You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by mjwillis <wi...@gmail.com> on 2010/08/31 18:36:27 UTC
Simulated users?
Hello Everyone,
I am a new JMeter user and am having difficulty setting up my tests to
accurately simulate users. I ran a test with 50 simulated users and it
completely hammers the server (which didnt seem right to me). So then I
tried with one simulated user and it appears to hammering the cache at an
unrealistic rate. Almost as if the single simulated user is looping through
the test at an incredible rate that a normal user couldnt do. Its worth
noting that the test is being run behind a few XServes as the actual test
would be 1000 simulated users (once I figure this out).
You must log into the site using username/password combination. After
logging in, your session is stored in a cookie. I used the cookie manager
for that piece and recorded my test using JMeters proxy feature. I didnt do
anything fancy when recording the test, just clicking on different pages and
waiting for them to load. With that said, here is how I have the test
setup:
-Thread (with the # of simulated users set inside it)
--Cookie Manager
--Recorded test plan
-Listener 1
-Listener 2
-Listener 3
Is there a key controller or component I am missing? Or is this really what
1 user will do to the website? Once I figure out how to properly test with
one user I will try to scale this number up to the 1000 that is needed.
Thanks for any help you can give!
--
View this message in context: http://jmeter.512774.n5.nabble.com/Simulated-users-tp2798333p2798333.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
Re: Simulated users?
Posted by Yasir Shoaib <ya...@gmail.com>.
Hey Willis,
Try adding think time if you don't already have it. Also, if you have
lots of samplers then it would be causing queueing at the network so for
demo-testing just keep the samplers that are computation intensive.
Yasir
-----Original Message-----
From: mjwillis <wi...@gmail.com>
Reply-to: "JMeter Users List" <jm...@jakarta.apache.org>
To: jmeter-user@jakarta.apache.org
Subject: Simulated users?
Date: Tue, 31 Aug 2010 09:36:27 -0700 (PDT)
Hello Everyone,
I am a new JMeter user and am having difficulty setting up my tests to
accurately simulate users. I ran a test with 50 simulated users and it
completely hammers the server (which didnt seem right to me). So then I
tried with one simulated user and it appears to hammering the cache at an
unrealistic rate. Almost as if the single simulated user is looping through
the test at an incredible rate that a normal user couldnt do. Its worth
noting that the test is being run behind a few XServes as the actual test
would be 1000 simulated users (once I figure this out).
You must log into the site using username/password combination. After
logging in, your session is stored in a cookie. I used the cookie manager
for that piece and recorded my test using JMeters proxy feature. I didnt do
anything fancy when recording the test, just clicking on different pages and
waiting for them to load. With that said, here is how I have the test
setup:
-Thread (with the # of simulated users set inside it)
--Cookie Manager
--Recorded test plan
-Listener 1
-Listener 2
-Listener 3
Is there a key controller or component I am missing? Or is this really what
1 user will do to the website? Once I figure out how to properly test with
one user I will try to scale this number up to the 1000 that is needed.
Thanks for any help you can give!
Re: Simulated users?
Posted by kblearner <sa...@citi.com>.
Hey "mjwillis",
1. Definately use Think time as suggested by other users in this thread.
>>Apart from using constant timer, you could also Use Test Action
Sampler with a timer as it's child.
2. It's advisable to test your application/server with incremental user
load and not the otehr way round.
>>May be do 20% - 50% - 75% - 100% - 120%. These are w.r.t Max user
load
3. Also make sure that you'r using the appropriate controller types for
your Http samplers.
Cheers..
--
View this message in context: http://jmeter.512774.n5.nabble.com/Simulated-users-tp2798333p2798467.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
Re: Simulated users?
Posted by Deepak Goel <de...@gmail.com>.
Hey
Namaskara~Nalama~Guten Tag
Have you set the think time?
Deepak
--
Keigu
Deepak
+91-9765089593
deicool@gmail.com
http://www.simtree.net
Skype: thumsupdeicool
Google talk: deicool
Blog: http://loveandfearless.wordpress.com
Facebook: http://www.facebook.com/deicool
"Contribute to the world, environment and more : http://www.gridrepublic.org
"
On Tue, Aug 31, 2010 at 10:06 PM, mjwillis <wi...@gmail.com> wrote:
>
> Hello Everyone,
> I am a new JMeter user and am having difficulty setting up my tests to
> accurately simulate users. I ran a test with 50 simulated users and it
> completely hammers the server (which didnt seem right to me). So then I
> tried with one simulated user and it appears to hammering the cache at an
> unrealistic rate. Almost as if the single simulated user is looping
> through
> the test at an incredible rate that a normal user couldnt do. Its worth
> noting that the test is being run behind a few XServes as the actual test
> would be 1000 simulated users (once I figure this out).
>
> You must log into the site using username/password combination. After
> logging in, your session is stored in a cookie. I used the cookie manager
> for that piece and recorded my test using JMeters proxy feature. I didnt
> do
> anything fancy when recording the test, just clicking on different pages
> and
> waiting for them to load. With that said, here is how I have the test
> setup:
>
> -Thread (with the # of simulated users set inside it)
> --Cookie Manager
> --Recorded test plan
> -Listener 1
> -Listener 2
> -Listener 3
>
> Is there a key controller or component I am missing? Or is this really
> what
> 1 user will do to the website? Once I figure out how to properly test with
> one user I will try to scale this number up to the 1000 that is needed.
>
> Thanks for any help you can give!
> --
> View this message in context:
> http://jmeter.512774.n5.nabble.com/Simulated-users-tp2798333p2798333.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
>
>
Re: Simulated users?
Posted by Deepak Shetty <sh...@gmail.com>.
Hi
Assuming your test script is correct (i.e. you have extracted any dynamic
parameters and you can see all your content correctly in view Results Tree)
then a difference between a JMeter test v/s a user is "think time" - The
amount of time a user spends reading the page / typing stuff before actually
clicking the next link. (so active users i.e. ones who have a session on
your site is always far > users who are actually doing something on your
site concurrently). By default JMeter has 0 think time. If you need JMeter
to simulate this behavior then you must use one of the timers available.
http://jakarta.apache.org/jmeter/usermanual/component_reference.html
Again , if your script is wrong then you need to fix that first (for e.g.
make sure your requests dont have any session ids , make sure you dont have
any dynamic parameters)
regards
deepak
On Tue, Aug 31, 2010 at 9:36 AM, mjwillis <wi...@gmail.com> wrote:
>
> Hello Everyone,
> I am a new JMeter user and am having difficulty setting up my tests to
> accurately simulate users. I ran a test with 50 simulated users and it
> completely hammers the server (which didnt seem right to me). So then I
> tried with one simulated user and it appears to hammering the cache at an
> unrealistic rate. Almost as if the single simulated user is looping
> through
> the test at an incredible rate that a normal user couldnt do. Its worth
> noting that the test is being run behind a few XServes as the actual test
> would be 1000 simulated users (once I figure this out).
>
> You must log into the site using username/password combination. After
> logging in, your session is stored in a cookie. I used the cookie manager
> for that piece and recorded my test using JMeters proxy feature. I didnt
> do
> anything fancy when recording the test, just clicking on different pages
> and
> waiting for them to load. With that said, here is how I have the test
> setup:
>
> -Thread (with the # of simulated users set inside it)
> --Cookie Manager
> --Recorded test plan
> -Listener 1
> -Listener 2
> -Listener 3
>
> Is there a key controller or component I am missing? Or is this really
> what
> 1 user will do to the website? Once I figure out how to properly test with
> one user I will try to scale this number up to the 1000 that is needed.
>
> Thanks for any help you can give!
> --
> View this message in context:
> http://jmeter.512774.n5.nabble.com/Simulated-users-tp2798333p2798333.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
>
>