jmeter-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Simon De Uvarow <sduva...@ces.com.uy>
Subject Re: Real User Analysis
Date Tue, 03 Mar 2009 13:49:25 GMT
You should do it backwards, first you want to know how many users you want
to simulate. Then, what things do the different types of users in the
system, and finally "write the scripts" or compose the testplan.

Each type of user could be a thread group. You will have one thread for each
virtual user.
A thread (or virtual user) will do things over the system, HTTP requests,
JDBC requests, etc. And maybe a user could have loops and conditions.


ThreadGroup1-UsersThatJustSeeFirstPage
   RequestFirstPage

ThreadGroup2-UsersThatLogin&Logout
   RequestFirstPage
   GetNamePasswordFromCSVFile
   Login
   Logout
ThreadGroup3-UsersThatBuyOneThing
   .....
ThreadGroup4-UsersThatBuySeeStateOfCreditCard
   .....


How many users of each type you will have? you have to investigate that.
So, you are going to have as many threads as users.

If you want to know how many users you have in a instant in the test you
will have to analyze the logs of the test. (You have to log the results in a
file, and analyze the file)


this is a explanation in a very few words, there are much many other
problems to solve, you will need enough good data to do a good test, you
will need the hardware and to monitor all the environment, and other things
...


regards.
Simon


"No olvides, no traiciones, lo que llevas bien dentro de ti. No olvides, no
traiciones, lo que siempre te ha hecho vivir."


On Mon, Mar 2, 2009 at 9:07 PM, sebb <sebbaz@gmail.com> wrote:

> On 02/03/2009, hoffm11 <hoffm11@gmail.com> wrote:
> >
> >  I have a script that has
> >  20 threads
> >  100 seconds of Ramp up
> >  5000 Milliseconds Constant Timer
> >
> >  I have to tell a manager what this means in "real" users.  What is the
> >  calculation to do this?  what if I run 20 concurrent threads without any
> >  ramp up time or timers?
>
> Impossible to say without knowing what a "real" user would be expected
> to do with your application. For example, the usage patterns for
> Google and www.apache.org are very different.
>
> I suggest you look at the number and type of requests (method,
> protocol and response size) made a real user, and compare that with
> the test plan.
>
> >  Thanks
> >
> > --
> >  View this message in context:
> http://www.nabble.com/Real-User-Analysis-tp22296648p22296648.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
> >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message