,
> >  > >  > >
> >  > >  > >  The web application I'm testing uses JMS queues for
> >  > >  > > communicating externally when certain events occur. The JMS
> >  > >  > > point to point sampler in JMeter only provides for Request and
> >  > >  > > Request/Response tests, however almost all of the messaging I
> >  > >  > > will be testing will be driven by a HTTP request, i.e. make a
> >  > >  > > HTTP request and there should be a corresponding message on a
> >  > >  > > JMS queue.
> >  > >  > >
> >  > >  > >  What is the best way to (at least) functionally test this? I
> >  > >  > > was thinking of writing a JMS Assertion that I could add to the
> >  > >  > > HTTP Request sampler, does that sound reasonable? The JMS
> >  > >  > > message occurs as a result of the HTTP request being made so to
> >  > >  > > me it makes sense for it to be an assertion.
> >  > >  >
> >  > >  > Is the JMS message sent before the HTTP response is returned? Can
> >  > >  > one guarantee this?
> >  > >  > I thought JMS was asynchronous.
> >  > >
> >  > > JMS is indeed asynchronous. The HTTP response will always return
> >  > > first, and the logic that puts the message on the queue is run in a
> >  > > background thread which has a default pause interval or 30 seconds.
> >  > > So there would be a maximum 30 second delay between the http request
> >  > > completing and the message appearing on the queue
> >  > >
> >  > >  > I would expect the HTTP Response Assertion to be used for
> >  > >  > checking that the HTTP response is OK, i.e. the page contains
> >  > >  > "message sent OK" or whatever.
> >  > >  >
> >  > >  > >  Is there any other way of having a "Response" only JMS
> >  > >  > > sampler?
> >  > >  >
> >  > >  > I've never used JMS, but isn't that what JMS Subscriber does?
> >  > >
> >  > > We are not using topic/subscriber model for JMS, but queues (FIFO).
> >  > > Will the JMS Subscriber sampler work with queues?
> >  >
> >  > No idea, sorry.
> >  >
> >  > >  > i.e what does the intended JMS recipient do?
> >  > >
> >  > > The intended JMS recipient parses the message and does further
> >  > > processing. In our example, it's an SMS gateway, but is outside the
> >  > > scope of our app, so I do not need to test it. I do need to test
> >  > > that the correct messages are coming off the queue (in quantity and
> >  > > content) and will need to performance test it in future
> >  >
> >  > I meant - what JMS actions does the recipient do to acquire the
> >  > message, not what it does with the message once received.
> >
> > I have no idea how the SMS gateway interacts with the queue, but as it's
> >  written in Java I assume it would use the onMessage() function.
>
> Presumably it will have to do something first in order to use that
> function.
>
> >  I will investigate further the JMS Subscriber sampler and report my
> > findings.
>
> Thanks.
>
> If it turns out that there is a missing aspect of JMS that JMeter
> could implement, then we can see about adding it for a future release.

I  tried using a JMS Subscriber sampler to take some messages off a queue, but 
the following exception happened in jmeter.log:

2009/02/16 17:04:20 ERROR - jmeter.threads.JMeterThread:  
java.lang.ClassCastException: com.swiftmq.jms.QueueImpl cannot be cast to 
javax.jms.Topic
at 
org.apache.jmeter.protocol.jms.client.InitialContextFactory.lookupTopic(InitialContextFactory.java:80)
at 
org.apache.jmeter.protocol.jms.client.OnMessageSubscriber.initConnection(OnMessageSubscriber.java:119)
at 
org.apache.jmeter.protocol.jms.client.OnMessageSubscriber.(OnMessageSubscriber.java:79)
at 
org.apache.jmeter.protocol.jms.sampler.SubscriberSampler.initListenerClient(SubscriberSampler.java:105)
at 
org.apache.jmeter.protocol.jms.sampler.Subscriber

Re: Building an XML string for response

[Noel O'Brien]

On Wednesday 18 February 2009 14:08:24 Hiro Protagonist wrote:
> sebb,
>
> Sorry for generating so much traffic here... I think I am now one step
> away from the solution to this.
> This works:
>
> ${__javaScript(var xml_string = ''; for (var i=0;
> i<"${USER_ID_matchNr}"; i++) {xml_string +="${USER_ID_matchNr}";
> xml_string.toUpperCase();})}
>
> But now I cannot 'build' the USER_ID_[number] variable reference that I
> need to get to each of the elements... I tried this:
>
> ${__javaScript(var xml_string = ''; for (var i=0;
> i<"${USER_ID_matchNr}"; i++) {xml_string +=${__V(USER_ID_${i})};
> xml_string.toUpperCase();})}
>
> but when I execute this JMeter omits the sampler altogether. Do you know
> how I need to reference/'build' the variable string so that I get
> USER_ID_1, USER_ID_2 and so on ?

I've a similar situation, where I need to "build" the variable 
all.minifeed.timestamps_1
all.minifeed.timestamps_2, etc

I do it like :

${__BeanShell(vars.get("all.minifeed.timestamps_" + vars.get("count")))}

where count is generated from a Couter test element between 1 and 
${all.minifeed.timestamps_matchNr}

> Thanks for your help so far !
>
> hiro
>
> On Wed, 2009-02-18 at 13:56 +, sebb wrote:
> > On 18/02/2009, Hiro Protagonist  wrote:
> > > Ok, making some progress but hitting another block now: I can get to
> > > the number of matches contained within the RegEx result by referencing
> > > MYVARNAME_matchNr - however, ONLY when not calling it from within
> > > javascript. For quick and dirty debugging purposes, I put
> > >  ${USER_ID_matchNr} as the name of the sampler, and it correctly
> > > returns 8... however, when I use the same variable in a parameter for a
> > > Google search as part of the search parameter, it always returns
> > > 'undefined' ! (And yes, I've tried referencing it like
> > > ${USER_ID_matchNr},
> > >  '${USER_ID_matchNr}', "${USER_ID_matchNr}" - none of which work). It
> > >  must be a scoping issue rather than syntax - am I right ?
> >
> > A variable reference will either return the contents, or the reference
> > will be left unchanged.
> > If it returns 'undefined' then that is the value of the variable at that
> > point.
> >
> > If VAR is undefined, then ${VAR} will remain as ${VAR}
> >
> > Try using the Debug Sampler.
> >
> > >  On Sat, 2009-02-14 at 23:51 +, sebb wrote:
> > > > On 14/02/2009, Hiro Protagonist  
wrote:
> > >  > > Hi all,
> > >  > >
> > >  > >  I would like to know if any of you have an elegant solution for
> > >  > > this problem: I capture user IDs from an HTTP request, and by
> > >  > > using the -1 option on the RegEx extractor, get a collection of
> > >  > > all the IDs I need. So far, so good. In the next POST to the
> > >  > > server, I then need to 'embed' these IDs (which are integers) in
> > >  > > XML that is being sent with the request.
> > >  > >  The problem is this: if there are multiple users found (which
> > >  > > will always be the case) then a subsection of the XML simply gets
> > >  > > repeated, sort of like this:
> > >  > >
> > >  > >  [other XML tags]
> > >  > >  
> > >  > >  
> > >  > >  123
> > >  > >  
> > >  > >  
> > >  > >  456
> > >  > >  
> > >  > >  
> > >  > >  789
> > >  > >  
> > >  > >  
> > >  > >  [other XML tags]
> > >  > >
> > >  > >  The problem is that I need to construct the response dynamically
> > >  > >  according to how many IDs were found in the previous call - so
> > >  > > the number of elements will always differ. What would be the best
> > >  > > solution - write the IDs to file then read them and construct the
> > >  > > XML, using the Beanshell... ?
> > >  >
> > >  > If there is a limited range of user ids - e.g. min 1, max 10 - then
> > >  > you could set up separate samplers for each value of the count, and
> > >  > then select between them using the Switch Controller.
> > >  >
> > >  > This will obviously only be sensible if the range is limited.
> > >  >
> > >  > Otherwise, you will need to build the string using BeanShell or Jexl
> > >  > or Javascript.
> > >  >
> > >  > This could be done in a Post-Processor (after the RE) or in a
> > >  > Pre-Processor as a child of the POST sampler.
> > >  >
> > >  > Or you could use a function call within the POST sampler. This would
> > >  > be easiest with the BeanShell function, as you can code a method in
> > >  > a startup file, and just call that. For Jexl and Javascript you
> > >  > would need to include the code within the function call. Though you
> > >  > can store the script in a variable on the Test Plan and just pass
> > >  > the variable to the function. This makes the function call a bit
> > >  > easier to read.
> > >  >
> > >  > e.g. on Test Plan:
> > >  >
> > >  > SCRIPT print("hello")
> > >  >
> > >  > And in the sampler:
> > >  >
> > >  > ${__jexl(${SCRIPT})}
> > >  >
> > >  > [not tested, but you get the idea, I hope]
> > >  >
> > >  > For testing, I suggest you use save the HTML to a file and use the
> > >  > HTTP file: protocol to 

Re: Not able to record Swing JDialog Box with JMeter

[Noel O'Brien]

On Thursday 19 February 2009 06:39:53 Chhavi Sharma wrote:
> Thanks Sebb and David for the prompt reply.
> Some more clarifications for my question (in response to sebb):
>
> Dialog Box is launched by a jar file, downloaded earlier at client side.
> Then the login information is transferred to the server through TCP.
> I tried adding TCP sampler manually but the problem is that I am not able
> to figure out how to make dialog box recognizable by JMeter.
> JMeter is not recording the DialogBox, and I could not find any other
> option to do it manually.
>
> I will try Badboy too and see how it goes.
>
> Swing support feature is mentioned at this page:
> http://jakarta.apache.org/jmeter/
>
> What does it do?

I think this reference to swing is in regard to how JMeter is developed; the 
GUI is written using Swing. I don't think it means that JMeter can interact 
with Swing components of other applications. Correct me if I'm wrong?

Regards,
Noel

> Apache JMeter features include:
>
>- Can load and performance test many different server types:
>   - Web - HTTP, HTTPS
>   - SOAP
>   - Database via JDBC
>   - LDAP
>   - JMS
>   - Mail - POP3
>- Complete portability and * 100% Java purity * .
>- Full * Swing * and lightweight component support (precompiled JAR uses
>packages javax.swing.* ).
>
> Thanks once again guys.
>
> Thanks,
> -csn
>
> On Wed, Feb 18, 2009 at 8:55 PM, sebb  wrote:
> > On 18/02/2009, Chhavi Sharma  wrote:
> > > Hi,
> > >
> > >  I am recording a java webstart based application using JMeter Proxy.
> > > The steps involved are as follows:
> > >
> > >  1. Connect to webserver and download a .jnlp file
> > >  2. Download all the jars required to launch the application.
> > >  3. Then a swing dialog box is raised asking for the login information
> > > to
> >
> > the
> >
> > >  same server but at different port.
> > >  4. Once login is successful, download some more jars and launch the
> > >  application.
> > >
> > >  The problem is that JMeter is not recording the dialog box and no
> >
> > connection
> >
> > >  is displayed at the server side.
> >
> > JMeter can only record HTTP requests.
> >
> > >  However, it is recording all the downloading information, even the one
> > >  started after the login.
> >
> > However, the communication with the server is presumably all HTTP(S),
> > so it should be able do what ever the application does.
> >
> > If the playback does not work, then probably the login used HTTPS,
> > which is not recorded.
> >
> > >  Can someone please help me figuring out how to do it.
> >
> > You'll need to find out how the login works, e.g. using a protocol
> > analyser such as Wireshark or a browser add-on which shows the HTTP
> > requests. You can then add the missing sampler(s) to the test plan.
> >
> > Or you might be successful using Badboy - compare the exported samples
> > with the one generated by JMeter (e.g. open them in two instances of
> > JMeter).
> >
> > >  As I read on the
> > >  website, Swing based are supported by JMeter.
> >
> > Where does it say that JMeter supports Swing?
> >
> > >  Thanks in advance,
> > >  -csn
> >
> > -
> > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: If Controller failing

[Noel O'Brien]

On Thursday 19 February 2009 09:57:16 Massas wrote:
> Hello!
>
> I'm trying to use an If Controller but aparently I'm not understanding the
> function so i'm reaching out for help.
>
> I want to check whether one of my regexp extractors have found a match or
> not, so I'm trying to match the default value.
>
> Here's what i put in my if controller:
>
> "${getResponse2}" != "NoRegExpMatch2"
>
> Meaning that i want the controller to trigger if the getResponse2 does not
> have the default value.
>
> But here's the problem, i only get a fault in the logfile, and i cannot
> find the solution.
>
> 2009/02/19 10:18:51 ERROR - jmeter.control.IfController: unterminated
> string literal (#1) org.mozilla.javascript.EvaluatorException:
> unterminated string literal (#1)
>   at
> org.mozilla.javascript.DefaultErrorReporter.runtimeError(DefaultErrorReport
>er.java:98)

from "unterminated string literal (#1)" it looks like ${getResponse2} may 
not exist. Try putting a debug sampler before the IfController and see what 
variable values are assigned by the Regexp

Regards,
Noel

> what am i missing? It feels like i've tried with and without quotes,
> different type of quotes (' "), quoting the default value in the RegExp
> extractor etc, but still the same error.
>
>
> Would deeply appreciate some help here.
>
> Thanks in advance
>
> p.s. i know that the statement is true



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Reading links from file

[Noel O'Brien]

Hi there,

Have you looked at the Best Practices section:
http://jakarta.apache.org/jmeter/usermanual/best-practices.html

Specifically:

* Use non-GUI mode
* Use as few Listeners as possible.
* Rather than using lots of similar samplers, use the same sampler in a 
loop, and use variables (CSV Data Set) to vary the sample. Or perhaps use the 
Access Log Sampler. [The Include Controller does not help here, as it adds 
all the test elements in the file to the test plan.]
* Don't use functional mode
* Use CSV output rather than XML
* Only save the data that you need
* Use as few Assertions as possible

Sounds like you're loading the links from a file? In that case try using a CSV 
Data Set Config to load them 
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#CSV_Data_Set_Config

Then as the best practices article suggests, have possibly a single sampler in 
a loop, to execute each link

Regards
Noel


On Friday 20 February 2009 04:40:12 Tarun Jain wrote:
> Hi,
> I am new to JMeter and I am trying to do simple load test of a web server.
> I have a file with about 300 links (one on each line) that I would like to
> call sequentially.
>
> What is the best way to do this in JMeter?
>
> I have read quite a bit of documentation & searched this mailing list but
> could not find an answer. Any help is appreciated.
>
> Tarun
> -=-



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Controlling loop

[Noel O'Brien]

I had the exact same problem and found to only solution was to create my own 
counter element, with a corresponding reset element:



Counter (Beanshell Pre-precessor):

c = Integer.parseInt(vars.get("count"));
c++;
vars.put("count", ""+c);



Reset/Initializer (Beanshell Pre-precessor):

vars.put("count", "0");



Regards,
noel

On Friday 20 February 2009 13:49:43 Massas wrote:
> Hello again all.
>
> I seem to have more problems..
>
> I'm having problems with variables in an embedded loop.
>
> From an regexp i get a variable which controls how many times a loop should
> run. In this loop, i want to use another variable which starts at 1 (lets
> give it the origninal name of ${counter}) and increments for each embedded
> loop. However, if looping the entire test case i always want to enter this
> embedded loop with the ${counter} value of 1.
>
> Here i have my problem. When running once, it works like a charm. But if i
> run the whole test case more than one loop the ${counter} keeps
> incrementing (not restarting at 1)
>
> Tring to put it in a more visual way:
>
> Thread group
> + Regexp out a variable (ex 3)
> + loop regexp variable times with ${counter}=1
> ++ do stuff in loop
> ++ ${counter}++
> + End of test
>
> On the next loop of the whole testcase the ${counter} will enter with the
> value of 4, but i need it to restart at 1. Is there a way to do this?
>
> Thanks in advance, and have a good weekend all.



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Setting number of threads via property not possible ?

[Noel O'Brien]

On Monday 23 February 2009 07:48:51 Hiro Protagonist wrote:
> Hi sebb,
>
> You are correct, I meant to say 'thread groups' rather than threads. I
> managed to set the property, and pass it to the number of users field in
> the 2nd thread group - however, it only picks up the value every second
> time I execute the test plan.
> This has probably to do with the fact that I don't start the second
> thread group *after* the first one (they start simultaneously).
> When you say starting them consecutively, do you mean specifying a
> certain time at which the second thread group starts, using the
> scheduling mechanism ? Specifying a time is the only way I could think
> of controlling the startup behaviour of the thread groups...
> Please let me know if there is any other way.

In the "Test Plan" element, there's a checkbox for "Run Thread Groups 
consecutively"

Regards,
Noel

> Thanks for your help,
> hiro
>
> On Sun, 2009-02-22 at 22:31 +, sebb wrote:
> > On 22/02/2009, Hiro Protagonist  wrote:
> > > Hi all,
> > >
> > >  I tried to dynamically set the number of users for the second thread
> > > in my test plan (based on how many results from a query in the first
> > > thread I find in the first) by giving ${__property(NUMBER_OF_CODES)}
> > > which seems to set the property value correctly
> >
> > The function __property() does not set a property.
> >
> > Did you mean _setProperty()?
> >
> > >  - I can see the right number
> > >  appear when using the debug sampler - and I can use this anywhere in
> > > the first thread, but when I try to specify the number of users in the
> > > second thread (in the "Number of Threads (users):" field) the thread
> > > doesn't get started !
> >
> > I assume you mean you have two thread *groups* and you are trying to
> > use the result of running the first thread group to control the number
> > of threads in the second group.
> >
> > Is that correct?
> >
> > >  Can this be done ? If so, how ?
> >
> > Works fine for me.
> >
> > Of course, you have to set the thread groups to run consecutively so
> > that the first group completes before the second one starts.
> >
> > >  Thanks,
> > >  hiro
> > >
> > >
> > >  -
> > >  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > >  For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
> >
> > -
> > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Problems with User logging in and creating contents when testing a web application

[Noel O'Brien]

On Tuesday 24 February 2009 06:28:53 MetalFang wrote:
> Hi all,
>
> My test plan will do the following,
>
> 1) Log in as a USER.
>
> 2) Go to update some personal information in that account.
>
> 3) Update the information
>
> 4) Log out
>
> The problem I encounter is when I change the USER parameter into another
> account(beside the one I've recorded), I am able to log in but not able to
> update the information, is anyone able help me? Thanks in advance

When you say you're "not able" to update the information I presume you mean 
the sample fails? If so, what are the failure details? 

Also, what kind of samplers are you using to make the requests? (there are 
three HTTP samplers..)

Regards,
Noel

-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Reg Expression & If Controller

[Noel O'Brien]

On Tuesday 24 February 2009 15:59:30 Woodcock, Simon wrote:
> I want to be able to do the following:
>
> If a regular expression is not found do not execute the next two nodes.
>
> First of all if a regular expression is not found how does the variable get
> populated? null, ""?

Try using the "refName_matchNr" variable created by the REGEX:
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Regular_Expression_Extractor

If there's no match it will be 0, so IFController will be:
"refName_matchNr" == "0"
or something like that

> Secondly using the 'If Controller' I want to be able to add the following
> condition:
>
> ${regex}!=null
>
> Where null is replaced by the value that gets populated by a regular
> expression find failure.

similar to above, try using its negative:

"refName_matchNr" != "0"

> I've tried a number of scenarios ( incl trying to use the Regular
> Expression Extractor default value ), but whatever I try fails to execute
> the If Controller as expected.
>
> Any help or examples would be greatly appreciated,
>
> Thanks in advance,
>
>
> Simon.
>
> --
> Simon Woodcock
> Staff Engineer / Manager
>
> QUALCOMM Poole Limited
> Voyager House, 253-257 High Street North, Poole, Dorset, BH15 1DX UK
> Tel: +44 (0)1202 440610, Fax: +44 (0)1202 440601
> www.qualcomm.com
> This e-mail transmission, and any documents, files or previous e-mail
> messages attached to it, may contain confidential information that is
> legally privileged. If you are not the intended recipient, or a person
> responsible for delivering it to the intended recipient, you are hereby
> notified that any disclosure, copying, distribution or use of any of the
> information contained in or attached to this message is STRICTLY
> PROHIBITED. Interception of e-mail is a crime under the Electronic
> Communications Privacy Act, 18 U.S.C. 2510-2521 and 2107-2709. If you have
> received this transmission in error, please immediately notify me by reply
> e-mail at simon.woodc...@qualcomm.com
> or by telephone at +44 (0)1202 207397 and destroy the original transmission
> and its attachments without reading them or saving them to disk. Thank you.



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Reg Expression & If Controller

[Noel O'Brien]

On Tuesday 24 February 2009 17:23:09 Woodcock, Simon wrote:
> Thanks for the response.
>
> I'm little confused by the syntax used within the 'If Controllers'
> condition field.
>
> If my regular expression extractor reference name is regex, following your
> example would my condition read:
>
> "regex_matchNr" != "0"

Apologies for my mistake, it should be:
"${regex_matchNr}" != "0"

See the examples here:
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#If_Controller

> Including all the "'s?
>
> Wouldn't this always return true?
>
> Apologies if I'm misunderstanding.
>
> Thanks again.
>
>
> Simon.
> -Original Message-
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: 24 February 2009 16:21
> To: jmeter-user@jakarta.apache.org
> Cc: Woodcock, Simon
> Subject: Re: Reg Expression & If Controller
>
> On Tuesday 24 February 2009 15:59:30 Woodcock, Simon wrote:
> > I want to be able to do the following:
> >
> > If a regular expression Is not found do not execute the next two nodes.
> >
> > First of all if a regular expression is not found how does the variable
> > get populated? null, ""?
>
> Try using the "refName_matchNr" variable created by the REGEX:
> http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Regula
>r_Expression_Extractor
>
> If there's no match it will be 0, so IFController will be:
> "refName_matchNr" == "0"
> or something like that
>
> > Secondly using the 'If Controller' I want to be able to add the following
> > condition:
> >
> > ${regex}!=null
> >
> > Where null is replaced by the value that gets populated by a regular
> > expression find failure.
>
> similar to above, try using its negative:
>
> "refName_matchNr" != "0"
>
> > I've tried a number of scenarios ( incl trying to use the Regular
> > Expression Extractor default value ), but whatever I try fails to execute
> > the If Controller as expected.
> >
> > Any help or examples would be greatly appreciated,
> >
> > Thanks in advance,
> >
> >
> > Simon.
> >
> > --
> > Simon Woodcock
> > Staff Engineer / Manager
> >
> > QUALCOMM Poole Limited
> > Voyager House, 253-257 High Street North, Poole, Dorset, BH15 1DX UK
> > Tel: +44 (0)1202 440610, Fax: +44 (0)1202 440601
> > www.qualcomm.com<http://www.qualcomm.com>
> > This e-mail transmission, and any documents, files or previous e-mail
> > messages attached to it, may contain confidential information that is
> > legally privileged. If you are not the intended recipient, or a person
> > responsible for delivering it to the intended recipient, you are hereby
> > notified that any disclosure, copying, distribution or use of any of the
> > information contained in or attached to this message is STRICTLY
> > PROHIBITED. Interception of e-mail is a crime under the Electronic
> > Communications Privacy Act, 18 U.S.C. 2510-2521 and 2107-2709. If you
> > have received this transmission in error, please immediately notify me by
> > reply e-mail at
> > simon.woodc...@qualcomm.com<mailto:simon.woodc...@qualcomm.com> or by
> > telephone at +44 (0)1202 207397 and destroy the original transmission and
> > its attachments without reading them or saving them to disk. Thank you.



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Problems with User logging in and creating contents when testing a web application

[Noel O'Brien]

On Wednesday 25 February 2009 00:46:34 MetalFang wrote:
> Noel O'Brien wrote:
> > On Tuesday 24 February 2009 06:28:53 MetalFang wrote:
> >> Hi all,
> >>
> >> My test plan will do the following,
> >>
> >> 1) Log in as a USER.
> >>
> >> 2) Go to update some personal information in that account.
> >>
> >> 3) Update the information
> >>
> >> 4) Log out
> >>
> >> The problem I encounter is when I change the USER parameter into another
> >> account(beside the one I've recorded), I am able to log in but not able
> >> to
> >> update the information, is anyone able help me? Thanks in advance
> >
> > When you say you're "not able" to update the information I presume you
> > mean
> > the sample fails? If so, what are the failure details?
> >
> > Also, what kind of samplers are you using to make the requests? (there
> > are three HTTP samplers..)
> >
> > Regards,
> > Noel
> >
> > -
> > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> Silly me, sorry for the lack of information, I am new to jmeter.
>
> I am using HTTP request.
> Sorry to ask but how do i see the failure details?

You need to add some listeners to your test plan. "View Results Tree" Listener 
will show you lots of detail.

http://jakarta.apache.org/jmeter/usermanual/component_reference.html#View_Results_Tree

Regards,
Noel

> Thanks for the reply though.



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

BackingStoreException

[Noel O'Brien]

Hi All,

I'm just wondering if anyone knows what causes the following warning while 
JMeter is running and if anything can be done to stop it. It's not causing 
any problems, just clogging up my console :)

Feb 25, 2009 1:02:49 PM java.util.prefs.FileSystemPreferences syncWorld
WARNING: Couldn't flush user prefs: java.util.prefs.BackingStoreException: 
java.lang.IllegalArgumentException: Not supported: indent-number

Regards,
Noel

-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: BackingStoreException

[Noel O'Brien]

On Wednesday 25 February 2009 14:31:07 sebb wrote:
> On 25/02/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  I'm just wondering if anyone knows what causes the following warning
> > while JMeter is running and if anything can be done to stop it. It's not
> > causing any problems, just clogging up my console :)
> >
> >  Feb 25, 2009 1:02:49 PM java.util.prefs.FileSystemPreferences syncWorld
> >  WARNING: Couldn't flush user prefs:
> > java.util.prefs.BackingStoreException:
> > java.lang.IllegalArgumentException: Not supported: indent-number
>
> This is a Java bug in certain implementations.

In certain implementations of JMeter or JVM? I'm using JMeter 2.3.2, Sun JDK 
1.6.0_03

Regards,
Noel

> >  Regards,
> >  Noel
> >
> >  -
> >  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Paralell running ThreadGroups

[Noel O'Brien]

On Wednesday 25 February 2009 15:40:35 aldana wrote:
> Hi,
>
> What I want to achieve is:
> -Hitting a certain URL with high load (20 threads, looping)
> -Hitting another URL every 20 seconds in parallel to the high load requests
>
> I tried to do this by applying two ThreadGroups as siblings (1st for heavy
> load looping, 2nd for scheduled firing every 20 seconds).
>
> But this doesn't work, JMeter is executing these ThreadGroups sequentially
> and not in parallel. I got a workaround, starting two JMeter instances but
> I guess this can be done easier...

Is the "Run Thread Groups consecutively" box ticked in the test plan?

Regards,
Noel

> thanks.
>
>
>
> -
> manuel aldana
> aldana((at))gmx.de
> software-engineering blog: http://www.aldana-online.de



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

When are paramerters/variables resolved?

[Noel O'Brien]

Hi,

I'm wondering at what point does JMeter resolve all variables (or 
parameters? :) ) stored in a User Defined Variables element?

I'm using serveral JMS samplers (point-to-point) and I'm using the same 
timeout value for each one, so for convenience I would like to define it in a 
User Defined Variables element and reference it in the Timeout field of the 
JMS Sampler as ${timeout}.

The problem is that when the GUI stores the data to the test element, or it is 
reloaded (i.e. once transfer() or configure() is called) the default timeout 
value (2000) is stored in the test element and displayed in the GUI 

I've debugged through it and the cause is that the getTimeout() method in 
JMSSampler.java is trying to return the value as an int, but as in my case 
timeout is ${timeout}, it is not resolved and the default value is returned 
instead.

Is this a bug? 

What can I do in the meantime to work around it? I see two approaches:
1. Modify getTimout() to return a string
2. Somehow resolve ${timeout} to its value

What are the ramifications of either of these approaches?

Thanks,
Noel

-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: When are paramerters/variables resolved?

[Noel O'Brien]

On Thursday 26 February 2009 16:51:28 Steve Kapinos wrote:
> Where are the UDVs compared to your samplers?  They are not valid until
> that element has been processed by the test plan.

They are defined at the very top of my test plan

> You can use properties which would be available from the start.

I don't even get to start my test plan, the value in the Timeout field of 
overridden once you select a different GUI element

It's easily reproducible:
Create new test plan
Add thread group
Add JMS Point-to-Point Sampler
Add non-numberic value to Timeout field
Save, or 
Select a different GUI element (e.g. thread group) then go back to the JMS 
Sampler

you will see the value has changed to "2000"

Regards,
Noel

> -----Original Message-
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: Thursday, February 26, 2009 10:30 AM
> To: jmeter-user@jakarta.apache.org
> Subject: When are paramerters/variables resolved?
>
> Hi,
>
> I'm wondering at what point does JMeter resolve all variables (or
> parameters? :) ) stored in a User Defined Variables element?
>
> I'm using serveral JMS samplers (point-to-point) and I'm using the same
> timeout value for each one, so for convenience I would like to define it
> in a
> User Defined Variables element and reference it in the Timeout field of
> the
> JMS Sampler as ${timeout}.
>
> The problem is that when the GUI stores the data to the test element, or
> it is
> reloaded (i.e. once transfer() or configure() is called) the default
> timeout
> value (2000) is stored in the test element and displayed in the GUI
>
> I've debugged through it and the cause is that the getTimeout() method
> in
> JMSSampler.java is trying to return the value as an int, but as in my
> case
> timeout is ${timeout}, it is not resolved and the default value is
> returned
> instead.
>
> Is this a bug?
>
> What can I do in the meantime to work around it? I see two approaches:
> 1. Modify getTimout() to return a string
> 2. Somehow resolve ${timeout} to its value
>
> What are the ramifications of either of these approaches?
>
> Thanks,
> Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org



-
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Using the __V function in an HTTP sampler

[Noel O'Brien]

Maybe have a look at the For Each Controller instead of the loop controller; 
it's more suited to looping over variables in the format that your variables 
are in (var_n)

If that doesn't work for you, you can try something like this (I use this and 
it works);

${__BeanShell(vars.get("sns.testdata." + vars.get("current.sns") + ".user"))}

which takes the value of "current.sns" e.g. 'BB' and when resolved, the value 
for the variable 'sns.testdata.BB.user' is returned successful

Regards,
Noel

On Thursday 05 March 2009 19:02:12 Jason Frank wrote:
> I'm trying to accomplish something relatively basic, and getting stuck.  I
> want to use the __V() function in order to get the value of a variable,
> whose name I have to generate from another varible.  (I've seen this
> referred to as a "nested variable".)
>
> I have an HTTP Sampler inside of a Loop controller, and I want to send the
> evaluated variable's value as one of the request params.  In the section
> titled "Send Parameters With the Request", I add a parameter with name foo,
> and value of ${__V(gameId_${C})}, where C is the name of a counter
> variable.  What I want to happen is that in the first iteration, the param
> foo gets the value of the evaluation of variable ${gameId_1}, then in the
> next iteration around the loop it would use the evaluation of ${gameId_2}
> and so forth.
>
> When I look at the HTTP Sampler request, what I see is
> foo=${__V(gameId_${C})}. This implies that the function is not running at
> all.  Perhaps I am confused about where to put function calls, but the
> reference says they can go just about anywhere in the test.  I also tried
> setting a different parameter's value to another function, ${__time(YMD)},
> just to see if it was something particular to the variables I was trying to
> use.  This also resulted in the literal ${__time(YMD)} being included in
> the request, rather than the function's result.
>
> If anyone can help me out, I would appreciate it.
>
>
>
>
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Simulating QoS and Speed

[Noel O'Brien]

Hi All,

I've been asked to simulate real traffic for our system using JMeter in 
relation 
to file uploads in terms of Quality of Service (approx 60%) and realistic 
upload speeds. 

Quality of Service seems reasonably possible with JMeter as its a matter of 
structuring samplers under a number of random and/or throughput controllers.

Simulating realistic upload speeds seems much harder. I'm told that clients 
will upload files to our server at approx 5 - 20 kbps, and to simaulte this 
correctly, I'll need to throttle the connection the sampler makes, rather than 
adding a delay timer. Is there a way to simulate this in JMeter? Nothing jumps 
out at me and I don't recall seeing anything in the User Manual to achieve 
this. Has anyone else had to deal with this situation? Does anyone have any 
thoughts on how this could be achieved?

Regards,
Noel

Re: Simulating QoS and Speed

[Noel O'Brien]

On Monday 09 March 2009 16:37:13 sebb wrote:
> On 09/03/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  I've been asked to simulate real traffic for our system using JMeter in
> > relation to file uploads in terms of Quality of Service (approx 60%) and
> > realistic upload speeds.
> >
> >  Quality of Service seems reasonably possible with JMeter as its a matter
> > of structuring samplers under a number of random and/or throughput
> > controllers.
> >
> >  Simulating realistic upload speeds seems much harder. I'm told that
> > clients will upload files to our server at approx 5 - 20 kbps, and to
> > simaulte this correctly, I'll need to throttle the connection the sampler
> > makes, rather than adding a delay timer. Is there a way to simulate this
> > in JMeter? Nothing jumps out at me and I don't recall seeing anything in
> > the User Manual to achieve this. Has anyone else had to deal with this
> > situation? Does anyone have any thoughts on how this could be achieved?
>
> jmeter.properties:

Excellent, I'd been using the Java HTTP Request sampler, so had missed this.

> # Define characters per second > 0 to emulate slow connections
> #httpclient.socket.http.cps=0
> #httpclient.socket.https.cps=0
>
> Only works with HttpClient version of HTTP sampler.

Does characters per second directly relate to speed? e.g. 2 CPS  == 20kbps

Regards,
Noel

> >  Regards,
> >
> > Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Simulating QoS and Speed

[Noel O'Brien]

On Monday 09 March 2009 17:34:17 sebb wrote:
> On 09/03/2009, Noel O'Brien  wrote:
> > On Monday 09 March 2009 16:37:13 sebb wrote:
> >  > On 09/03/2009, Noel O'Brien  wrote:
> >  > > Hi All,
> >  > >
> >  > >  I've been asked to simulate real traffic for our system using
> >  > > JMeter in relation to file uploads in terms of Quality of Service
> >  > > (approx 60%) and realistic upload speeds.
> >  > >
> >  > >  Quality of Service seems reasonably possible with JMeter as its a
> >  > > matter of structuring samplers under a number of random and/or
> >  > > throughput controllers.
> >  > >
> >  > >  Simulating realistic upload speeds seems much harder. I'm told that
> >  > > clients will upload files to our server at approx 5 - 20 kbps, and
> >  > > to simaulte this correctly, I'll need to throttle the connection the
> >  > > sampler makes, rather than adding a delay timer. Is there a way to
> >  > > simulate this in JMeter? Nothing jumps out at me and I don't recall
> >  > > seeing anything in the User Manual to achieve this. Has anyone else
> >  > > had to deal with this situation? Does anyone have any thoughts on
> >  > > how this could be achieved?
> >  >
> >  > jmeter.properties:
> >
> > Excellent, I'd been using the Java HTTP Request sampler, so had missed
> > this.
> >
> >  > # Define characters per second > 0 to emulate slow connections
> >  > #httpclient.socket.http.cps=0
> >  > #httpclient.socket.https.cps=0
> >  >
> >  > Only works with HttpClient version of HTTP sampler.
> >
> > Does characters per second directly relate to speed? e.g. 2 CPS  ==
> > 20kbps
>
> CPS is a measure of speed.
>
> Whether 1 CPS == 10 bps I don't know; sounds about right.

I can confirm that there is an almost (in not) direct mapping under the 
following situation:

JPEG file  62356 bytes in size
httpclient.socket.http.cps=1
sample time for upload: 6567ms

6.567s * 1cps = 65670c ~= 62356b

Regards,
Noel

> >  Regards,
> >  Noel
> >
> >  > >  Regards,
> >  > >
> >  > > Noel
> >  >
> >  > -
> >  > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >  > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Simulating QoS and Speed

[Noel O'Brien]

On Monday 09 March 2009 18:32:03 da...@davidwbrown.name wrote:
> You might want to try removing JMeter the fat client from the actual load
> testing with the following:

What do you mean by "fat client"? The GUI?

> http://jakarta.apache.org/jmeter/usermanual/jmeter_distributed_testing_step
>_by_step.pdf
>
> My experience with this is more realistic. JMeter at the command-line is
> more work but more control: beanshell et. al.
>
> Noel O'Brien wrote ..
>
> > Hi All,
> >
> > I've been asked to simulate real traffic for our system using JMeter in
> > relation to file uploads in terms of Quality of Service (approx 60%) and
> > realistic upload speeds.
> >
> > Quality of Service seems reasonably possible with JMeter as its a matter
> > of structuring samplers under a number of random and/or throughput
> > controllers.
> >
> > Simulating realistic upload speeds seems much harder. I'm told that
> > clients will upload files to our server at approx 5 - 20 kbps, and to
> > simaulte this correctly, I'll need to throttle the connection the sampler
> > makes, rather than adding a delay timer. Is there a way to simulate this
> > in JMeter? Nothing jumps out at me and I don't recall seeing anything in
> > the User Manual to achieve this. Has anyone else had to deal with this
> > situation? Does anyone have any thoughts on how this could be achieved?
> >
> > Regards,
> > Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Specifying Variables on the Command Line

[Noel O'Brien]

Hi All,

Is it possible to specify variables (User Defined Variables, e.g.) on the 
command line in the same way as specifying properties?

I currently specify the host and port of the server I'm testing in a User 
Defined Variables config element in the JMX file but would like to override 
this 
from the command line. Is it possible?

Regards,
Noel

Re: Specifying Variables on the Command Line

[Noel O'Brien]

On Wednesday 11 March 2009 12:56:18 sebb wrote:
> On 11/03/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  Is it possible to specify variables (User Defined Variables, e.g.) on
> > the command line in the same way as specifying properties?
>
> No. Only Java and JMeter properties can be specified on the command-line.
>
> >  I currently specify the host and port of the server I'm testing in a
> > User Defined Variables config element in the JMX file but would like to
> > override this from the command line. Is it possible?
>
> Yes.
>
> http://wiki.apache.org/jakarta-jmeter/JMeterFAQ#head-81e8fdc6d08e5b574a0655
>6f597f788df8ed12fe

Cool, that worked, thanks!

Regards,
Noel

> >  Regards,
> >
> > Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: timer between calls

[Noel O'Brien]

Add a constant timer element to the top level of the thread group

Regards,
Noel

On Thursday 12 March 2009 15:52:11 Jose Castro wrote:
> If I have a sequence of SOAP calls using only one thread. How can I
> put a timer between each call of 2 seconds(for example)
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Stress Testing in a Virtual Server

[Noel O'Brien]

Hi All,

I've been running some load testing in a virtual server (Xen) as that's where 
our product will be deployed. I've noticed some unusual behavior and I can't 
seem to be able to pinpoint what's going wrong. 

Basically, I start 1 JMeter instance running 10 threads against the product (a 
http REST server in Tomcat). It connects to several auxillary programs (DB, 
other HTTP servers, SwiftMQ, etc), all of which are on different virtual 
machines within the same subnet. I'm using top and lsof to monitor the 
resources of each running application. Virtualised OS is RHEL 5, running 
Tomcat 6.0.16 and Java 1.6.0_07

The problem is that every few mins (about 3) the CPU usage of all the 
processes I'm monitoring (product, JMeter, DB, etc.) all drop to 0% for about  
30 seconds. None of the logs suggest that anything is wrong, so it's 
incredibly hard to track down what's causing this. I'm thinking that this 
issue is caused by the virtual servers upon which the programs are running, 
perhaps mis- / mal-configuration of the virtual network cards.

Has anyone else experienced anything like this while stress testing in a 
virtualised environment? 

Regards,
Noel

Re: Stress Testing in a Virtual Server

[Noel O'Brien]

On Monday 16 March 2009 12:01:53 sebb wrote:
> On 16/03/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  I've been running some load testing in a virtual server (Xen) as that's
> > where our product will be deployed. I've noticed some unusual behavior
> > and I can't seem to be able to pinpoint what's going wrong.
> >
> >  Basically, I start 1 JMeter instance running 10 threads against the
> > product (a http REST server in Tomcat). It connects to several auxillary
> > programs (DB, other HTTP servers, SwiftMQ, etc), all of which are on
> > different virtual machines within the same subnet. I'm using top and lsof
> > to monitor the resources of each running application. Virtualised OS is
> > RHEL 5, running Tomcat 6.0.16 and Java 1.6.0_07
> >
> >  The problem is that every few mins (about 3) the CPU usage of all the
> >  processes I'm monitoring (product, JMeter, DB, etc.) all drop to 0% for
> > about 30 seconds. None of the logs suggest that anything is wrong, so
> > it's incredibly hard to track down what's causing this. I'm thinking that
> > this issue is caused by the virtual servers upon which the programs are
> > running, perhaps mis- / mal-configuration of the virtual network cards.
> >
> >  Has anyone else experienced anything like this while stress testing in a
> >  virtualised environment?
>
> What happens to I/O during this time?

I'm not sure. lsof blocks, so running it during this time is fruitless. Is 
there another command I can run to check I/O?

> Do top and lsof continue responding OK?

Top continues to respond (unless it blocks and shows 0% CPU for the duration), 
but lsof blocks until the CPU appears to be given back to the processes.

> >  Regards,
> >
> > Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Using UDV in SOAP Request

[Noel O'Brien]

On Wednesday 18 March 2009 03:31:41 sebb wrote:
> On 18/03/2009, Mark Musone  wrote:
> > Ok, so here's my dilema..hopefully someone can give me some suggestions
> > on a high level what would make the most sense for JMeter to do:
> >  I've got literally about 1000 different SOAP/XML requests, each one
> > being different enough to warrent being in their own file. For example,
> > I've got a general SOAP request to "Store a Document", Well the document
> > can be very many different types (PDF, word, plain text..etc..), each
> > type has their own properties So i've got 100 or so different "Store a
> > Document" SOAP/XML files, each one storing a different type.
>
> I find it a bit surprising that there are so many different SOAP requests.
> Is there no way to parameterize these and use fewer requests?
>
> In the limit, I suppose one could have a variable for the entire SOAP
> body, and just read that  from a CSV file, but that would be very
> awkard to maintain.
>
> Can you not find more similarities that would allow one to use small
> chunks of text as variables?
>
> >  However, for each document "type", there are properties i need to be
> >  variable (I want a random word document name, or  if it's a JPEG image
> > the height and width should be random..etc..)
> >
> >  My original idea was going to hopefully be simple in such that in each
> > of these 1,000 XML files, i'd just reference a ${variable} whenever I
> > wanted dynamic info..all of these XML files themselves would simply be
> > pointed to by the "Message Folder" attribute, and JMeter'll magically
> > pick any of the 1,000 XML files and magically do the variable
> > substitution.
> >
> >  Now that I understand it's not as cut and dry, with JMeter not
> > processing external files, is there an JMeter-best-practices way to
> > accomplish it?
>
> It's best to reduce the amount of processing that JMeter has to do at
> run-time, so if you can create the personalised files outside JMeter,
> that would be best.
>
> For example, when I first started using JMeter, I needed to generate
> many thousands of randomly orderd unique account ids. That would have
> required a lot of memory to keep track of which ones had been used, so
> I created many smaller batches of accound ids, and used those in turn
> (with the StringFromFile function, which predates CSV Dataset).
>
> >  I believe I can most likely have 1,000 SOAP request samplers (probably
> >  inside a random controller) and copy and paste each of these XML files
> > in each sampler and have the variable substitution work...I guess that's
> > the brute force way..I'm just not sure if there's a better option. I
> > probably just need a pointer from the pro's of "Try this
> > function/config"..
>
> If you are prepared to write some Java, then you could create your own
> copy of the SOAP/XML-RPC sampler to process the files on the fly.
>
> Or you could perhaps use a BeanShell or BSF Pre-Processor to generate
> the contents for the sampler. Or create a BeanShell function that
> returns the SOAP body. Again, coding is needed.

Or, you could store the SOAP requests in files like you're doing, but use 
placeholders where variables are needed. Then use a pre-processor to 
substitute the placeholders for variable data

Regards,
Noel

> >  Thanks,
> >
> > Mark
> >
> >  On Tue, Mar 17, 2009 at 11:04 AM, sebb  wrote:
> >  > On 17/03/2009, Mark Musone  wrote:
> >  > > Yep, I think you hit the nail on the head..they should rename it to
> >  > > "User Define Constant" or "User Defined Global"!
> >  >
> >  > This has been made clearer in the documentation.
> >  >
> >  > Once the Test Plan and all UDVs have been processed, the resulting set
> >  > of variables is copied to each thread to provide the initial set of
> >  > variables.
> >  >
> >  > I.e. UDVs create regular JMeter variables, but only at test startup.
> >  >
> >  > >  ok, so if i understand it correctly,I basically need to put
> >  > > whatever function call I plan on using directly in the XML Data..
> >  >
> >  > Yes, or pre-create the files and use a variable file name.
> >  >
> >  > >  Will any of the methods you mention below work on external files,
> >  > > or am
> >  >
> >  > i
> >  >
> >  > >  still screwed with being able to dynamically change anything in an
> >  >
> >  > external
> >  >
> >  > >  file?
> >  > >  Thanks,
> >  > >
> >  > >
> >  > >  Mark
> >  > >
> >  > >
> >  > >
> >  > >  On Tue, Mar 17, 2009 at 10:50 AM, Ian Blavins
> >  > > 
> >  >
> >  > wrote:
> >  > >  > G'day
> >  > >  >
> >  > >  > Despite the name User Defined Variables are in fact (global) user
> >  >
> >  > defined
> >  >
> >  > >  > constants. UDVs are evaluated once prior to the run. Each UDV
> >  > >  > will
> >  >
> >  > take the
> >  >
> >  > >  > final value seen in the test plan and will not change at run
> >  > >  > time.
> >  > >  >
> >  > >  > You can use UDVs to initialize local variables but it is the
> >  > >  > local variables whose values you have to modify 

Re: Stress Testing in a Virtual Server

[Noel O'Brien]

Hi All,

After much investigation and help from our ops  department, it was found that 
the ip_conntrack buffer on the host OS was filling up under load tests. 

Disabling iptables on the host (/etc/init.d/iptables stop) resolved the issue 
and did not seem to impact the hosted VMs.

Regards,
Noel

On Monday 16 March 2009 12:32:54 sebb wrote:
> On 16/03/2009, Noel O'Brien  wrote:
> > On Monday 16 March 2009 12:01:53 sebb wrote:
> > > On 16/03/2009, Noel O'Brien  wrote:
> > > > Hi All,
> > > >
> > > >
> > > >
> > > > I've been running some load testing in a virtual server (Xen) as
> > > > that's
> > > >
> > > > where our product will be deployed. I've noticed some unusual
> > > > behavior
> > > >
> > > > and I can't seem to be able to pinpoint what's going wrong.
> > > >
> > > >
> > > >
> > > > Basically, I start 1 JMeter instance running 10 threads against the
> > > >
> > > > product (a http REST server in Tomcat). It connects to several
> > > > auxillary
> > > >
> > > > programs (DB, other HTTP servers, SwiftMQ, etc), all of which are on
> > > >
> > > > different virtual machines within the same subnet. I'm using top and
> >
> > lsof
> >
> > > > to monitor the resources of each running application. Virtualised OS
> > > > is
> > > >
> > > > RHEL 5, running Tomcat 6.0.16 and Java 1.6.0_07
> > > >
> > > >
> > > >
> > > > The problem is that every few mins (about 3) the CPU usage of all the
> > > >
> > > > processes I'm monitoring (product, JMeter, DB, etc.) all drop to 0%
> > > > for
> > > >
> > > > about 30 seconds. None of the logs suggest that anything is wrong, so
> > > >
> > > > it's incredibly hard to track down what's causing this. I'm thinking
> >
> > that
> >
> > > > this issue is caused by the virtual servers upon which the programs
> > > > are
> > > >
> > > > running, perhaps mis- / mal-configuration of the virtual network
> > > > cards.
> > > >
> > > >
> > > >
> > > > Has anyone else experienced anything like this while stress testing
> > > > in a
> > > >
> > > > virtualised environment?
> > >
> > > What happens to I/O during this time?
> >
> > I'm not sure. lsof blocks, so running it during this time is fruitless.
> > Is there another command I can run to check I/O?
>
> Doesn't top show I/O?
>
> Also, try running ping to each of the systems (ideally from a non-Xen
> system).
>
> This should help show if it is a single system locking up and causing
> the others to wait for it.
>
> Are all the Xen servers on the same host?
>
> > > Do top and lsof continue responding OK?
> >
> > Top continues to respond (unless it blocks and shows 0% CPU for the
> > duration), but lsof blocks until the CPU appears to be given back to the
> > processes.
> >
> > > > Regards,
> > > >
> > > >
> > > >
> > > > Noel
> >
> > -
> >
> > > To unsubscribe, e-mail:
> >
> > jmeter-user-unsubscr...@jakarta.apache.org
> >
> > > For additional commands, e-mail:
> >
> > jmeter-user-h...@jakarta.apache.org

Accuracy of Elapsed Time

[Noel O'Brien]

Hi,

I'm running some load testing, nothing too heavy. 10 JMeter threads, ramp up 
is 30. JMeter is run from the command line and the test plan has no listeners. 
using the -l flag, I capture the results to a CSV file.

I then load the results file into JMeter and look at hem from a Summary 
Listener, View Tree Listener etc. I'm a little confused by the elapsed time 
results. 

The app under test has the ability to log the time that's taken to receive the 
request and complete sending the response, which for an auth API call it lists 
as taking 114ms for example (ronded to the closest ms)

I've set up tcpdump on the machine running the app (refer to as server 
henceforth) and the machine running  JMeter (refer to as client henceforth). 

tcpdump on the server reports that the time between the request and response 
packets to be 113.83ms, while tcpdump on the client reports that the time 
between the request and response packets 113.96ms

However for that particular HTTP request, JMeter reports that the load time 
(elasped time) is 174ms and that the latency is 174 ms also. How is this 
discrepancy explained?

As I understand it, the latency is the time taken to receive the first byte of 
the response and since the load time and latency are the same then it 
indicated that the response payload was received within 1 ms. Is this correct? 
Or is it even possible to achieve this? The payload size for this response is 
106 bytes.

From the tcpdump on the client, it's clear that the response packets are 
available to JMeter after 113ms so I'm concerned about what's causing the 
increase.

FWIW, the auth request sampler has an XPath Extracter Post processor, but 
processing time for that's not included in the load time right??. All HTTP 
Samplers are Java, not HTTPClient.

Regards,
Noel

Re: Accuracy of Elapsed Time

[Noel O'Brien]

On Thursday 26 March 2009 13:00:02 sebb wrote:
> On 26/03/2009, Noel O'Brien  wrote:
> > Hi,
> >
> >  I'm running some load testing, nothing too heavy. 10 JMeter threads,
> > ramp up is 30. JMeter is run from the command line and the test plan has
> > no listeners. using the -l flag, I capture the results to a CSV file.
>
> Good.
>
> JMeter version?

JMeter 2.3.2:r665936
Java: Sun 1.6.0_06-b02
OS: RHEL 5.2 running in RHEL hypervisor

> >  I then load the results file into JMeter and look at hem from a Summary
> >  Listener, View Tree Listener etc. I'm a little confused by the elapsed
> > time results.
> >
> >  The app under test has the ability to log the time that's taken to
> > receive the request and complete sending the response, which for an auth
> > API call it lists as taking 114ms for example (ronded to the closest ms)
> >
> >  I've set up tcpdump on the machine running the app (refer to as server
> >  henceforth) and the machine running  JMeter (refer to as client
> > henceforth).
> >
> >  tcpdump on the server reports that the time between the request and
> > response packets to be 113.83ms, while tcpdump on the client reports that
> > the time between the request and response packets 113.96ms
> >
> >  However for that particular HTTP request, JMeter reports that the load
> > time (elasped time) is 174ms and that the latency is 174 ms also. How is
> > this discrepancy explained?
>
> There is overhead in the OS, Java and JMeter on both sending and receiving.

Some of the other API calls I make only show a difference of a few ms in JMeter 
compared to the app/tcpdump

> Note also that the timer resolution will affect the reported elapsed time.
> However if you are running 2.3.2+ on Java 1.5+ it will use a higher
> resolution timer.

Does it matter that the auth API call is the first call for each thread? (Is 
the thread doing extra work at it's start-up that would impact the time?)

> >  As I understand it, the latency is the time taken to receive the first
> > byte of the response and since the load time and latency are the same
> > then it indicated that the response payload was received within 1 ms. Is
> > this correct? Or is it even possible to achieve this? The payload size
> > for this response is 106 bytes.
>
> Latency is time to first response.
> This may be the entire response, especially for small payloads.

Ah, I see :)

> >  From the tcpdump on the client, it's clear that the response packets are
> >  available to JMeter after 113ms (*) so I'm concerned about what's
> > causing the increase.
>
> (*) The OS (and Java) have to process the request before tcpdump sees
> it, and likewise after tcpdump sees the response.

Hmmm, see above re: other API calls.

> >  FWIW, the auth request sampler has an XPath Extracter Post processor,
> > but processing time for that's not included in the load time right??. All
> > HTTP Samplers are Java, not HTTPClient.
>
> Post-Processors are not included in individual sample-times.

Good :)

> The HttpSampler does as little processing as possible whilst timing
> the sample, but it has to issue the connect - retrying if necessary -
> then send the data and process the response.
>
> The connection time is not currently measured separately.
>
> I don't know why the times differ by as much as you are seeing.
> Is this the case for all samples, or only a few?

Only a few when the tests are run for a short period of time. Even with a 
single threaded test the times are off a little for some calls.

When the test is left for more than a few hours the times seem to become more 
in-accurate. 

> And does it really matter, so long as the server can handle the load
> it's supposed to?

Part of our tests are to test how many "users" the app can handle before a 
degradation of quality beyond a certain point (e.g. 2 seconds), so the 
inaccuracies I'm seeing are not helping to determine that number :( because 
the average time per call for 10 threads grows over time

> >  Regards,
> >
> > Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Accuracy of Elapsed Time

[Noel O'Brien]

On Friday 27 March 2009 01:39:27 sebb wrote:
> I've not used it much, but the only problem I can think of with 1.6 is
> the change to the socket code which causes the "unconnected socket"
> error when using HTTPS.
>
> [This was a bug in JMeter which has been fixed; the nightlies work with
> 1.6]

The only issue I saw was the bug in Java (BackingStoreException), which did 
not impact testing; it only printed out lots of warnings and resulted in the 
"Recent File List" not being saved.

Solution was to install a version of Java >= 1.6.0_10

Regarding the original issue, I guess I've nothing to lose by trying with 1.5. 
I'll give it a go and report back

Regards,
Noel

> Are there any other problems with 1.6 you know of?
>
> On 27/03/2009, Oliver Erlewein [DATACOM]  
wrote:
> > Hi Noel,
> >
> >  I'd stay away from using Java 1.6.0 for now. Hasn't yet proven to be
> >  that dependable. Sebb might have some more info on that.
> >
> >  Cheers Oliver
> >
> >
> >  -Original Message-
> >  From: Noel O'Brien [mailto:nobr...@newbay.com]
> >  Sent: Friday, March 27, 2009 3:17 AM
> >  To: jmeter-user@jakarta.apache.org
> >  Cc: sebb
> >  Subject: Re: Accuracy of Elapsed Time
> >
> >  On Thursday 26 March 2009 13:00:02 sebb wrote:
> >  > On 26/03/2009, Noel O'Brien  wrote:
> >  > > Hi,
> >  > >
> >  > >  I'm running some load testing, nothing too heavy. 10 JMeter
> >
> >  threads,
> >
> >  > > ramp up is 30. JMeter is run from the command line and the test plan
> >
> >  has
> >
> >  > > no listeners. using the -l flag, I capture the results to a CSV
> >
> >  file.
> >
> >  > Good.
> >  >
> >  > JMeter version?
> >
> >  JMeter 2.3.2:r665936
> >  Java: Sun 1.6.0_06-b02
> >  OS: RHEL 5.2 running in RHEL hypervisor
> >
> >  > >  I then load the results file into JMeter and look at hem from a
> >
> >  Summary
> >
> >  > >  Listener, View Tree Listener etc. I'm a little confused by the
> >
> >  elapsed
> >
> >  > > time results.
> >  > >
> >  > >  The app under test has the ability to log the time that's taken to
> >  > > receive the request and complete sending the response, which for an
> >
> >  auth
> >
> >  > > API call it lists as taking 114ms for example (ronded to the closest
> >
> >  ms)
> >
> >  > >  I've set up tcpdump on the machine running the app (refer to as
> >
> >  server
> >
> >  > >  henceforth) and the machine running  JMeter (refer to as client
> >  > > henceforth).
> >  > >
> >  > >  tcpdump on the server reports that the time between the request and
> >  > > response packets to be 113.83ms, while tcpdump on the client reports
> >
> >  that
> >
> >  > > the time between the request and response packets 113.96ms
> >  > >
> >  > >  However for that particular HTTP request, JMeter reports that the
> >
> >  load
> >
> >  > > time (elasped time) is 174ms and that the latency is 174 ms also.
> >
> >  How is
> >
> >  > > this discrepancy explained?
> >  >
> >  > There is overhead in the OS, Java and JMeter on both sending and
> >
> >  receiving.
> >
> >  Some of the other API calls I make only show a difference of a few ms in
> >  JMeter
> >  compared to the app/tcpdump
> >
> >  > Note also that the timer resolution will affect the reported elapsed
> >
> >  time.
> >
> >  > However if you are running 2.3.2+ on Java 1.5+ it will use a higher
> >  > resolution timer.
> >
> >  Does it matter that the auth API call is the first call for each thread?
> >  (Is
> >  the thread doing extra work at it's start-up that would impact the
> >  time?)
> >
> >  > >  As I understand it, the latency is the time taken to receive the
> >
> >  first
> >
> >  > > byte of the response and since the load time and latency are the
> >
> >  same
> >
> >  > > then it indicated that the response payload was received within 1
> >
> >  ms. Is
> >
> >  > > this correct? Or is it even possible to achieve this? The payload
> >
> >  size
> >
> >  > > for this response is 106 bytes.
> >  >
> >  > Latency is time to first 

Re: Accuracy of Elapsed Time

[Noel O'Brien]

Just wondering: would adding a think time to a load test help JMeter in terms 
of elapsed time accuracy?

I ask because, I configured the app under testing to temporarily eliminate 
activity logging which was expensive in terms of time on the server side. Now 
the average test times in JMeter are much lower when viewed in the summary 
report listener from 2sec to about 300 ms per call. 

I looked at the app logs which output the time it took to process the 
requests; the longest time was 8sec, but JMeter is reporting the max times for 
each call are about 15sec! 

I'm thinking that server is super fast (but that's probably just the optimist 
in me ;)) and JMeter is struggling to keep up in processing. Again the test 
plan has no listeners, outputs only critical results data in CSV, JMeter runs 
in non-gui mode, with 10 threads, 30 sec ramp-up. 

When I arrived into work this morning, JMeter was at 100% CPU, but the app 
server was cruising at about 10%. I remember reading somewhere on the JMeter 
site (possibly the wiki) that a faster server will put more load on JMeter to 
keep up. What can be done in this situation to have a stable load test?

Will a think time help if I increase the number of threads? Or will a higher 
number of threads make things worse?

Regards,
Noel

On Friday 27 March 2009 09:13:06 Noel O'Brien wrote:
> On Friday 27 March 2009 01:39:27 sebb wrote:
> > I've not used it much, but the only problem I can think of with 1.6 is
> > the change to the socket code which causes the "unconnected socket"
> > error when using HTTPS.
> >
> > [This was a bug in JMeter which has been fixed; the nightlies work with
> > 1.6]
>
> The only issue I saw was the bug in Java (BackingStoreException), which did
> not impact testing; it only printed out lots of warnings and resulted in
> the "Recent File List" not being saved.
>
> Solution was to install a version of Java >= 1.6.0_10
>
> Regarding the original issue, I guess I've nothing to lose by trying with
> 1.5. I'll give it a go and report back
>
> Regards,
> Noel
>
> > Are there any other problems with 1.6 you know of?
> >
> > On 27/03/2009, Oliver Erlewein [DATACOM] 
>
> wrote:
> > > Hi Noel,
> > >
> > >  I'd stay away from using Java 1.6.0 for now. Hasn't yet proven to be
> > >  that dependable. Sebb might have some more info on that.
> > >
> > >  Cheers Oliver
> > >
> > >
> > >  -Original Message-
> > >  From: Noel O'Brien [mailto:nobr...@newbay.com]
> > >  Sent: Friday, March 27, 2009 3:17 AM
> > >  To: jmeter-user@jakarta.apache.org
> > >  Cc: sebb
> > >  Subject: Re: Accuracy of Elapsed Time
> > >
> > >  On Thursday 26 March 2009 13:00:02 sebb wrote:
> > >  > On 26/03/2009, Noel O'Brien  wrote:
> > >  > > Hi,
> > >  > >
> > >  > >  I'm running some load testing, nothing too heavy. 10 JMeter
> > >
> > >  threads,
> > >
> > >  > > ramp up is 30. JMeter is run from the command line and the test
> > >  > > plan
> > >
> > >  has
> > >
> > >  > > no listeners. using the -l flag, I capture the results to a CSV
> > >
> > >  file.
> > >
> > >  > Good.
> > >  >
> > >  > JMeter version?
> > >
> > >  JMeter 2.3.2:r665936
> > >  Java: Sun 1.6.0_06-b02
> > >  OS: RHEL 5.2 running in RHEL hypervisor
> > >
> > >  > >  I then load the results file into JMeter and look at hem from a
> > >
> > >  Summary
> > >
> > >  > >  Listener, View Tree Listener etc. I'm a little confused by the
> > >
> > >  elapsed
> > >
> > >  > > time results.
> > >  > >
> > >  > >  The app under test has the ability to log the time that's taken
> > >  > > to receive the request and complete sending the response, which
> > >  > > for an
> > >
> > >  auth
> > >
> > >  > > API call it lists as taking 114ms for example (ronded to the
> > >  > > closest
> > >
> > >  ms)
> > >
> > >  > >  I've set up tcpdump on the machine running the app (refer to as
> > >
> > >  server
> > >
> > >  > >  henceforth) and the machine running  JMeter (refer to as client
> > >  > > henceforth).
> > >  > >
> > >  > >  tcpdump on the server reports that the time between the req

Re: Jmeter to load test a java application which takes XML input

[Noel O'Brien]

How does one connect to the Java app? Through a socket connection or does the 
Java app support a http connection? This will determine the appropriate 
sampler for you to use

Regards,
Noel

On Monday 30 March 2009 14:35:37 nawaz_farhan wrote:
> Application gets input through a XML file.
> Application displays output to a XML file.
>
> sebb-2-2 wrote:
> > On 30/03/2009, nawaz_farhan  wrote:
> >>  I'm involved in testing a java application.
> >>  This application takes input as a XML and proccess it sends it to a
> >> Derby
> >>  DB, and retrives results from the DB, process the results and outputs a
> >> XML.
> >>  I need to do concurrent users load testing on that application.
> >>  Any inputs on how to do it.
> >
> > If you want to test the application, then JMeter needs to send work to
> > the application.
> >
> > How does the application get its input and return its ouput?
> >
> >>  Thanks in advance,
> >>  Nawaz
> >>
> >> --
> >>  View this message in context:
> >> http://www.nabble.com/Jmeter-to-load-test-a-java-application-which-takes
> >>-XML-input-tp22781169p22781169.html Sent from the JMeter - User mailing
> >> list archive at Nabble.com.
> >>
> >>
> >>  -
> >>  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >>  For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
> >
> > -
> > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Starting thread group once other is finished.

[Noel O'Brien]

In the test plan element you can check "Run threads groups consecutively" 
option

Regards,
Noel

On Monday 30 March 2009 15:28:05 Salman Ghaffar wrote:
> I have around ten thread groups, I want to run them serially, like only
> start 1st one than let it finish than start second one and so on.
>
> Tried constant timer,but that's constant, and the running time of my
> threads could be variable so does JMeter provide anything for that out
> of the box ?
>
> Salman

Re: Response assertion after if

[Noel O'Brien]

It's an issue of scope; see http://jakarta.apache.org/jmeter/usermanual/build-
test-plan.html section 3.6

Response assertion will only be checked if it's a *direct* child of the 
sampler, is a sibling test element or in a parent controller

Regards,
Noel

On Monday 30 March 2009 16:53:33 Salman Ghaffar wrote:
> I have following layout
>
>
> --HTTP Request
>
> --IF Controller
>
>  --Reponse Assertion.
>
>
> This response assertion is not being checked in either case of the if
> condition pass or fail, is that a bug in 2.3.1 ?
>
> Thanks
>
> salman

Re: Response assertion after if

[Noel O'Brien]

What are you checking with the If Controller? 


Regards,
Noel

On Monday 30 March 2009 17:13:43 Salman Ghaffar wrote:
> But than, if I want to check responses based on condition(if controller)
> being true or false it won't be possible ?
>
> Because my assertion would always be child of If controller ?
>
>
> -Original Message-
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: Monday, March 30, 2009 4:58 PM
> To: JMeter Users List
> Subject: Re: Response assertion after if
>
> It's an issue of scope; see
> http://jakarta.apache.org/jmeter/usermanual/build-
> test-plan.html section 3.6
>
> Response assertion will only be checked if it's a *direct* child of the
> sampler, is a sibling test element or in a parent controller
>
> Regards,
> Noel
>
> On Monday 30 March 2009 16:53:33 Salman Ghaffar wrote:
> > I have following layout
> >
> >
> > --HTTP Request
> >
> > --IF Controller
> >
> >  --Reponse Assertion.
> >
> >
> > This response assertion is not being checked in either case of the if
> > condition pass or fail, is that a bug in 2.3.1 ?
> >
> > Thanks
> >
> > salman
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Response assertion after if

[Noel O'Brien]

A BSFAssertion / Beanshell Assertion would work i this case, seeing as both 
cases are 

Would something like this work:

1. define a retry variable
2. while controller with retry condition
3.  HTTP Sampler
4.  Beanshell Assertion checks response, if retry link is present, 
re-iterate 
in while loop, otherwise make other assertions and exit

Regards,
Noel

On Monday 30 March 2009 17:46:02 Salman Ghaffar wrote:
> Basically in my response I get the splash screen which refreshes itself,
> so I can either get the content or a link to retry
>
>
>
> Through IF controller I check whether I've got the actual content or the
> retry link. ( a regex tries to extract retry link and I check regex's
> value in if controller)
>
>
>
> In case I get retry link, I try that link again through another request,
> In case I get original contents I want to check those contents for some
> strings in them.
>
>
>
> thanks
>
>
>
> Salman
>
>
>
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: Monday, March 30, 2009 5:28 PM
> To: jmeter-user@jakarta.apache.org
> Cc: Salman Ghaffar
> Subject: Re: Response assertion after if
>
>
>
> What are you checking with the If Controller?
>
> Regards,
>
> Noel
>
> On Monday 30 March 2009 17:13:43 Salman Ghaffar wrote:
> > But than, if I want to check responses based on condition(if
>
> controller)
>
> > being true or false it won't be possible ?
> >
> >
> >
> > Because my assertion would always be child of If controller ?
> >
> >
> >
> >
> >
> > -Original Message-
> >
> > From: Noel O'Brien [mailto:nobr...@newbay.com]
> >
> > Sent: Monday, March 30, 2009 4:58 PM
> >
> > To: JMeter Users List
> >
> > Subject: Re: Response assertion after if
> >
> >
> >
> > It's an issue of scope; see
> >
> > http://jakarta.apache.org/jmeter/usermanual/build-
> >
> > test-plan.html section 3.6
> >
> >
> >
> > Response assertion will only be checked if it's a *direct* child of
>
> the
>
> > sampler, is a sibling test element or in a parent controller
> >
> >
> >
> > Regards,
> >
> > Noel
> >
> > On Monday 30 March 2009 16:53:33 Salman Ghaffar wrote:
> > > I have following layout
> > >
> > >
> > >
> > >
> > >
> > > --HTTP Request
> > >
> > >
> > >
> > > --IF Controller
> > >
> > >
> > >
> > > --Reponse Assertion.
> > >
> > >
> > >
> > >
> > >
> > > This response assertion is not being checked in either case of the
>
> if
>
> > > condition pass or fail, is that a bug in 2.3.1 ?
> > >
> > >
> > >
> > > Thanks
> > >
> > >
> > >
> > > salman
> >
> > -
> >
> > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >
> > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: how to record the HTTPS Requests Through Proxy server

[Noel O'Brien]

Please do not contact people directly, use the mailing list instead. Your 
questions will reach a greater audience and you won't be annoying anyone.

http://jakarta.apache.org/jmeter/usermanual/component_reference.html#HTTP_Proxy_Server

You cannot record HTTPS traffic with JMeter's proxy server

Regards,
Noel

On Tuesday 07 April 2009 14:26:20 M.Vijaya Bhaskar wrote:
> Hi All,
>
> Can you please tell me how to record the HTTPS Request with Proxy server
> settings.
>
> i am able to record the HTTPS Requests with BadBoy.
>
> But i want to Rcord the HTTPS Request with Proxy server.
>
> Thank
> M.Vijaya Bhaskar

Re: how stuff works Jmeter ??

[Noel O'Brien]

http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Thread_Group

Regards,
Noel

On Tuesday 07 April 2009 15:32:37 ricardo13 wrote:
> Hi,
>
> More once thank you !!
>
> Ok but Jmeter create Threads, Processes to request webserver ??
> When I configure Jmeter I typed 10 users and ramp-up . How does Jmeter
> do it internally ??
>
> I would want know this details.
>
> Thank You
>
> Ricardo

Re: Creating JMeter test cases programatically and running through program

[Noel O'Brien]

On Wednesday 08 April 2009 14:57:38 Rajesh Balamohan wrote:
> Hello Folks,
>
> I have some experience in using this wonderful JMeter tool. It is really
> useful.
>
> I was wondering if we could the following with latest release of JMeter.
>
> 1. Create test plan programatically (using java client code, is it possible
> to create test-plan and pass it on to JMeter?)
> 2. Is it possible to kick off JMeter from some other java program?. (I know
> only java-ant task is available. However, I would like to know if it is
> possible to kick off from some other java code). This will help in doing
> some performance automation.

You could use Runtime.exec() to run JMeter from Java code

Regards,
Noel

>
> Any pointers on this would be of great help.

Re: MD5 conversion of a string.

[Noel O'Brien]

Here's some code I use in a Beanshell PreProcessor to do MD5 encryption:

import java.math.BigInteger;
import java.security.MessageDigest;
import java.util.TreeMap;
import java.net.URLEncoder;

.
...
...

authSigString = strList.toString();

// Log the auth signature string
log.debug("authSigString: " + authSigString);

digest = MessageDigest.getInstance("MD5");
bytes = authSigString.getBytes();
digest.reset(); 
theHash = new StringBuilder(new BigInteger(1, 
digest.digest(bytes)).toString(16));
// Pad to correct length
while (theHash.length() < 32)
theHash.insert(0, "0");

log.debug("authSigHash: " + theHash.toString());
return theHash.toString();

Regards,
Noel

On Wednesday 08 April 2009 16:49:16 Zakir Sayed wrote:
> Hello everybody,
>
> I need to get MD5 format of a concatenated string, which is then
> converted to Hex format. For MD5 conversion, I downloaded the js script
> file from http://pajhome.org.uk/crypt/md5/ (also attached <> ).
> Also that same script file (md5.js) has 'binl2hex(binarray)' function
> which convert to hex value.. My real problem is (perhaps simple!) that I
> don't know the way to access those functions inside that script in my
> SOAP XML/RPC data request. I mean, I don't how to call that js file in
> my SOAP request to process some variable value to get MD5 hash of it.
> Ideally I would like to add or call that hex_md5 function from "Function
> Helper dialog" but any other possible solution would also do. So in
> brief, I need to have convert a string to MD5 through JavaScript or
> BeanShell or any other way possible.
>
> I really need that to complete my project and it's not possible to work
> around with them, so any kind of even suggestion is highly appreciated.
>
> 
> Thanks,
> Zack Sayed.

Re: MD5 conversion of a string.

[Noel O'Brien]

I only posted a snippet, because some of the other code has company specific 
details in it, but what I posted shows how to create a hash. 

AFAIK, you can put Java (but not javascript) code straight into a beanshell 
script and it will work, but I may have wrong. So if you've got any Java 
experience, that will help 

In my case, `strList` is a java StringBuffer used to build up a lot of 
substrings into the string I want to get the hash of (`authSigString`). Also 
the code I posted was part of a function, as you can see from the return 
statement at the end.

You can run this code procedurally (without methods), you just need to build 
your string before the block of code I posted and then manipulate your SOAP 
request to include the hash.

I don't have any experience with SOAP requests unfortunately

Regards,
Noel

On Thursday 09 April 2009 00:14:29 Zakir Sayed wrote:
> Thanks Noel for the code.
>
> Is there anything similar in JavaScript because I don't have experience
> on BeanShell?
> I installed BeanShell jar file, I copied this code to tried BeanShell
> Preprocessor script and provided the script file File Name (md5.bsh) and
> parameter (strList) as argument but couldn't comprehend how I can call
> this script by sending a string variable as parameter in my SOAP request
> data. I also ran few __BeanShell functions successfully in my SOAP
> request data just to make sure BeanShell in installed properly.
>
> I would highly oblige if you could spoon feed this time in showing how
> to call that code to have MD5 conversion.
>
> Thanks,
> Zakir Sayed
>
>
> -Original Message-
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: Wednesday, April 08, 2009 8:59 AM
> To: jmeter-user@jakarta.apache.org
> Cc: Zakir Sayed; Ronan Klyne
> Subject: Re: MD5 conversion of a string.
>
> Here's some code I use in a Beanshell PreProcessor to do MD5 encryption:
>
> import java.math.BigInteger;
> import java.security.MessageDigest;
> import java.util.TreeMap;
> import java.net.URLEncoder;
>
> .
> ...
> ...
>
> authSigString = strList.toString();
>
>   // Log the auth signature string
>   log.debug("authSigString: " + authSigString);
>
>   digest = MessageDigest.getInstance("MD5");
>   bytes = authSigString.getBytes();
>   digest.reset();
>   theHash = new StringBuilder(new BigInteger(1,
> digest.digest(bytes)).toString(16));
>   // Pad to correct length
>   while (theHash.length() < 32)
>   theHash.insert(0, "0");
>
>   log.debug("authSigHash: " + theHash.toString());
>   return theHash.toString();
>
> Regards,
> Noel
>
> On Wednesday 08 April 2009 16:49:16 Zakir Sayed wrote:
> > Hello everybody,
> >
> > I need to get MD5 format of a concatenated string, which is then
> > converted to Hex format. For MD5 conversion, I downloaded the js
>
> script
>
> > file from http://pajhome.org.uk/crypt/md5/ (also attached <>
>
> ).
>
> > Also that same script file (md5.js) has 'binl2hex(binarray)' function
> > which convert to hex value.. My real problem is (perhaps simple!) that
>
> I
>
> > don't know the way to access those functions inside that script in my
> > SOAP XML/RPC data request. I mean, I don't how to call that js file in
> > my SOAP request to process some variable value to get MD5 hash of it.
> > Ideally I would like to add or call that hex_md5 function from
>
> "Function
>
> > Helper dialog" but any other possible solution would also do. So in
> > brief, I need to have convert a string to MD5 through JavaScript or
> > BeanShell or any other way possible.
> >
> > I really need that to complete my project and it's not possible to
>
> work
>
> > around with them, so any kind of even suggestion is highly
>
> appreciated.
>
> > 
> > Thanks,
> > Zack Sayed.
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Bug in Loop Controller??

[Noel O'Brien]

Hi All,

I've found some unusual behavior when I use a loop controller inside a for 
each controller. I think it may be a bug, but hopefully someone can explain 
why it's happening:

I can demonstrate it by using the following test plan:



thread group
-User defined variables
-ForEach Controller
-Debug Sampler 1
-Loop Controller
-Debug Sampler 2
View Results Tree


User defined variables element contains the following values:
val_1   1
val_2   2
val_3   0
val_4   0
val_5   3
val_6   0
val_7   7
val_8   8

The basic objective of the test plan is: "for each `val`, loop that number of 
times"

ForEach Controller is configured:
Input Variable Prefix: val
Output Variable Name: this.val
Add '_' before number: yes

Loop Controller is configured:
Loop Count: ${this.val}
Forever: No

I expected this to produce the following sequence in the View Results tree:
Debug Sampler 1
Debug Sampler 2
Debug Sampler 1
Debug Sampler 2
Debug Sampler 2
Debug Sampler 1
Debug Sampler 1
Debug Sampler 1
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 1
Debug Sampler 1
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 1
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2
Debug Sampler 2

but instead is produced the following results:
Debug Sampler 1
Debug Sampler 2
Debug Sampler 1
Debug Sampler 2
Debug Sampler 2
Debug Sampler 1
Debug Sampler 2
Debug Sampler 1
Debug Sampler 1
Debug Sampler 1
Debug Sampler 1
Debug Sampler 1

It's really weird, because in the first case where the loop controller will be 
0 (third iteration, val_3=0), Debug Sampler 2 is executed once, but it lists 
`this.val=0`, which is the correct value, but since it's 0, the loop 
controller should not have been entered at all, right? It seems too that once 
a loop controller has been executed with a value of 0, any subsequent 
iterations of the foreach controller which should enter the loop controller 
(i.e. iterations 5, 7, and 8 of the foreach) do not enter it.

Thoughts?

Regards,
Noel

Re: Regular Expressions

[Noel O'Brien]

Do you need to have line breaks in your data?
Is the data in question sent as a parameter or in the request body?

Regards,
Noel

On Wednesday 15 April 2009 17:05:47 Jannetta Steyn wrote:
> Hi Folks
>
> I posted this message earlier today but it didn't seem to go through.
> Apologies if it now ends up having gone through twice.
>
> I wonder if someone can perhaps help. I am using JMeter to test a web site
> which consist of forms submitted by POST. I read data from a CSV file and
> then submit the information using HTTP Request. The problem I have is with
> the CSV Data Set Config not being able to read data with line breaks in the
> fields. So what I have done is to replace the line breaks with an odd
> character (or with ). However before I submit I would like to replace
> the odd character (or the ) with a line break.
>
> I have tried a few things but just can't get it to work. Does anybody have
> any advice regarding this?
>
> Thanks in advance.
> Jannetta

Re: Regular Expressions

[Noel O'Brien]

I have a beanshell pre processor script that I use to replace values in 
parameters of a HTTP Request, maybe you can modify it to do a find and replace 
on the  placeholders:

import org.apache.jmeter.protocol.http.util.HTTPArgument;




// Get HTTP Params
params = sampler.getArguments();

passArg = params.getArgument(0);
userArg = params.getArgument(1);

if (!passArg.getName().equals("password")) {
log.error("Password not first HTTP Param: " + userArg.getName());
return;
}
if (!userArg.getName().trim().equals("username")) {
log.error("Username not second HTTP Param: " + passArg.getName());
return;
}

pass = passArg.getValue();
user = userArg.getValue();

if (user == null || pass == null)
return;

...
...
...

userArg.setValue(user);
passArg.setValue(pass);


Regards,
Noel

On Wednesday 15 April 2009 18:50:35 Jannetta Steyn wrote:
> Yes the data of one of the fields I want to submit is a textarea which
> sometimes contains line breaks.
>
> I have two ways of retrieving the data. The first is as I mentioned before
> from the CSV file, so it is in a variable. Because the CSV data config
> can't read line breaks, I replaced the line breaks with .I can replace
> it with anything really as long as when I submit it to the text field I can
> replace it with line breaks again.
>
> Alternatively I can retrieve the data from a database, but then I don't
> know how to get it into a variable to submit in the form.
>
> I tried to use a Regular Expression Extractor to retrieve the data from the
> Response Data. The data is then broken up into separate variables by each
> line break. For instance, if the reference name I specify is DESCRIPTION,
> the Regular expression is (.+), template is $1$, match no is -1, then
> DESCRIPTION_matchNr will contain the number of lines read and the lines
> will be in DESCRIPTION_1 to DESCRIPTION_n. What I then can't figure out is
> how to submit the whole lot to the form again - in an automated manner of
> course.
>
> Regards
> Jannetta
>
>
>
>
> 2009/4/15 Noel O'Brien 
>
>  Do you need to have line breaks in your data?
>
> > Is the data in question sent as a parameter or in the request body?
> >
> > Regards,
> >
> > Noel
> >
> > On Wednesday 15 April 2009 17:05:47 Jannetta Steyn wrote:
> > > Hi Folks
> > >
> > >
> > >
> > > I posted this message earlier today but it didn't seem to go through.
> > >
> > > Apologies if it now ends up having gone through twice.
> > >
> > >
> > >
> > > I wonder if someone can perhaps help. I am using JMeter to test a web
> >
> > site
> >
> > > which consist of forms submitted by POST. I read data from a CSV file
> > > and
> > >
> > > then submit the information using HTTP Request. The problem I have is
> >
> > with
> >
> > > the CSV Data Set Config not being able to read data with line breaks in
> >
> > the
> >
> > > fields. So what I have done is to replace the line breaks with an odd
> > >
> > > character (or with ). However before I submit I would like to
> >
> > replace
> >
> > > the odd character (or the ) with a line break.
> > >
> > >
> > >
> > > I have tried a few things but just can't get it to work. Does anybody
> >
> > have
> >
> > > any advice regarding this?
> > >
> > >
> > >
> > > Thanks in advance.
> > >
> > > Jannetta

Re: Regular Expressions

[Noel O'Brien]

Hi Jannetta,

Happy to help :)

I'm not sure whether it's a beanshell thing or a JMeter thing, but 
System.out.println doesn't seem to print to standard out for me. I use the 
print() function instead for stdout, or log.info() for writing to jmeter.log

Regards,
Noel

On Friday 17 April 2009 10:36:55 Jannetta Steyn wrote:
> Hi Noel
>
> The beanshell worked. Thank you very much for the suggestion. It was much
> simpler than your example though. This is all I had to do:
>
> String desc = vars.get("DESCRIPTION");
>
> desc = desc.replaceAll("","\r\n\r\n");
> vars.put("DESCRIPTION",desc);
>
> I was just wondering. Where does the output of a System.out.println go?
>
> Kind Regards
> Jannetta
>
> 2009/4/16 Noel O'Brien 
>
> >  I have a beanshell pre processor script that I use to replace values in
> > parameters of a HTTP Request, maybe you can modify it to do a find and
> > replace on the  placeholders:
> >
> > import org.apache.jmeter.protocol.http.util.HTTPArgument;
> >
> >  
> >
> > 
> >
> >  // Get HTTP Params
> >
> > params = sampler.getArguments();
> >
> >  passArg = params.getArgument(0);
> >
> > userArg = params.getArgument(1);
> >
> >  if (!passArg.getName().equals("password")) {
> >
> > log.error("Password not first HTTP Param: " + userArg.getName());
> >
> > return;
> >
> > }
> >
> > if (!userArg.getName().trim().equals("username")) {
> >
> > log.error("Username not second HTTP Param: " + passArg.getName());
> >
> > return;
> >
> > }
> >
> >  pass = passArg.getValue();
> >
> > user = userArg.getValue();
> >
> >  if (user == null || pass == null)
> >
> > return;
> >
> >  ...
> >
> > ...
> >
> > ...
> >
> > userArg.setValue(user);
> >
> > passArg.setValue(pass);
> >
> >  Regards,
> >
> > Noel
> >
> > On Wednesday 15 April 2009 18:50:35 Jannetta Steyn wrote:
> > > Yes the data of one of the fields I want to submit is a textarea which
> > >
> > > sometimes contains line breaks.
> > >
> > >
> > >
> > > I have two ways of retrieving the data. The first is as I mentioned
> >
> > before
> >
> > > from the CSV file, so it is in a variable. Because the CSV data config
> > >
> > > can't read line breaks, I replaced the line breaks with .I can
> >
> > replace
> >
> > > it with anything really as long as when I submit it to the text field I
> >
> > can
> >
> > > replace it with line breaks again.
> > >
> > >
> > >
> > > Alternatively I can retrieve the data from a database, but then I don't
> > >
> > > know how to get it into a variable to submit in the form.
> > >
> > >
> > >
> > > I tried to use a Regular Expression Extractor to retrieve the data from
> >
> > the
> >
> > > Response Data. The data is then broken up into separate variables by
> > > each
> > >
> > > line break. For instance, if the reference name I specify is
> > > DESCRIPTION,
> > >
> > > the Regular expression is (.+), template is $1$, match no is -1, then
> > >
> > > DESCRIPTION_matchNr will contain the number of lines read and the lines
> > >
> > > will be in DESCRIPTION_1 to DESCRIPTION_n. What I then can't figure out
> >
> > is
> >
> > > how to submit the whole lot to the form again - in an automated manner
> > > of
> > >
> > > course.
> > >
> > >
> > >
> > > Regards
> > >
> > > Jannetta
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > 2009/4/15 Noel O'Brien 
> > >
> > >
> > >
> > > Do you need to have line breaks in your data?
> > >
> > > > Is the data in question sent as a parameter or in the request body?
> > > >
> > > >
> > > >
> > > > Regards,
> > > >
> > > >
> > > >
> > > > Noel
> > > >
> > > > On Wednesday 15 April 2009 17:05:47 Jannetta Steyn wrote:
> > > > > Hi Folks
> > > > >
> > > > >
> > >

Re: Regular Expressions

[Noel O'Brien]

Try replacing 

&

with

&

Regards,
Noel

On Friday 17 April 2009 14:24:43 Jannetta Steyn wrote:
> Does anyone perhaps have any advice with regards to using the ampersand
> character in a textarea? At the moment the ampersand is being interpreted
> by the server as another parameter. I'll use the beanshell to replace the
> ampersand, but what do I replace it with?
>
> With regards to System.out.println, I remembered that I changed java.exe to
> javaw.exe in my jmeter.bat file because I like to be able to close the DOS
> box (or whatever it is called nowadays). When using java.exe though, it
> does print to console.
>
> Regards
> Jannetta
>
> 2009/4/17 sebb 
>
> > On 17/04/2009, Noel O'Brien  wrote:
> > > Hi Jannetta,
> > >
> > >  Happy to help :)
> > >
> > >  I'm not sure whether it's a beanshell thing or a JMeter thing, but
> > >  System.out.println doesn't seem to print to standard out for me. I use
> >
> > the
> >
> > Works for me - and print() just calls it anyway.
> >
> > Both appear in the console window.
> >
> > >  print() function instead for stdout, or log.info() for writing to
> >
> > jmeter.log
> >
> > >  Regards,
> > >
> > > Noel
> > >
> > >  On Friday 17 April 2009 10:36:55 Jannetta Steyn wrote:
> > >  > Hi Noel
> > >  >
> > >  > The beanshell worked. Thank you very much for the suggestion. It was
> >
> > much
> >
> > >  > simpler than your example though. This is all I had to do:
> > >  >
> > >  > String desc = vars.get("DESCRIPTION");
> > >  >
> > >  > desc = desc.replaceAll("","\r\n\r\n");
> > >  > vars.put("DESCRIPTION",desc);
> >
> > Or you could use a script function to return the changed string, for
> > example in BeanShell:
> >
> > ${__BeanShell("${DESCRIPTION}".replaceAll(""\,"\r\n\r\n"))}
> >
> > N.B. the comma in the replaceAll() parameter list needs to be escaped
> > with backslash so it is treated as part of the script.
> >
> > and if you want to change the variable as well:
> >
> >
> > ${__BeanShell("${DESCRIPTION}".replaceAll(""\,"\r\n\r\n"),DESCRIPTIO
> >N)}
> >
> > Similar code should work for Javascript and Jexl.
> >
> > Have a look at BeanShellFunction.bshrc, as that contains an example
> > function definition for replacing "&" with "&". It is called as
> > follows:
> >
> > ${__BeanShell(fixamps(vars.get("DESCRIPTION")))}
> >
> > You could create your own function:
> >
> > String fixBRinVar(varname) {
> >  return vars.get(varname).replaceAll("","\r\n\r\n");
> > }
> >
> > and call as
> >
> > ${__BeanShell(fixBRinVar("DESCRIPTION"))}
> >
> > >  > I was just wondering. Where does the output of a System.out.println
> >
> > go?
> >
> > >  > Kind Regards
> > >  > Jannetta
> > >  >
> > >  > 2009/4/16 Noel O'Brien 
> > >  >
> > >  > >  I have a beanshell pre processor script that I use to replace
> >
> > values in
> >
> > >  > > parameters of a HTTP Request, maybe you can modify it to do a find
> >
> > and
> >
> > >  > > replace on the  placeholders:
> > >  > >
> > >  > > import org.apache.jmeter.protocol.http.util.HTTPArgument;
> > >  > >
> > >  > >  
> > >  > >
> > >  > > 
> > >  > >
> > >  > >  // Get HTTP Params
> > >  > >
> > >  > > params = sampler.getArguments();
> > >  > >
> > >  > >  passArg = params.getArgument(0);
> > >  > >
> > >  > > userArg = params.getArgument(1);
> > >  > >
> > >  > >  if (!passArg.getName().equals("password")) {
> > >  > >
> > >  > > log.error("Password not first HTTP Param: " + userArg.getName());
> > >  > >
> > >  > > return;
> > >  > >
> > >  > > }
> > >  > >
> > >  > > if (!userArg.getName().trim().equals("username")) {
> > >  > >
> > >  > > log.error("Username not second HTTP Param: " + passArg.getName());
&g

Re: Weird pause

[Noel O'Brien]

Are you running the server and/or JMeter in a Virtual Machine or a physical 
machine?

I had a similar issue , which turned out to be caused by IPTables on the VM 
host 

Regards,
Noel

On Tuesday 21 April 2009 05:22:19 Ashley Martens wrote:
> Hello
>
> I created my first JMeter test and it seems to work fine, however,
> after about 250 connections the test just stops for about 1 minute
> then continues and begins to pause every so often for a minute or two.
> I can't tell if this is JMeter or my web application. When I tail my
> web application log it seems to be responding fine and with fast
> response times. However, I can't tell what JMeter is doing. I ran
> JMeter without a GUI and with "-L jmeter=DEBUG" but everything just
> seems to stop in these periods. Does any one have any ideas as to how
> I can tell what's going on?
>
> Ashley
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Replace JDBC Pool

[Noel O'Brien]

Hi All,

I'm seeing some issues with slow INSERT/COMMIT times (about 4sec!) to MySQL 
from our product and would like to recreate the same setup with JMeter to rule 
out certain components (hibernate, DBCP, etc.)

Our product uses DBCP for connection pooling but JMeter uses Avalon if I'm not 
mistaken. Is it an easy task to replace Avalon with DBCP in JMeter?

Regards,
Noel

Re: Replace JDBC Pool

[Noel O'Brien]

On Tuesday 21 April 2009 15:47:16 sebb wrote:
> On 21/04/2009, Noel O'Brien  wrote:
> > Hi All,
> >
> >  I'm seeing some issues with slow INSERT/COMMIT times (about 4sec!) to
> > MySQL from our product and would like to recreate the same setup with
> > JMeter to rule out certain components (hibernate, DBCP, etc.)
> >
> >  Our product uses DBCP for connection pooling but JMeter uses Avalon if
> > I'm not mistaken. Is it an easy task to replace Avalon with DBCP in
> > JMeter?
>
> No, but if you set the JDBC Config Max Number of Connections to 0, it
> will use a connection per thread, i.e. no pooling.

Interesting, I'll try  this to see what the effect is.

> I'd like to remove connection pooling altogether from JMeter.
> JMeter threads are supposed to be independent users, and users don't
> share DB pools.

That's a shame, as it pooling was useful in my case to simulate how our 
product works and to help point towards the DB as a bottleneck. I can see your 
point thought, from an end-user point of view there's no sharing of DB pools. 
Maybe it could be left an an option?

> >  Regards,
> >
> > Noel
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Response Time of Http Request

[Noel O'Brien]

Elapsed Time  == Load Time in  "View Results Tree" listener. 

Regards,
Noel

On Wednesday 22 April 2009 08:30:09 linuxos wrote:
> Thanks. But I can't find the Elapsed time in the "Sampler result".
> Here is my result from the "View Results Tree":
>
> Thread Name: Thread Group 1-1
> Sample Start: 2009-04-22 14:54:01 CST
> Load time: 1489
> Latency: 1360
> Size in bytes: 81648
> Sample Count: 1
> Error Count: 0
> Response code: 200
> Response message: OK
>
> Is there any wrong step in using the JMeter?
>
> sebb-2-2 wrote:
> > On 21/04/2009, linuxos  wrote:
> >>  I've just started using JMeter 2.3.2. I created a "Thread Group", added
> >> "HTTP
> >>  Request Sampler" and "View Results Tree" for my testing.
> >>
> >>  For calculation of "Response Time" of a request, is it correct to say
> >> it's
> >>  the "Load time" minus "Latency"?
> >
> > No, Load Time is the same as Response Time and overall Elapsed time.
> >
> >>  I can't find the Start Time and End Time! Do they exist in JMeter?
> >
> > The default Time Stamp is start time;  end time = start + elapsed.
> >
> >>  Help!
> >>
> >> --
> >>  View this message in context:
> >> http://www.nabble.com/Response-Time-of-Http-Request-tp23148393p23148393.
> >>html Sent from the JMeter - User mailing list archive at Nabble.com.
> >>
> >>
> >>  -
> >>  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >>  For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
> >
> > -
> > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Response Time of Http Request

[Noel O'Brien]

Latency is the time taken until the first response is received by JMeter. Load 
time is the total time to complete the request/response.

Quoting sebb from a previous mail thread where I had a similar question:

"Latency is time to first response.
This may be the entire response, especially for small payloads."

Regards,
Noel

On Wednesday 22 April 2009 09:56:38 Kannan, Sengamalam wrote:
> Response time for a HTTP request = Load Time - Elapsed time
>
> Is this correct? Or is Load time and Response time are the same?
>
>
> Thanks & Regards
> Sengamalam Kannan
>
> -Original Message-
> From: Noel O'Brien [mailto:nobr...@newbay.com]
> Sent: Wednesday, April 22, 2009 2:23 PM
> To: jmeter-user@jakarta.apache.org
> Cc: linuxos
> Subject: Re: Response Time of Http Request
>
> Elapsed Time  == Load Time in  "View Results Tree" listener.
>
> Regards,
> Noel
>
> On Wednesday 22 April 2009 08:30:09 linuxos wrote:
> > Thanks. But I can't find the Elapsed time in the "Sampler result".
> > Here is my result from the "View Results Tree":
> >
> > Thread Name: Thread Group 1-1
> > Sample Start: 2009-04-22 14:54:01 CST
> > Load time: 1489
> > Latency: 1360
> > Size in bytes: 81648
> > Sample Count: 1
> > Error Count: 0
> > Response code: 200
> > Response message: OK
> >
> > Is there any wrong step in using the JMeter?
> >
> > sebb-2-2 wrote:
> > > On 21/04/2009, linuxos  wrote:
> > >>  I've just started using JMeter 2.3.2. I created a "Thread Group",
>
> added
>
> > >> "HTTP
> > >>  Request Sampler" and "View Results Tree" for my testing.
> > >>
> > >>  For calculation of "Response Time" of a request, is it correct to
>
> say
>
> > >> it's
> > >>  the "Load time" minus "Latency"?
> > >
> > > No, Load Time is the same as Response Time and overall Elapsed time.
> > >
> > >>  I can't find the Start Time and End Time! Do they exist in JMeter?
> > >
> > > The default Time Stamp is start time;  end time = start + elapsed.
> > >
> > >>  Help!
> > >>
> > >> --
> > >>  View this message in context:
>
> http://www.nabble.com/Response-Time-of-Http-Request-tp23148393p23148393.
>
> > >>html Sent from the JMeter - User mailing list archive at Nabble.com.
>
> -
>
> > >>  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > >>  For additional commands, e-mail:
>
> jmeter-user-h...@jakarta.apache.org
>
>
> -
>
> > > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> > > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
>
> -
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Response Time of Http Request

[Noel O'Brien]

On Wednesday 22 April 2009 11:29:04 linuxos wrote:
> I don't know why the HTTP Request Samples are started randomly. So, I guess
> waiting time is included in the "Load time" or "Latency", and I can't
> calculate the actual time used between the request and response.

What do you mean "started randomly"?

AFAIK, for the HTTP Sampler the load time is the time taken to establish a 
connection (retrying if necessary), send the request and receive the entire 
response.

How do you define "Waiting Time"? If you mean delays caused by saturated 
networks and/or processing time on the server side, there's no way to measure 
that independently; it's included both the load and latency time. 

> Also, I found the "Load time" is very similar to the "Latency"!

I've found that in most cases latency is equal to load time, because my 
product sends back small response payloads, which are received by JMeter as 
the one and only response packet for the request send.

If in doubt, use Wireshark / tcpdump to sniff the network data. You'll be able 
to see from the timestamps in the wireshark dump when packets were sent

> Any suggestion?
>
> Noel O'Brien wrote:
> > Latency is the time taken until the first response is received by JMeter.
> > Load
> > time is the total time to complete the request/response.
> >
> > Quoting sebb from a previous mail thread where I had a similar question:
> >
> > "Latency is time to first response.
> > This may be the entire response, especially for small payloads."
> >
> > Regards,
> > Noel
> >
> > On Wednesday 22 April 2009 09:56:38 Kannan, Sengamalam wrote:
> >> Response time for a HTTP request = Load Time - Elapsed time
> >>
> >> Is this correct? Or is Load time and Response time are the same?
> >>
> >>
> >> Thanks & Regards
> >> Sengamalam Kannan
> >>
> >> -Original Message-
> >> From: Noel O'Brien [mailto:nobr...@newbay.com]
> >> Sent: Wednesday, April 22, 2009 2:23 PM
> >> To: jmeter-user@jakarta.apache.org
> >> Cc: linuxos
> >> Subject: Re: Response Time of Http Request
> >>
> >> Elapsed Time  == Load Time in  "View Results Tree" listener.
> >>
> >> Regards,
> >> Noel
> >>
> >> On Wednesday 22 April 2009 08:30:09 linuxos wrote:
> >> > Thanks. But I can't find the Elapsed time in the "Sampler result".
> >> > Here is my result from the "View Results Tree":
> >> >
> >> > Thread Name: Thread Group 1-1
> >> > Sample Start: 2009-04-22 14:54:01 CST
> >> > Load time: 1489
> >> > Latency: 1360
> >> > Size in bytes: 81648
> >> > Sample Count: 1
> >> > Error Count: 0
> >> > Response code: 200
> >> > Response message: OK
> >> >
> >> > Is there any wrong step in using the JMeter?
> >> >
> >> > sebb-2-2 wrote:
> >> > > On 21/04/2009, linuxos  wrote:
> >> > >>  I've just started using JMeter 2.3.2. I created a "Thread Group",
> >>
> >> added
> >>
> >> > >> "HTTP
> >> > >>  Request Sampler" and "View Results Tree" for my testing.
> >> > >>
> >> > >>  For calculation of "Response Time" of a request, is it correct to
> >>
> >> say
> >>
> >> > >> it's
> >> > >>  the "Load time" minus "Latency"?
> >> > >
> >> > > No, Load Time is the same as Response Time and overall Elapsed time.
> >> > >
> >> > >>  I can't find the Start Time and End Time! Do they exist in JMeter?
> >> > >
> >> > > The default Time Stamp is start time;  end time = start + elapsed.
> >> > >
> >> > >>  Help!
> >> > >>
> >> > >> --
> >> > >>  View this message in context:
> >>
> >> http://www.nabble.com/Response-Time-of-Http-Request-tp23148393p23148393.
> >>
> >> > >>html Sent from the JMeter - User mailing list archive at Nabble.com.
> >>
> >> -
> >>
> >> > >>  To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >> > >>  For additional commands, e-mail:
> >>
> >> jmeter-user-h...@jakarta.apache.org
> >>
> >>
> >> -
> >>
> >> > > To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >> > > For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
> >>
> >> -
> >> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org
> >> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org

Re: Response Time of Http Request

[Noel O'Brien]

Are the times still 1-2 with one thread? 

If not, then that may indicate: 
1. High network delays 
2. Possible Resource Contention in your server due to multi-threading, etc. 

Use tcpdump or some other network analysis tool to confirm the times are high. 
You should be able to deduce where the discrepancy is coming from 

Regards, 
Noel 

- Original Message - 
From: "linuxos"  
To: jmeter-user@jakarta.apache.org 
Sent: Friday, 24 April, 2009 06:36:42 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: Response Time of Http Request 


In the Sampler result, I don't know why the "Load time" and "Latency" are 
often over 1 and sometimes over 2, but the actual time per request 
should be 2-4 seconds if run in browser. 

My "Thread Properties": 
Number of Threads (users): 5 
Ramp-Up Period (in seconds): 0 
Loop Count: 1 

If "Load time" is divided by "Number of Threads", then I think the result is 
reasonable. Did I set something wrong? 



Noel O'Brien wrote: 
> 
> On Wednesday 22 April 2009 11:29:04 linuxos wrote: 
>> I don't know why the HTTP Request Samples are started randomly. So, I 
>> guess 
>> waiting time is included in the "Load time" or "Latency", and I can't 
>> calculate the actual time used between the request and response. 
> 
> What do you mean "started randomly"? 
> 
> AFAIK, for the HTTP Sampler the load time is the time taken to establish a 
> connection (retrying if necessary), send the request and receive the 
> entire 
> response. 
> 
> How do you define "Waiting Time"? If you mean delays caused by saturated 
> networks and/or processing time on the server side, there's no way to 
> measure 
> that independently; it's included both the load and latency time. 
> 
>> Also, I found the "Load time" is very similar to the "Latency"! 
> 
> I've found that in most cases latency is equal to load time, because my 
> product sends back small response payloads, which are received by JMeter 
> as 
> the one and only response packet for the request send. 
> 
> If in doubt, use Wireshark / tcpdump to sniff the network data. You'll be 
> able 
> to see from the timestamps in the wireshark dump when packets were sent 
> 
>> Any suggestion? 
>> 
>> Noel O'Brien wrote: 
>> > Latency is the time taken until the first response is received by 
>> JMeter. 
>> > Load 
>> > time is the total time to complete the request/response. 
>> > 
>> > Quoting sebb from a previous mail thread where I had a similar 
>> question: 
>> > 
>> > "Latency is time to first response. 
>> > This may be the entire response, especially for small payloads." 
>> > 
>> > Regards, 
>> > Noel 
>> > 
>> > On Wednesday 22 April 2009 09:56:38 Kannan, Sengamalam wrote: 
>> >> Response time for a HTTP request = Load Time - Elapsed time 
>> >> 
>> >> Is this correct? Or is Load time and Response time are the same? 
>> >> 
>> >> 
>> >> Thanks & Regards 
>> >> Sengamalam Kannan 
>> >> 
>> >> -Original Message- 
>> >> From: Noel O'Brien [mailto:nobr...@newbay.com] 
>> >> Sent: Wednesday, April 22, 2009 2:23 PM 
>> >> To: jmeter-user@jakarta.apache.org 
>> >> Cc: linuxos 
>> >> Subject: Re: Response Time of Http Request 
>> >> 
>> >> Elapsed Time == Load Time in "View Results Tree" listener. 
>> >> 
>> >> Regards, 
>> >> Noel 
>> >> 
>> >> On Wednesday 22 April 2009 08:30:09 linuxos wrote: 
>> >> > Thanks. But I can't find the Elapsed time in the "Sampler result". 
>> >> > Here is my result from the "View Results Tree": 
>> >> > 
>> >> > Thread Name: Thread Group 1-1 
>> >> > Sample Start: 2009-04-22 14:54:01 CST 
>> >> > Load time: 1489 
>> >> > Latency: 1360 
>> >> > Size in bytes: 81648 
>> >> > Sample Count: 1 
>> >> > Error Count: 0 
>> >> > Response code: 200 
>> >> > Response message: OK 
>> >> > 
>> >> > Is there any wrong step in using the JMeter? 
>> >> > 
>> >> > sebb-2-2 wrote: 
>> >> > > On 21/04/2009, linuxos  wrote: 
>> >> > >> I've just started using JMeter 2.3.2. I created a "Thread 
&g

Feasibility of addind a "blocked" test result

[Noel O'Brien]

Hi, 

Is is possible to have samplers report a "blocked" status if they're not 
executed (e.g. when they're inside an IF Controller that does not execute it's 
children) ? 

-- 
Regards, 
Noel 

HTTP Request File Path Dynamically Calculated?

[Noel O'Brien]

Hi , 

Does anyone know if variables in the file paths of HTTP requests are calculated 
each time the sampler is executed or only once, upon initialisation? I'm trying 
to use a HTTP sampler in a Loop controller to send a file chunk each time. I 
use the following path: 

${testplan.root}/data/resources/${medium.3gpp.filename}_${current.chunk.index} 

where ${current.chunk.index} increases each time the Loop Controller interates, 
but it seems to keep sending the first file chunk, which leads me to believe 
this is not evaluated each time the sampler is executed. 

Is there a workaround for this? 

-- 
Regards, 
Noel 

Re: HTTP Request File Path Dynamically Calculated?

[Noel O'Brien]

Ok, I'll try that, thanks Sebb 

Regards, 
Noel 
- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 5 May, 2009 20:27:54 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: HTTP Request File Path Dynamically Calculated? 

On 05/05/2009, Noel O'Brien  wrote: 
> Hi , 
> 
> Does anyone know if variables in the file paths of HTTP requests are 
> calculated each time the sampler is executed or only once, upon 
> initialisation? I'm trying to use a HTTP sampler in a Loop controller to send 
> a file chunk each time. I use the following path: 
> 
> ${testplan.root}/data/resources/${medium.3gpp.filename}_${current.chunk.index}
>  
> 
> where ${current.chunk.index} increases each time the Loop Controller 
> interates, but it seems to keep sending the first file chunk, which leads me 
> to believe this is not evaluated each time the sampler is executed. 

There is a bug in 2.3.2 - the first file entry does not get re-evaluated. 
This has been fixed in SVN. 

> Is there a workaround for this? 

Use a dummy first file if you can. 

> -- 
> Regards, 
> 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: HTTP Request File Path Dynamically Calculated?

[Noel O'Brien]

Well, the option of setting a dummy file (empty .txt) as the first file didn't 
work for me for various reasons; the Content-Length was being set to 0 and our 
product required it to be accurate. I wrote a pre-processor for it and this 
works for me. Hopefully it will help someone else: 

>>>>>>>>>>>>>>>>> 

import org.apache.jmeter.protocol.http.util.*; 

/* 
This script is require as a workaround for a bug in JMeter 2.3.2. From the 
JMeter forums: 

"There is a bug in 2.3.2 - the first file entry does not get re-evaluated. This 
has been fixed in SVN." 

Basically, if using variables to point at file chunks while iterating through a 
loop, the variables are 
evaluated on the first iteration and not on subsequent iterations. This script 
works around the issue by 
setting the correct file chunk on each iteration. 
*/ 

// Parameters are split on `|` and are assinged as follows (defaults in 
brackets): 
// 1. Filename of media chunk file (REQUIRED). 

log.info("Starting AddMediaFileChunk.bsh"); 
log.debug("AddMediaFileChunk.bsh: Parameters: " + Parameters); 
startTime = System.currentTimeMillis(); 

params = Parameters.split("\\|"); 

httpSampler = ctx.getCurrentSampler(); 
httpFiles = httpSampler.getHTTPFiles(); 

log.debug("AddMediaFileChunk.bsh: Adding: "+params[0]); 

// Create file arg, add it to the file arg list and set it. 
// Existing files specified inthe HTTP Sampler will be lost. 
HTTPFileArg fileChunk = new HTTPFileArg(params[0]); 
HTTPFileArg [] newList = {fileChunk}; 
httpSampler.setHTTPFiles(newList); 

finishTime = System.currentTimeMillis(); 
log.info("Finished AddMediaFileChunk.bsh (" + (finishTime-startTime) + "ms)"); 


>>>>>>>>>>>>>>> 

I pass a single parameter into the script, which is the location of the file I 
need to upload (in my case, it's a file which it a chunk of a larger file, 
broken up using `split`). By omitting the Content-Length header from the HTTP 
Header config element the Sampler seems to set it correctly. 

When I try to set the Content-Length header programatically (using 
fileChunk.setHeader("Content-Length: 1") for example) it doesn't seem to 
get set, because when I read it back using getHeader() it always returns 
`null`. Is this a bug? 

Regards, 
Noel 

- Original Message - 
From: "Noel O'Brien"  
To: "JMeter Users List"  
Sent: Wednesday, 6 May, 2009 09:16:38 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: HTTP Request File Path Dynamically Calculated? 

Ok, I'll try that, thanks Sebb 

Regards, 
Noel 
- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 5 May, 2009 20:27:54 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: HTTP Request File Path Dynamically Calculated? 

On 05/05/2009, Noel O'Brien  wrote: 
> Hi , 
> 
> Does anyone know if variables in the file paths of HTTP requests are 
> calculated each time the sampler is executed or only once, upon 
> initialisation? I'm trying to use a HTTP sampler in a Loop controller to send 
> a file chunk each time. I use the following path: 
> 
> ${testplan.root}/data/resources/${medium.3gpp.filename}_${current.chunk.index}
>  
> 
> where ${current.chunk.index} increases each time the Loop Controller 
> interates, but it seems to keep sending the first file chunk, which leads me 
> to believe this is not evaluated each time the sampler is executed. 

There is a bug in 2.3.2 - the first file entry does not get re-evaluated. 
This has been fixed in SVN. 

> Is there a workaround for this? 

Use a dummy first file if you can. 

> -- 
> Regards, 
> 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 


-- 
Regards, 
Noel 

Recording from J2ME WTK App

[Noel O'Brien]

Hi All, 

I'm trying to use the Recording Proxy to capture the traffic from our client (a 
J2ME app which uses HTTP) to our server. I've set up JMeter (2.3.2) correctly 
and when I enable my browser to user the JMeter proxy (as a test) it 
successfully records requests. 

I've set the proxy settings in the Wireless Toolkit (Sun, 2.5.2) and run the 
emulator. Traffic is hitting the proxy but I'm getting "[Sample Failed]" and 
"Cannot handle CONNECT - probably used HTTPS". I'm not using HTTPS , all the 
traffic is HTTP. From looking at the code, the problem seems to be that the 
JMeter proxy won't handle a CONNECT request (I've substituted the url): 


2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: browser 
request: CONNECT :80 HTTP/1.1 

2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parser 
input: CONNECT :80 HTTP/1.1 

2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parsed 
method: CONNECT 
2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parsed 
url: :80 
2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parsed 
version:HTTP/1.1 
2009/05/19 13:50:45 ERROR - jmeter.protocol.http.proxy.Proxy: Not implemented 
(probably used https) java.lang.IllegalArgumentException: Cannot handle CONNECT 
- probably used HTTPS 
at 
org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parseFirstLine(HttpRequestHdr.java:212)
 
at 
org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parse(HttpRequestHdr.java:164)
 
at org.apache.jmeter.protocol.http.proxy.Proxy.run(Proxy.java:165) 

Any thoughts on how to get around this problem? 

-- 
Regards, 
Noel 

Re: Recording from J2ME WTK App

[Noel O'Brien]

Hi, 

Thanks for you input sebb. I was planning to use it for building a traffic 
model for use in performance testing, which is due to start tomorrow. I don't 
know how much work it is to set up a real proxy so I think it would be more 
wise to spend my time picking up the traffic model with Wireshark and manually 
inputting it into JMeter. 

For what it's worth, this seems to be the answer to my question: 
http://forums.java.net/jive/thread.jspa?messageID=296830&tstart=0#296830 

The WTK tunnels both HTTP and HTTPS, so unfortunately it won't work. There is a 
code workaround posted in that link so I might be abe to convince the 
developers here to stick it in for me ;) 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 19 May, 2009 14:42:29 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: Recording from J2ME WTK App 

On 19/05/2009, Noel O'Brien  wrote: 
> Hi All, 
> 
> I'm trying to use the Recording Proxy to capture the traffic from our client 
> (a J2ME app which uses HTTP) to our server. I've set up JMeter (2.3.2) 
> correctly and when I enable my browser to user the JMeter proxy (as a test) 
> it successfully records requests. 
> 
> I've set the proxy settings in the Wireless Toolkit (Sun, 2.5.2) and run the 
> emulator. Traffic is hitting the proxy but I'm getting "[Sample Failed]" and 
> "Cannot handle CONNECT - probably used HTTPS". I'm not using HTTPS , all the 
> traffic is HTTP. From looking at the code, the problem seems to be that the 
> JMeter proxy won't handle a CONNECT request (I've substituted the url): 
> 
> 
> 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> browser request: CONNECT :80 HTTP/1.1 
> 
> 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parser 
> input: CONNECT :80 HTTP/1.1 
> 
> 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parsed 
> method: CONNECT 
> 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parsed 
> url: :80 
> 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: parsed 
> version:HTTP/1.1 
> 2009/05/19 13:50:45 ERROR - jmeter.protocol.http.proxy.Proxy: Not implemented 
> (probably used https) java.lang.IllegalArgumentException: Cannot handle 
> CONNECT - probably used HTTPS 
> at 
> org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parseFirstLine(HttpRequestHdr.java:212)
>  
> at 
> org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parse(HttpRequestHdr.java:164)
>  
> at org.apache.jmeter.protocol.http.proxy.Proxy.run(Proxy.java:165) 
> 
> Any thoughts on how to get around this problem? 

JMeter Proxy does not support CONNECT; I doubt it ever will, as it is 
not intended as a general-purpose proxy. 

Not sure why the application should be sending a CONNECT for HTTP; 
perhaps it is due to the way the proxy was set up? Maybe because the 
Sun toolkit is wireless, it uses CONNECT to tunnel SSL over HTTP? 

You could try using a "real" HTTP proxy to see what traffic is 
actually being sent. If that can record the requests to a file, you 
should be able to use that to build up a JMeter test plan. 

> -- 
> Regards, 
> 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: Recording from J2ME WTK App

[Noel O'Brien]

Hi, 

As it turns out, the developers had that code in already, just needed to set 
some properties and re-built the app 

Now it's working :) 

Thanks for your help. 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 19 May, 2009 15:38:20 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: Recording from J2ME WTK App 

On 19/05/2009, Noel O'Brien  wrote: 
> Hi, 
> 
> Thanks for you input sebb. I was planning to use it for building a traffic 
> model for use in performance testing, which is due to start tomorrow. I don't 
> know how much work it is to set up a real proxy so I think it would be more 
> wise to spend my time picking up the traffic model with Wireshark and 
> manually inputting it into JMeter. 

There are several proxies that are quite easy to set up. 

For example: 

http://wiki.apache.org/jakarta-jmeter/NetworkSniffer 

Try TCPMon. 

If the requests are not all that complicated, you can record the URLs 
in a file and use CSV Dataset to read them. Full URLs can be used in 
the HTTP Path: field. 

It gets a bit more complicated for POST requests, unless these always 
have the same number of parameters. 

> For what it's worth, this seems to be the answer to my question: 
> http://forums.java.net/jive/thread.jspa?messageID=296830&tstart=0#296830 

Thanks! 

> The WTK tunnels both HTTP and HTTPS, so unfortunately it won't work. There is 
> a code workaround posted in that link so I might be abe to convince the 
> developers here to stick it in for me ;) 
> 
> Regards, 
> Noel 
> 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> Sent: Tuesday, 19 May, 2009 14:42:29 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: Recording from J2ME WTK App 
> 
> On 19/05/2009, Noel O'Brien  wrote: 
> > Hi All, 
> > 
> > I'm trying to use the Recording Proxy to capture the traffic from our 
> > client (a J2ME app which uses HTTP) to our server. I've set up JMeter 
> > (2.3.2) correctly and when I enable my browser to user the JMeter proxy (as 
> > a test) it successfully records requests. 
> > 
> > I've set the proxy settings in the Wireless Toolkit (Sun, 2.5.2) and run 
> > the emulator. Traffic is hitting the proxy but I'm getting "[Sample 
> > Failed]" and "Cannot handle CONNECT - probably used HTTPS". I'm not using 
> > HTTPS , all the traffic is HTTP. From looking at the code, the problem 
> > seems to be that the JMeter proxy won't handle a CONNECT request (I've 
> > substituted the url): 
> > 
> > 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > browser request: CONNECT :80 HTTP/1.1 
> > 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parser input: CONNECT :80 HTTP/1.1 
> > 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parsed method: CONNECT 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parsed url: :80 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parsed version:HTTP/1.1 
> > 2009/05/19 13:50:45 ERROR - jmeter.protocol.http.proxy.Proxy: Not 
> > implemented (probably used https) java.lang.IllegalArgumentException: 
> > Cannot handle CONNECT - probably used HTTPS 
> > at 
> > org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parseFirstLine(HttpRequestHdr.java:212)
> >  
> > at 
> > org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parse(HttpRequestHdr.java:164)
> >  
> > at org.apache.jmeter.protocol.http.proxy.Proxy.run(Proxy.java:165) 
> > 
> > Any thoughts on how to get around this problem? 
> 
> JMeter Proxy does not support CONNECT; I doubt it ever will, as it is 
> not intended as a general-purpose proxy. 
> 
> Not sure why the application should be sending a CONNECT for HTTP; 
> perhaps it is due to the way the proxy was set up? Maybe because the 
> Sun toolkit is wireless, it uses CONNECT to tunnel SSL over HTTP? 
> 
> You could try using a "real" HTTP proxy to see what traffic is 
> actually being sent. If that can record the requests to a file, you 
> should be able to use that to build up a JMeter test plan. 
> 
> > -- 
> > Regards, 
> > 
> > Noel 
> > 
> 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 
> 
> -- 
> Regards, 
> 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: Recording from J2ME WTK App

[Noel O'Brien]

Sure, it was proxyHost and proxyPort 

I'm not sure whether they're standard J2ME properties (I presume they are :) ) 

I needed to update them to point at the JMeter proxy running on port 9000 
: 
proxyHost=localhost 
proxyPort=9000 

Regards, 
Noel 

- Original Message - 
From: "Suvendu_Mohapatra"  
To: "JMeter Users List"  
Sent: Tuesday, 19 May, 2009 15:49:35 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: RE: Recording from J2ME WTK App 

Hi, 

Kindly would you like to mention which properties you have set?? 

With Regards, 
Suvendu 

-Original Message- 
From: Noel O'Brien [mailto:nobr...@newbay.com] 
Sent: Tuesday, May 19, 2009 8:16 PM 
To: JMeter Users List 
Subject: Re: Recording from J2ME WTK App 

Hi, 

As it turns out, the developers had that code in already, just needed to set 
some properties and re-built the app 

Now it's working :) 

Thanks for your help. 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 19 May, 2009 15:38:20 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: Recording from J2ME WTK App 

On 19/05/2009, Noel O'Brien  wrote: 
> Hi, 
> 
> Thanks for you input sebb. I was planning to use it for building a traffic 
> model for use in performance testing, which is due to start tomorrow. I don't 
> know how much work it is to set up a real proxy so I think it would be more 
> wise to spend my time picking up the traffic model with Wireshark and 
> manually inputting it into JMeter. 

There are several proxies that are quite easy to set up. 

For example: 

http://wiki.apache.org/jakarta-jmeter/NetworkSniffer 

Try TCPMon. 

If the requests are not all that complicated, you can record the URLs 
in a file and use CSV Dataset to read them. Full URLs can be used in 
the HTTP Path: field. 

It gets a bit more complicated for POST requests, unless these always 
have the same number of parameters. 

> For what it's worth, this seems to be the answer to my question: 
> http://forums.java.net/jive/thread.jspa?messageID=296830&tstart=0#296830 

Thanks! 

> The WTK tunnels both HTTP and HTTPS, so unfortunately it won't work. There is 
> a code workaround posted in that link so I might be abe to convince the 
> developers here to stick it in for me ;) 
> 
> Regards, 
> Noel 
> 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> Sent: Tuesday, 19 May, 2009 14:42:29 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: Recording from J2ME WTK App 
> 
> On 19/05/2009, Noel O'Brien  wrote: 
> > Hi All, 
> > 
> > I'm trying to use the Recording Proxy to capture the traffic from our 
> > client (a J2ME app which uses HTTP) to our server. I've set up JMeter 
> > (2.3.2) correctly and when I enable my browser to user the JMeter proxy (as 
> > a test) it successfully records requests. 
> > 
> > I've set the proxy settings in the Wireless Toolkit (Sun, 2.5.2) and run 
> > the emulator. Traffic is hitting the proxy but I'm getting "[Sample 
> > Failed]" and "Cannot handle CONNECT - probably used HTTPS". I'm not using 
> > HTTPS , all the traffic is HTTP. From looking at the code, the problem 
> > seems to be that the JMeter proxy won't handle a CONNECT request (I've 
> > substituted the url): 
> > 
> > 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > browser request: CONNECT :80 HTTP/1.1 
> > 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parser input: CONNECT :80 HTTP/1.1 
> > 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parsed method: CONNECT 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parsed url: :80 
> > 2009/05/19 13:50:45 DEBUG - jmeter.protocol.http.proxy.HttpRequestHdr: 
> > parsed version:HTTP/1.1 
> > 2009/05/19 13:50:45 ERROR - jmeter.protocol.http.proxy.Proxy: Not 
> > implemented (probably used https) java.lang.IllegalArgumentException: 
> > Cannot handle CONNECT - probably used HTTPS 
> > at 
> > org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parseFirstLine(HttpRequestHdr.java:212)
> >  
> > at 
> > org.apache.jmeter.protocol.http.proxy.HttpRequestHdr.parse(HttpRequestHdr.java:164)
> >  
> > at org.apache.jmeter.protocol.http.proxy.Proxy.run(Proxy.java:165) 
> > 
> > Any thoughts on how to get around this problem? 
> 
> JMeter Proxy does not support CONNECT; I doubt it ever will, as it is 
> not intended a

User Defined Beanshell Function

[Noel O'Brien]

Hi All, 

Is it possible to define a beanshell function that I can use throughout my test 
in the same way as the built in functions, e.g. instead of using 
${__threadNum()} 
I can use ${__encrypt("raw string")} ? 
-- 
Regards, 
Noel 

Re: User Defined Beanshell Function

[Noel O'Brien]

Thanks for your explanation of this Sebb, very insightful! 

Sounds like a bit of work to make a native function, and to be honest 
${__BeanShell(functionName())} would suit my needs perfectly. 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Thursday, 21 May, 2009 18:56:24 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: User Defined Beanshell Function 

On 21/05/2009, Noel O'Brien  wrote: 
> Hi All, 
> 
> Is it possible to define a beanshell function that I can use throughout my 
> test in the same way as the built in functions, e.g. instead of using 
> ${__threadNum()} 
> I can use ${__encrypt("raw string")} ? 

You can define a Beanhell function in an init file such as 
BeanShellFunction.bshrc, 
but this will need to be invoked using __BeanShell(functionName()) 

It's not possible to register the BeanShell function as a "native" function. 

However, you can of course create your own function, just make sure it 
implements org.apache.jmeter.functions.Function 

The package name must contain ".functions." but not ".gui." - see 
jmeter.properties "classfinder.functions" for details. 

The code would need to be put in a jar and stored in lib/ext. 

> -- 
> Regards, 
> 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: how to confirm that my .jmx script is pass or failed?

[Noel O'Brien]

Have a look here: 
http://jakarta.apache.org/jmeter/usermanual/best-practices.html 

Look at section 16.6 Reducing resource requirements 
in particular. 

Namely, you need to start JMeter with the -l flag and also set it up to log 
results in CSV format. Additionally, only log the result data that you 
absolutely need. 

Regards, 
Noel 

- Original Message - 
From: "gpub"  
To: jmeter-user@jakarta.apache.org 
Sent: Friday, 22 May, 2009 11:43:12 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: how to confirm that my .jmx script is pass or failed? 


Hi, 

how do i need to confirm that my jmeter script has passed or got failed. 
I am not using any listener at all(due to heavy load performance) 
from where i came to know about the errors ?? 

Thanks 
-Gp 
-- 
View this message in context: 
http://www.nabble.com/how-to-confirm-that-my-.jmx-script-is-pass-or-failed--tp23668149p23668149.html
 
Sent from the JMeter - User mailing list archive at Nabble.com. 


- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: how to confirm that my .jmx script is pass or failed?

[Noel O'Brien]

In linux I run non-gui mode like this: 

/usr/bin/jmeter/bin/jmeter -n -t MyTest.jmx -l MyResults.jtl 

If you're using windows, adjust appropriately 

Regards, 
Noel 

- Original Message - 
From: "gpub"  
To: jmeter-user@jakarta.apache.org 
Sent: Friday, 22 May, 2009 12:09:45 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: RE: how to confirm that my .jmx script is pass or failed? 


I am using GUi mode. 
and with non gui mode you told to pass .jtl file on command line argument. 
i dont uderstand about this. please let me understand to pass .jtl on cmmd 
line. 
so that i can try to work on non gui. 


suvendu wrote: 
> 
> Are you running the jmeter in gui mode or non-gui mode?? 
> In non-gui mode you can pass .jtl file as a command line argument. 
> Just try the same method when running in gui modeI have never tried 
> it... 
> 
> With Regards, 
> Suvendu 
> 
> -Original Message- 
> From: gpub [mailto:affablepr...@gmail.com] 
> Sent: Friday, May 22, 2009 4:13 PM 
> To: jmeter-user@jakarta.apache.org 
> Subject: how to confirm that my .jmx script is pass or failed? 
> 
> 
> Hi, 
> 
> how do i need to confirm that my jmeter script has passed or got failed. 
> I am not using any listener at all(due to heavy load performance) 
> from where i came to know about the errors ?? 
> 
> Thanks 
> -Gp 
> -- 
> View this message in context: 
> http://www.nabble.com/how-to-confirm-that-my-.jmx-script-is-pass-or-failed--tp23668149p23668149.html
>  
> Sent from the JMeter - User mailing list archive at Nabble.com. 
> 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 
> 
> DISCLAIMER: 
> This email (including any attachments) is intended for the sole use of the 
> intended recipient/s and may contain material that is CONFIDENTIAL AND 
> PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying 
> or distribution or forwarding of any or all of the contents in this 
> message is STRICTLY PROHIBITED. If you are not the intended recipient, 
> please contact the sender by email and delete all copies; your cooperation 
> in this regard is appreciated. 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 
> 

-- 
View this message in context: 
http://www.nabble.com/how-to-confirm-that-my-.jmx-script-is-pass-or-failed--tp23668149p23668460.html
 
Sent from the JMeter - User mailing list archive at Nabble.com. 


- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: which driver file should i use??

[Noel O'Brien]

>From the ClassNotFoundException I'd guess JMeter can't find the driver class 
>you're referencing. Either: 

1. You need to add the ojdbc jar file to JMETER_HOME/lib or, 
2. You've referenced an invalid class for the driver 

Regards, 
Noel 

- Original Message - 
From: "gpub"  
To: jmeter-user@jakarta.apache.org 
Sent: Monday, 25 May, 2009 12:05:14 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: which driver file should i use?? 


hi, 
i am getting this errror 
WARN - jmeter.protocol.jdbc.config.DataSourceElement: Could not load 
driver: oracle.jdbc.driver.OracleDriver java.lang.ClassNotFoundException: 
oracle/jdbc/driver/OracleDriver 

i have use following details in JDBC config 
jdbc:oracle:thin:@//db_name_abc.XXX.com:1524:CRP3 
JDBC Driver Class: oracle.jdbc.driver.OracleDriver 
username:  
password: 

My db details is: 
CRP3 = (DESCRIPTION = (ADDRESS = (PROTOCOL = tcp)(HOST = 
db_name_abc..com)(PORT = 1524)) 
(CONNECT_DATA = (SID = CRP3) 
) ) 

what could be the prob 

-- 
View this message in context: 
http://www.nabble.com/which-driver-file-should-i-use---tp23705248p23705248.html 
Sent from the JMeter - User mailing list archive at Nabble.com. 


- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: which driver file should i use??

[Noel O'Brien]

I can successfully make an Oracle connection with: 


1. Only ojdbc14.jar added to JMETER_HOME/lib 
2. Database URL: jdbc:oracle:thin:@(DESCRIPTION = ..(omitted for 
security reasons) 
3. JDBC Driver Class: oracle.jdbc.OracleDriver 

If in doubt, make a connection from another tool like SQLExplorer to test the 
connection 

Regards, 
Noel 

- Original Message - 
From: "gpub"  
To: jmeter-user@jakarta.apache.org 
Sent: Monday, 25 May, 2009 12:23:10 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: which driver file should i use?? 


hey Noel 

I have added 4 jar files in Jmeter\lib 
classes12.jar, odbc14.jar, classes111_g.jar and classes12dms.jar 

now, which 1 should i use all or which one among thez?? 
using oracle DB 
and >You've referenced an invalid class for the driver ??(reference ; 
oracle.jdbc.driver.OracleDriver ) 
get jar from D:\oracle\ora92\jdbc\lib 



Noel O'Brien wrote: 
> 
> From the ClassNotFoundException I'd guess JMeter can't find the driver 
> class you're referencing. Either: 
> 
> 1. You need to add the ojdbc jar file to JMETER_HOME/lib or, 
> 2. You've referenced an invalid class for the driver 
> 
> Regards, 
> Noel 
> 
> - Original Message - 
> From: "gpub"  
> To: jmeter-user@jakarta.apache.org 
> Sent: Monday, 25 May, 2009 12:05:14 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: which driver file should i use?? 
> 
> 
> hi, 
> i am getting this errror 
> WARN - jmeter.protocol.jdbc.config.DataSourceElement: Could not load 
> driver: oracle.jdbc.driver.OracleDriver java.lang.ClassNotFoundException: 
> oracle/jdbc/driver/OracleDriver 
> 
> i have use following details in JDBC config 
> jdbc:oracle:thin:@//db_name_abc.XXX.com:1524:CRP3 
> JDBC Driver Class: oracle.jdbc.driver.OracleDriver 
> username:  
> password: 
> 
> My db details is: 
> CRP3 = (DESCRIPTION = (ADDRESS = (PROTOCOL = tcp)(HOST = 
> db_name_abc..com)(PORT = 1524)) 
> (CONNECT_DATA = (SID = CRP3) 
> ) ) 
> 
> what could be the prob 
> 
> -- 
> View this message in context: 
> http://www.nabble.com/which-driver-file-should-i-use---tp23705248p23705248.html
>  
> Sent from the JMeter - User mailing list archive at Nabble.com. 
> 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 
> 
> -- 
> Regards, 
> Noel 
> 
> 

-- 
View this message in context: 
http://www.nabble.com/which-driver-file-should-i-use---tp23705248p23705511.html 
Sent from the JMeter - User mailing list archive at Nabble.com. 


- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Stac

[Noel O'Brien]

Hi All, 

I'm using JMeter (2.3.2) to simulate a traffic model for our application 
(server) based on the traffic that our client generates. I used the JMeter 
proxy to sniff the traffic and then to make it more realistic I re-organised 
the captured samples into a series or Random and Throughput controllers. This 
works fine until the point in the test plan where the random/throuput 
controller come into play nad the the following happens, halting the test: 

2009/05/25 16:07:54 ERROR - jmeter.threads.JMeterThread: Test failed! 
java.lang.StackOverflowError 
at java.util.HashMap.getEntry(HashMap.java:344) 
at java.util.LinkedHashMap.get(LinkedHashMap.java:280) 
at java.util.Collections$SynchronizedMap.get(Collections.java:1975) 
at 
org.apache.jmeter.testelement.AbstractTestElement.getProperty(AbstractTestElement.java:140)
 
at 
org.apache.jmeter.testelement.AbstractTestElement.getPropertyAsInt(AbstractTestElement.java:183)
 
at 
org.apache.jmeter.control.ThroughputController.getStyle(ThroughputController.java:109)
 
at 
org.apache.jmeter.control.ThroughputController.isDone(ThroughputController.java:207)
 
at 
org.apache.jmeter.control.InterleaveControl.currentReturnedNull(InterleaveControl.java:153)
 
at 
org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:94)
 
at org.apache.jmeter.control.GenericController.next(GenericController.java:146) 
at org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
at 
org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
 
at org.apache.jmeter.control.GenericController.next(GenericController.java:146) 
at org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
at 
org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
 
at org.apache.jmeter.control.GenericController.next(GenericController.java:146) 
at org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
at 
org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
 


Has anyone seen this before? If so what's causing it and is thete a way to work 
around it? I've tries the test plan with 2.3.3 also, but the same thing happens 

-- 
Regards, 
Noel 

Re: StackOverflowError

[Noel O'Brien]

Hi, 

I find it hard to reproduce with a simple test plan, but I'll try to put 
something together that will reproduce the issue 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 12:21:38 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: Stac 

On 25/05/2009, Noel O'Brien  wrote: 
> Hi All, 
> 
> I'm using JMeter (2.3.2) to simulate a traffic model for our application 
> (server) based on the traffic that our client generates. I used the JMeter 
> proxy to sniff the traffic and then to make it more realistic I re-organised 
> the captured samples into a series or Random and Throughput controllers. This 
> works fine until the point in the test plan where the random/throuput 
> controller come into play nad the the following happens, halting the test: 
> 
> 2009/05/25 16:07:54 ERROR - jmeter.threads.JMeterThread: Test failed! 
> java.lang.StackOverflowError 
> at java.util.HashMap.getEntry(HashMap.java:344) 
> at java.util.LinkedHashMap.get(LinkedHashMap.java:280) 
> at java.util.Collections$SynchronizedMap.get(Collections.java:1975) 
> at 
> org.apache.jmeter.testelement.AbstractTestElement.getProperty(AbstractTestElement.java:140)
>  
> at 
> org.apache.jmeter.testelement.AbstractTestElement.getPropertyAsInt(AbstractTestElement.java:183)
>  
> at 
> org.apache.jmeter.control.ThroughputController.getStyle(ThroughputController.java:109)
>  
> at 
> org.apache.jmeter.control.ThroughputController.isDone(ThroughputController.java:207)
>  
> at 
> org.apache.jmeter.control.InterleaveControl.currentReturnedNull(InterleaveControl.java:153)
>  
> at 
> org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:94)
>  
> at 
> org.apache.jmeter.control.GenericController.next(GenericController.java:146) 
> at 
> org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
> at 
> org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
>  
> at 
> org.apache.jmeter.control.GenericController.next(GenericController.java:146) 
> at 
> org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
> at 
> org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
>  
> at 
> org.apache.jmeter.control.GenericController.next(GenericController.java:146) 
> at 
> org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
> at 
> org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
>  
> 
> 
> Has anyone seen this before? If so what's causing it and is thete a way to 
> work around it? I've tries the test plan with 2.3.3 also, but the same thing 
> happens 
> 

It's not been reported before, as far as I know. 

Could you create a Bugzilla issue for it? 
Please attach a (simple) test plan that exibits the behaiour, and the 
log from running it. 

Thanks. 

> -- 
> Regards, 
> 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: StackOverflowError

[Noel O'Brien]

Hi, 

I've generated the HTML schema file. Can I send it to you personally, i.e. I 
don't want to post it publically for security reasons 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 14:56:18 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 

OK, that could be added later. 

Meanwhile, perhaps you can use the extras/schematic.* files to 
generate a summary of the test plan that fails and attach that. 
[There's no schematic.sh version, but it should be obvious what to do 
if you are not using Windows]. 

On 26/05/2009, Noel O'Brien  wrote: 
> Hi, 
> 
> I find it hard to reproduce with a simple test plan, but I'll try to put 
> something together that will reproduce the issue 
> 
> Regards, 
> Noel 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> Sent: Tuesday, 26 May, 2009 12:21:38 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: Stac 
> 
> On 25/05/2009, Noel O'Brien  wrote: 
> > Hi All, 
> > 
> > I'm using JMeter (2.3.2) to simulate a traffic model for our application 
> > (server) based on the traffic that our client generates. I used the JMeter 
> > proxy to sniff the traffic and then to make it more realistic I 
> > re-organised the captured samples into a series or Random and Throughput 
> > controllers. This works fine until the point in the test plan where the 
> > random/throuput controller come into play nad the the following happens, 
> > halting the test: 
> > 
> > 2009/05/25 16:07:54 ERROR - jmeter.threads.JMeterThread: Test failed! 
> > java.lang.StackOverflowError 
> > at java.util.HashMap.getEntry(HashMap.java:344) 
> > at java.util.LinkedHashMap.get(LinkedHashMap.java:280) 
> > at java.util.Collections$SynchronizedMap.get(Collections.java:1975) 
> > at 
> > org.apache.jmeter.testelement.AbstractTestElement.getProperty(AbstractTestElement.java:140)
> >  
> > at 
> > org.apache.jmeter.testelement.AbstractTestElement.getPropertyAsInt(AbstractTestElement.java:183)
> >  
> > at 
> > org.apache.jmeter.control.ThroughputController.getStyle(ThroughputController.java:109)
> >  
> > at 
> > org.apache.jmeter.control.ThroughputController.isDone(ThroughputController.java:207)
> >  
> > at 
> > org.apache.jmeter.control.InterleaveControl.currentReturnedNull(InterleaveControl.java:153)
> >  
> > at 
> > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:94)
> >  
> > at 
> > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> >  
> > at 
> > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
> > at 
> > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> >  
> > at 
> > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> >  
> > at 
> > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
> > at 
> > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> >  
> > at 
> > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> >  
> > at 
> > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83) 
> > at 
> > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> >  
> > 
> > 
> > Has anyone seen this before? If so what's causing it and is thete a way to 
> > work around it? I've tries the test plan with 2.3.3 also, but the same 
> > thing happens 
> > 
> 
> It's not been reported before, as far as I know. 
> 
> Could you create a Bugzilla issue for it? 
> Please attach a (simple) test plan that exibits the behaiour, and the 
> log from running it. 
> 
> Thanks. 
> 
> > -- 
> > Regards, 
> > 
> > Noel 
> > 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 
> 
> -- 
> Regards, 
> 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: StackOverflowError

[Noel O'Brien]

Hmm, doesn't look like attachyments are supported in this mailing list. Is 
there anywhere else I can put it? 

Regards, 
Noel 

- Original Message - 
From: "Noel O'Brien"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 16:12:16 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 


Ok, I've replaced it all, see attached 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 15:52:24 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 

That's fine - or you could just change the sensitive parts - e.g. use 
different URLs etc. 

On 26/05/2009, Noel O'Brien  wrote: 
> Hi, 
> 
> I've generated the HTML schema file. Can I send it to you personally, i.e. I 
> don't want to post it publically for security reasons 
> 
> 
> Regards, 
> Noel 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> 
> Sent: Tuesday, 26 May, 2009 14:56:18 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: StackOverflowError 
> 
> OK, that could be added later. 
> 
> Meanwhile, perhaps you can use the extras/schematic.* files to 
> generate a summary of the test plan that fails and attach that. 
> [There's no schematic.sh version, but it should be obvious what to do 
> if you are not using Windows]. 
> 
> On 26/05/2009, Noel O'Brien  wrote: 
> > Hi, 
> > 
> > I find it hard to reproduce with a simple test plan, but I'll try to put 
> > something together that will reproduce the issue 
> > 
> > Regards, 
> > Noel 
> > 
> > - Original Message - 
> > From: "sebb"  
> > To: "JMeter Users List"  
> > Sent: Tuesday, 26 May, 2009 12:21:38 GMT +00:00 GMT Britain, Ireland, 
> > Portugal 
> > Subject: Re: Stac 
> > 
> > On 25/05/2009, Noel O'Brien  wrote: 
> > > Hi All, 
> > > 
> > > I'm using JMeter (2.3.2) to simulate a traffic model for our application 
> > > (server) based on the traffic that our client generates. I used the 
> > > JMeter proxy to sniff the traffic and then to make it more realistic I 
> > > re-organised the captured samples into a series or Random and Throughput 
> > > controllers. This works fine until the point in the test plan where the 
> > > random/throuput controller come into play nad the the following happens, 
> > > halting the test: 
> > > 
> > > 2009/05/25 16:07:54 ERROR - jmeter.threads.JMeterThread: Test failed! 
> > > java.lang.StackOverflowError 
> > > at java.util.HashMap.getEntry(HashMap.java:344) 
> > > at java.util.LinkedHashMap.get(LinkedHashMap.java:280) 
> > > at java.util.Collections$SynchronizedMap.get(Collections.java:1975) 
> > > at 
> > > org.apache.jmeter.testelement.AbstractTestElement.getProperty(AbstractTestElement.java:140)
> > >  
> > > at 
> > > org.apache.jmeter.testelement.AbstractTestElement.getPropertyAsInt(AbstractTestElement.java:183)
> > >  
> > > at 
> > > org.apache.jmeter.control.ThroughputController.getStyle(ThroughputController.java:109)
> > >  
> > > at 
> > > org.apache.jmeter.control.ThroughputController.isDone(ThroughputController.java:207)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.currentReturnedNull(InterleaveControl.java:153)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:94)
> > >  
> > > at 
> > > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> > >  
> > > at 
> > > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> > >  
> > > at 
> > > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83)
> > >  
&g

Re: StackOverflowError

[Noel O'Brien]

Done: 
https://issues.apache.org/bugzilla/show_bug.cgi?id=47269 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 16:19:36 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 

Create the Bugzilla issue, re-open and add the attachment. 

On 26/05/2009, Noel O'Brien  wrote: 
> Hmm, doesn't look like attachyments are supported in this mailing list. Is 
> there anywhere else I can put it? 
> 
> Regards, 
> Noel 
> 
> 
> - Original Message - 
> From: "Noel O'Brien"  
> To: "JMeter Users List"  
> 
> Sent: Tuesday, 26 May, 2009 16:12:16 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: StackOverflowError 
> 
> 
> Ok, I've replaced it all, see attached 
> 
> Regards, 
> Noel 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> Sent: Tuesday, 26 May, 2009 15:52:24 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: StackOverflowError 
> 
> That's fine - or you could just change the sensitive parts - e.g. use 
> different URLs etc. 
> 
> On 26/05/2009, Noel O'Brien  wrote: 
> > Hi, 
> > 
> > I've generated the HTML schema file. Can I send it to you personally, i.e. 
> > I don't want to post it publically for security reasons 
> > 
> > 
> > Regards, 
> > Noel 
> > 
> > - Original Message - 
> > From: "sebb"  
> > To: "JMeter Users List"  
> > 
> > Sent: Tuesday, 26 May, 2009 14:56:18 GMT +00:00 GMT Britain, Ireland, 
> > Portugal 
> > Subject: Re: StackOverflowError 
> > 
> > OK, that could be added later. 
> > 
> > Meanwhile, perhaps you can use the extras/schematic.* files to 
> > generate a summary of the test plan that fails and attach that. 
> > [There's no schematic.sh version, but it should be obvious what to do 
> > if you are not using Windows]. 
> > 
> > On 26/05/2009, Noel O'Brien  wrote: 
> > > Hi, 
> > > 
> > > I find it hard to reproduce with a simple test plan, but I'll try to put 
> > > something together that will reproduce the issue 
> > > 
> > > Regards, 
> > > Noel 
> > > 
> > > - Original Message - 
> > > From: "sebb"  
> > > To: "JMeter Users List"  
> > > Sent: Tuesday, 26 May, 2009 12:21:38 GMT +00:00 GMT Britain, Ireland, 
> > > Portugal 
> > > Subject: Re: Stac 
> > > 
> > > On 25/05/2009, Noel O'Brien  wrote: 
> > > > Hi All, 
> > > > 
> > > > I'm using JMeter (2.3.2) to simulate a traffic model for our 
> > > > application (server) based on the traffic that our client generates. I 
> > > > used the JMeter proxy to sniff the traffic and then to make it more 
> > > > realistic I re-organised the captured samples into a series or Random 
> > > > and Throughput controllers. This works fine until the point in the test 
> > > > plan where the random/throuput controller come into play nad the the 
> > > > following happens, halting the test: 
> > > > 
> > > > 2009/05/25 16:07:54 ERROR - jmeter.threads.JMeterThread: Test failed! 
> > > > java.lang.StackOverflowError 
> > > > at java.util.HashMap.getEntry(HashMap.java:344) 
> > > > at java.util.LinkedHashMap.get(LinkedHashMap.java:280) 
> > > > at java.util.Collections$SynchronizedMap.get(Collections.java:1975) 
> > > > at 
> > > > org.apache.jmeter.testelement.AbstractTestElement.getProperty(AbstractTestElement.java:140)
> > > >  
> > > > at 
> > > > org.apache.jmeter.testelement.AbstractTestElement.getPropertyAsInt(AbstractTestElement.java:183)
> > > >  
> > > > at 
> > > > org.apache.jmeter.control.ThroughputController.getStyle(ThroughputController.java:109)
> > > >  
> > > > at 
> > > > org.apache.jmeter.control.ThroughputController.isDone(ThroughputController.java:207)
> > > >  
> > > > at 
> > > > org.apache.jmeter.control.InterleaveControl.currentReturnedNull(InterleaveControl.java:153)
> > > >  
> > > > at 
> > > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:94)
> > > >  
> > > > at 
> > > > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> >

Re: StackOverflowError

[Noel O'Brien]

Ok, I've replaced it all, see attached 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 15:52:24 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 

That's fine - or you could just change the sensitive parts - e.g. use 
different URLs etc. 

On 26/05/2009, Noel O'Brien  wrote: 
> Hi, 
> 
> I've generated the HTML schema file. Can I send it to you personally, i.e. I 
> don't want to post it publically for security reasons 
> 
> 
> Regards, 
> Noel 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> 
> Sent: Tuesday, 26 May, 2009 14:56:18 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: StackOverflowError 
> 
> OK, that could be added later. 
> 
> Meanwhile, perhaps you can use the extras/schematic.* files to 
> generate a summary of the test plan that fails and attach that. 
> [There's no schematic.sh version, but it should be obvious what to do 
> if you are not using Windows]. 
> 
> On 26/05/2009, Noel O'Brien  wrote: 
> > Hi, 
> > 
> > I find it hard to reproduce with a simple test plan, but I'll try to put 
> > something together that will reproduce the issue 
> > 
> > Regards, 
> > Noel 
> > 
> > - Original Message - 
> > From: "sebb"  
> > To: "JMeter Users List"  
> > Sent: Tuesday, 26 May, 2009 12:21:38 GMT +00:00 GMT Britain, Ireland, 
> > Portugal 
> > Subject: Re: Stac 
> > 
> > On 25/05/2009, Noel O'Brien  wrote: 
> > > Hi All, 
> > > 
> > > I'm using JMeter (2.3.2) to simulate a traffic model for our application 
> > > (server) based on the traffic that our client generates. I used the 
> > > JMeter proxy to sniff the traffic and then to make it more realistic I 
> > > re-organised the captured samples into a series or Random and Throughput 
> > > controllers. This works fine until the point in the test plan where the 
> > > random/throuput controller come into play nad the the following happens, 
> > > halting the test: 
> > > 
> > > 2009/05/25 16:07:54 ERROR - jmeter.threads.JMeterThread: Test failed! 
> > > java.lang.StackOverflowError 
> > > at java.util.HashMap.getEntry(HashMap.java:344) 
> > > at java.util.LinkedHashMap.get(LinkedHashMap.java:280) 
> > > at java.util.Collections$SynchronizedMap.get(Collections.java:1975) 
> > > at 
> > > org.apache.jmeter.testelement.AbstractTestElement.getProperty(AbstractTestElement.java:140)
> > >  
> > > at 
> > > org.apache.jmeter.testelement.AbstractTestElement.getPropertyAsInt(AbstractTestElement.java:183)
> > >  
> > > at 
> > > org.apache.jmeter.control.ThroughputController.getStyle(ThroughputController.java:109)
> > >  
> > > at 
> > > org.apache.jmeter.control.ThroughputController.isDone(ThroughputController.java:207)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.currentReturnedNull(InterleaveControl.java:153)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:94)
> > >  
> > > at 
> > > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> > >  
> > > at 
> > > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> > >  
> > > at 
> > > org.apache.jmeter.control.GenericController.next(GenericController.java:146)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.next(InterleaveControl.java:83)
> > >  
> > > at 
> > > org.apache.jmeter.control.InterleaveControl.nextIsAController(InterleaveControl.java:95)
> > >  
> > > 
> > > 
> > > Has anyone seen this before? If so what's causing it and is thete a way 
> > > to work around it? I've tries the test plan with 2.3.3 also, bu

Re: StackOverflowError

[Noel O'Brien]

Yes, but won't that execute the same requests each iteration (just like the 
proxy)? What I was trying to achieve was to simulate the randomness of choosing 
different paths through the client application, which leads to realistic groups 
of API calls being made to the server 

Either way I still had to re-organise the requests into Random/Throughput 
Controllers to simulating random branching with weighting. 

Regards, 
Noel 

- Original Message - 
From: "Peter Lin"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 17:12:37 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 

If you want to simulate "real production traffic", one option is to 
use the access log sampler. 

I wrote that sampler so that I could simulate production traffic. It 
might be easier than using a test plan that generates random requests. 

peter 

On Tue, May 26, 2009 at 12:08 PM, Noel O'Brien  wrote: 
> Done: 
> https://issues.apache.org/bugzilla/show_bug.cgi?id=47269 
> 
> Regards, 
> Noel 
> 

- 
To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 



-- 
Regards, 
Noel 

Re: StackOverflowError

[Noel O'Brien]

I've added the details to the ticket in Bugzilla, but here it is for 
convenience: 

Notes about the test plan schema: 

1. All Random Controllers have "Ignore sub-controller blocks" unchecked 
2. All Throughput Controllers are "Percent Executions", with siblings adding to 
100%. e.g. 4 sibling Throughput Controllers could have "Throughput" set to 60, 
25, 10, 5 respectively. 

>From some simple tests I had run previously, haveing (for example) 4 
>Throughput Controllers without a Random Controller as a parent would execute 
>Throughput Controllers in sequence. That's why each set of Throughput 
>Controllers has a Random Controler as a parent. 

I found that it was convenient just to reference out to the captured traffic 
using Module Controllers, but if you think this adds too much complexity to the 
test plan, I will try refactor it. 

Regards, 
Noel 


- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 17:37:34 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 

Thanks! 

That looks far too complicated to use as a test case, however it's 
still useful to have the information in case simpler test cases don't 
show the problem. 

Unfortunately the schematic does not show the "Execution Style" for 
the Throughput Controller - which do you use? Also, the Random 
Controller "Ignore sub-controller blocks" setting is not shown - which 
did you use? 

On 26/05/2009, Noel O'Brien  wrote: 
> Done: 
> https://issues.apache.org/bugzilla/show_bug.cgi?id=47269 
> 
> 
> Regards, 
> Noel 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> 
> Sent: Tuesday, 26 May, 2009 16:19:36 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: StackOverflowError 
> 
> Create the Bugzilla issue, re-open and add the attachment. 
> 
> On 26/05/2009, Noel O'Brien  wrote: 
> > Hmm, doesn't look like attachyments are supported in this mailing list. Is 
> > there anywhere else I can put it? 
> > 
> > Regards, 
> > Noel 
> > 
> > 
> > - Original Message - 
> > From: "Noel O'Brien"  
> > To: "JMeter Users List"  
> > 
> > Sent: Tuesday, 26 May, 2009 16:12:16 GMT +00:00 GMT Britain, Ireland, 
> > Portugal 
> > Subject: Re: StackOverflowError 
> > 
> > 
> > Ok, I've replaced it all, see attached 
> > 
> > Regards, 
> > Noel 
> > 
> > - Original Message - 
> > From: "sebb"  
> > To: "JMeter Users List"  
> > Sent: Tuesday, 26 May, 2009 15:52:24 GMT +00:00 GMT Britain, Ireland, 
> > Portugal 
> > Subject: Re: StackOverflowError 
> > 
> > That's fine - or you could just change the sensitive parts - e.g. use 
> > different URLs etc. 
> > 
> > On 26/05/2009, Noel O'Brien  wrote: 
> > > Hi, 
> > > 
> > > I've generated the HTML schema file. Can I send it to you personally, 
> > > i.e. I don't want to post it publically for security reasons 
> > > 
> > > 
> > > Regards, 
> > > Noel 
> > > 
> > > - Original Message - 
> > > From: "sebb"  
> > > To: "JMeter Users List"  
> > > 
> > > Sent: Tuesday, 26 May, 2009 14:56:18 GMT +00:00 GMT Britain, Ireland, 
> > > Portugal 
> > > Subject: Re: StackOverflowError 
> > > 
> > > OK, that could be added later. 
> > > 
> > > Meanwhile, perhaps you can use the extras/schematic.* files to 
> > > generate a summary of the test plan that fails and attach that. 
> > > [There's no schematic.sh version, but it should be obvious what to do 
> > > if you are not using Windows]. 
> > > 
> > > On 26/05/2009, Noel O'Brien  wrote: 
> > > > Hi, 
> > > > 
> > > > I find it hard to reproduce with a simple test plan, but I'll try to 
> > > > put something together that will reproduce the issue 
> > > > 
> > > > Regards, 
> > > > Noel 
> > > > 
> > > > - Original Message - 
> > > > From: "sebb"  
> > > > To: "JMeter Users List"  
> > > > Sent: Tuesday, 26 May, 2009 12:21:38 GMT +00:00 GMT Britain, Ireland, 
> > > > Portugal 
> > > > Subject: Re: Stac 
> > > > 
> > > > On 25/05/2009, Noel O'Brien  wrote: 
> > > > > Hi All, 
> > > > > 
> > &

Re: StackOverflowError

[Noel O'Brien]

I can reproduce the issue using a Switch Controller. I've posted a simple 
reproducible test plan to the issue: 

https://issues.apache.org/bugzilla/show_bug.cgi?id=47269 

It seems to cause the stack overflow when there's some dodgy BeanShell 
evaluation, i.e. it works fine if I escape the commas in the ${__BeanShell()} 

Regards, 
Noel 

- Original Message - 
From: "sebb"  
To: "JMeter Users List"  
Sent: Tuesday, 26 May, 2009 17:59:42 GMT +00:00 GMT Britain, Ireland, Portugal 
Subject: Re: StackOverflowError 

On 26/05/2009, Noel O'Brien  wrote: 
> I've added the details to the ticket in Bugzilla, but here it is for 
> convenience: 
> 
> Notes about the test plan schema: 
> 
> 1. All Random Controllers have "Ignore sub-controller blocks" unchecked 
> 2. All Throughput Controllers are "Percent Executions", with siblings adding 
> to 100%. e.g. 4 sibling Throughput Controllers could have "Throughput" set to 
> 60, 25, 10, 5 respectively. 

In that case, using the Switch Controller with a suitably chosen value 
might solve the crash. 

> From some simple tests I had run previously, haveing (for example) 4 
> Throughput Controllers without a Random Controller as a parent would execute 
> Throughput Controllers in sequence. That's why each set of Throughput 
> Controllers has a Random Controler as a parent. 
> 
> I found that it was convenient just to reference out to the captured traffic 
> using Module Controllers, but if you think this adds too much complexity to 
> the test plan, I will try refactor it. 

No, that's probably not causing the problem. It's most likely the deep 
nesting of Random and Throughput Controllers. 

It would obviously be useful to know where the failure occurs. 

> Regards, 
> 
> Noel 
> 
> 
> 
> - Original Message - 
> From: "sebb"  
> To: "JMeter Users List"  
> 
> Sent: Tuesday, 26 May, 2009 17:37:34 GMT +00:00 GMT Britain, Ireland, 
> Portugal 
> Subject: Re: StackOverflowError 
> 
> Thanks! 
> 
> That looks far too complicated to use as a test case, however it's 
> still useful to have the information in case simpler test cases don't 
> show the problem. 
> 
> Unfortunately the schematic does not show the "Execution Style" for 
> the Throughput Controller - which do you use? Also, the Random 
> Controller "Ignore sub-controller blocks" setting is not shown - which 
> did you use? 
> 
> On 26/05/2009, Noel O'Brien  wrote: 
> > Done: 
> > https://issues.apache.org/bugzilla/show_bug.cgi?id=47269 
> > 
> > 
> > Regards, 
> > Noel 
> > 
> > - Original Message - 
> > From: "sebb"  
> > To: "JMeter Users List"  
> > 
> > Sent: Tuesday, 26 May, 2009 16:19:36 GMT +00:00 GMT Britain, Ireland, 
> > Portugal 
> > Subject: Re: StackOverflowError 
> > 
> > Create the Bugzilla issue, re-open and add the attachment. 
> > 
> > On 26/05/2009, Noel O'Brien  wrote: 
> > > Hmm, doesn't look like attachyments are supported in this mailing list. 
> > > Is there anywhere else I can put it? 
> > > 
> > > Regards, 
> > > Noel 
> > > 
> > > 
> > > - Original Message - 
> > > From: "Noel O'Brien"  
> > > To: "JMeter Users List"  
> > > 
> > > Sent: Tuesday, 26 May, 2009 16:12:16 GMT +00:00 GMT Britain, Ireland, 
> > > Portugal 
> > > Subject: Re: StackOverflowError 
> > > 
> > > 
> > > Ok, I've replaced it all, see attached 
> > > 
> > > Regards, 
> > > Noel 
> > > 
> > > - Original Message - 
> > > From: "sebb"  
> > > To: "JMeter Users List"  
> > > Sent: Tuesday, 26 May, 2009 15:52:24 GMT +00:00 GMT Britain, Ireland, 
> > > Portugal 
> > > Subject: Re: StackOverflowError 
> > > 
> > > That's fine - or you could just change the sensitive parts - e.g. use 
> > > different URLs etc. 
> > > 
> > > On 26/05/2009, Noel O'Brien  wrote: 
> > > > Hi, 
> > > > 
> > > > I've generated the HTML schema file. Can I send it to you personally, 
> > > > i.e. I don't want to post it publically for security reasons 
> > > > 
> > > > 
> > > > Regards, 
> > > > Noel 
> > > > 
> > > > - Original Message - 
> > > > From: "sebb"  
> > > > To: "JMeter Users List&

Re: Password field in Swing application not recoginized

[Noel O'Brien]

JMeter cannot test (or record) Swing based applications. It can record HTTP 
traffic and can test a multitude if services (FTP, Mail, JMS, JDBC, HTTP, etc.) 

Regards, 
Noel 

- "nawaz_farhan"  wrote: 
> 
> I'm recording a Swing based application, a pop-window with username and 
> password field , problem is with password field (JPasswordField) not getting 
> reocoginized. 
> Expecting a quick reply 
> -- 
> View this message in context: 
> http://www.nabble.com/Password-field-in-Swing-application-not-recoginized-tp23865312p23865312.html
>  
> Sent from the JMeter - User mailing list archive at Nabble.com. 
> 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 

-- 
Regards, 
Noel 

Re: expected Response code: 500

[Noel O'Brien]

In the Response Assertion check the "ignore response" box and add an assertion 
for response code equals 500 

Regards, 
Noel 

- "mattlamignat"  wrote: 
> 
> I'm getting 'Response code: 500' responces for several SOAP/XML-RPC requests 
> and they are expected responses. 
> How I can force JMeter to not mark such responces as failures. I mean marked 
> as red transactions in 'View Results Tree' 
> -- 
> View this message in context: 
> http://www.nabble.com/expected-Response-code%3A-500-tp24040330p24040330.html 
> Sent from the JMeter - User mailing list archive at Nabble.com. 
> 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 

-- 
Regards, 
Noel 

Re: Does JMeter support URL list?

[Noel O'Brien]

You could use a CSV data config element with a single sampler to achieve this 
:) It would probably be less intensive on resources too. 

Otherwise you could write a small program to construct the .jmx file, creating 
a sampler for each URL 

Regards, 
Noel 

- "Zheng Jian-Ming"  wrote: 
> Hi, 
> 
> I want to construct a web test plan which accesses a lot of web pages, for 
> example 1000 pages. 
> 
> Is it the only way to add 1000 "HTTP Request Sampler" to the Thread Group 
> element? Does JMeter supports a file of URL list? 
> 
> Thank you. 
> 
> 
> 
> - 
> To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org 
> For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org 
> 
> 

-- 
Regards, 
Noel