jmeter-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From santiagoloso <santiagol...@gmail.com>
Subject Re: Is this working or not?
Date Mon, 13 Apr 2009 13:45:36 GMT

That's exactly what I've been doing (in fact, I dunno how to generate all the
requests by myself). The problem is that there are too many requests
(especially image requests) and I'm not sure if those requests are really
important when I'm load testing.

Michael McDonnell wrote:
> 
> You can pre-record your test script and then afterwards you can swap out
> the
> posted data from your forms with variables like username and password.
> 
> -Michael P.
> 
> On Mon, Apr 13, 2009 at 7:19 AM, santiagoloso
> <santiagoloso@gmail.com>wrote:
> 
>>
>> I'm not sure I'm understanding well what you're proposing: the Once Only
>> Controller is used in order to log in only once that username? How does
>> it
>> works?
>>
>> In another subject, I'm willing to know how useful are all the HTTP
>> request
>> self-generated, I mean, how important are those requests recorded by the
>> HTTP proxy server that are not a consequence of user action. I'm asking
>> this, because when I executed my test plan, I'd had 17k hits on the
>> webserver on two hours, because each request on the test plan is
>> considered
>> as a hit, but when manually executing the same actions hits weren't more
>> than 15 (test plan has over 60 requests, like images and stuff, and I
>> don't
>> know if those requests are really that important).
>>
>>
>> Michael McDonnell wrote:
>> >
>> > I do the same thing in my testing for my company
>> >
>> >
>> > Here's my set up:
>> >
>> > Test Plan
>> >     -Thread Group
>> >         -HTTP Cookie Manager
>> >         -CSV Configuration
>> >             Defines USERNAME,PASSWORD
>> >         -HTTP Request Defaults
>> >         -Once Only Controller
>> >             -Http Request to Login.jsp
>> >                 Passes ${USERNAME} and ${PASSWORD} through POST
>> >         -Do other things after I'm logged in in a loop.
>> >
>> > Hopefully this helps you.
>> >
>> > On Thu, Apr 9, 2009 at 8:37 AM, santiagoloso <santiagoloso@gmail.com>
>> > wrote:
>> >
>> >>
>> >> I've resolved this problem by using the Login Configuration Element.
>> >> Now I have a new problem and it's that the application I'm testing
>> does
>> >> not
>> >> allow the same user to log in simultaneously more than once.
>> >>
>> >> So, when I select 1 thread group and infinite loop there's no problem.
>> >> When
>> >> I select n users and 1 loop, using n different usernames, there's no
>> >> problem, but when I select n thread groups, n loops and I have n
>> >> usernames
>> >> (using csv data config element, that provides usernames to the Login
>> >> Config
>> >> Element), some of the samples (the most significative) returns a
>> negative
>> >> response code (500).
>> >>
>> >> I've thought about creating separate Thread Groups, using a different
>> >> username for each Thread Group. But there are some concepts I'm not
>> quite
>> >> sure about: when I'm remote testing, using 5 different computers, each
>> >> Thread Group will be repeated on each machine, so, I'll be in the same
>> >> place
>> >> again (since I'll have the same username on 5 different machines, and
>> >> response code will be 500 again).
>> >>
>> >> I've thought about this once more and may be, if I use csv files to
>> >> provide
>> >> usernames to each thread group (different file for different Thread
>> >> Groups),
>> >> since each csv file will be local on each machine, then all the users
>> can
>> >> be
>> >> different.
>> >>
>> >> Any better idea?
>> >>
>> >> Thanks a lot for your help,
>> >>
>> >> Santiago
>> >>
>> >> Michael McDonnell wrote:
>> >> >
>> >> > Are you checking the actual response using the tree to see if your
>> page
>> >> > reflect the log in correctly?
>> >> >
>> >> > You need to run in GUI mode to check that.
>> >> >
>> >> > On Wed, Apr 8, 2009 at 12:34 PM, santiagoloso
>> >> > <santiagoloso@gmail.com>wrote:
>> >> >
>> >> >>
>> >> >> Yeap, I'm passing them. Results for ALL the samples are positive
>> (on
>> >> the
>> >> >> whole test plan). But the logs on the DB says anonymous instead.
>> >> >>
>> >> >>
>> >> >> Michael McDonnell wrote:
>> >> >> >
>> >> >> > Are you passing in the username/password variables when you
POST
>> the
>> >> >> login
>> >> >> > form?
>> >> >> >
>> >> >> > On Wed, Apr 8, 2009 at 12:14 PM, santiagoloso
>> >> >> > <santiagoloso@gmail.com>wrote:
>> >> >> >
>> >> >> >>
>> >> >> >> Hi everyone, I'm working on Test Plan for an application
that
>> >> requires
>> >> >> >> username and password. Due to some permissions thing only
some
>> of
>> >> the
>> >> >> >> users
>> >> >> >> loaded on the DB are allowed to use this feature I'll
be
>> testing.
>> >> >> >>
>> >> >> >> Said that, I've configured a CSV file with usernames and
>> passwords
>> >> >> (all
>> >> >> >> allowed to use this feature I'm testing) and setted the
thread
>> >> group
>> >> >> to
>> >> >> >> log
>> >> >> >> in to the website, and then perform the required steps
to get
>> into
>> >> the
>> >> >> >> feature and operate over it.
>> >> >> >>
>> >> >> >> After executing THOUSANDS of times, and checked results,
when I
>> ask
>> >> >> for
>> >> >> >> information, the required information is retrieved (analyzed
>> using
>> >> >> View
>> >> >> >> Results in Tree Listener). It's not really important data
>> veracity,
>> >> so
>> >> >> >> I'm
>> >> >> >> not using assertions, just need the performance.
>> >> >> >>
>> >> >> >> Well, my doubt comes this way: when taking a look at db
logs,
>> >> servers
>> >> >> >> been
>> >> >> >> hitted a lot of times, but always by an anonymous user
(not by
>> the
>> >> >> >> usernames
>> >> >> >> and passwords that are sent within the test plan).
>> >> >> >>
>> >> >> >> My deduction power tells me that the first steps (those
executed
>> >> >> before
>> >> >> >> the
>> >> >> >> log in) must be anonymous, but, from that step and all
that
>> follow,
>> >> >> >> username
>> >> >> >> on db logs should be the one inserted on each iteration
(csv
>> file
>> >> >> brings
>> >> >> >> multiple username/password combinations, all valid).
>> >> >> >>
>> >> >> >> Am I wrong? What am I missing?
>> >> >> >> --
>> >> >> >> View this message in context:
>> >> >> >>
>> >> http://www.nabble.com/Is-this-working-or-not--tp22956547p22956547.html
>> >> >> >> Sent from the JMeter - User mailing list archive at Nabble.com.
>> >> >> >>
>> >> >> >>
>> >> >> >>
>> >> ---------------------------------------------------------------------
>> >> >> >> To unsubscribe, e-mail:
>> jmeter-user-unsubscribe@jakarta.apache.org
>> >> >> >> For additional commands, e-mail:
>> >> jmeter-user-help@jakarta.apache.org
>> >> >> >>
>> >> >> >>
>> >> >> >
>> >> >> >
>> >> >>
>> >> >> --
>> >> >> View this message in context:
>> >> >>
>> http://www.nabble.com/Is-this-working-or-not--tp22956547p22956898.html
>> >> >> Sent from the JMeter - User mailing list archive at Nabble.com.
>> >> >>
>> >> >>
>> >> >>
>> ---------------------------------------------------------------------
>> >> >> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
>> >> >> For additional commands, e-mail:
>> jmeter-user-help@jakarta.apache.org
>> >> >>
>> >> >>
>> >> >
>> >> >
>> >>
>> >> --
>> >> View this message in context:
>> >> http://www.nabble.com/Is-this-working-or-not--tp22956547p22972288.html
>> >> Sent from the JMeter - User mailing list archive at Nabble.com.
>> >>
>> >>
>> >> ---------------------------------------------------------------------
>> >> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
>> >> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>> >>
>> >>
>> >
>> >
>>
>> --
>> View this message in context:
>> http://www.nabble.com/Is-this-working-or-not--tp22956547p23021624.html
>> Sent from the JMeter - User mailing list archive at Nabble.com.
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
>> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>>
>>
> 
> 

-- 
View this message in context: http://www.nabble.com/Is-this-working-or-not--tp22956547p23022559.html
Sent from the JMeter - User mailing list archive at Nabble.com.


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


Mime
View raw message