constant throughput timer and response time

[Toni Menendez Lopez]

<jmeter-***@jakarta.apache.org>Hello everybody,

Is there any way I can check relation between constant throughput timer and
response time in requests.
The thing is that in my application has a response time of 80mseg and
jmeter is not able to follow the throghput I set up in the constant
throughput timer !

Can anyone explain me why ?

[Andrei Ghimus]

Toni,

The Constant Throughput Timer will NOT affect the response time of
your requests.
It will, however, introduce a variable delay between samples, in order
to limit your Actions Per Minute to the specified value.


On Mon, Nov 8, 2010 at 10:21 PM, Toni Menendez Lopez

Post by Toni Menendez Lopez
Is there any way I can check relation between constant throughput timer and
response time in requests.
The thing is that in my application has a response time of 80mseg and
jmeter is not able to follow the throghput I set up in the constant
throughput timer !
Can anyone explain me why ?

[Toni Menendez Lopez]

Look Andrei,

I have following scenario : Figure1.png and Figure2.png

With to scenarios 1st ) with average response time of 9 mseg 2nd ) with
average of 100mseg

The 1st case : Figure3.png, I am able to manage the 100reqxsec as is
specified in the constant throughput timer, but 2nd case ) Figure4.png I am
only able to send 10 reqxsec. The only difference in the scenario is the
response time.

Do you find any explanation ?

Toni.

Post by Andrei Ghimus
Toni,
The Constant Throughput Timer will NOT affect the response time of
your requests.
It will, however, introduce a variable delay between samples, in order
to limit your Actions Per Minute to the specified value.
On Mon, Nov 8, 2010 at 10:21 PM, Toni Menendez Lopez

Post by Toni Menendez Lopez
Is there any way I can check relation between constant throughput timer

and

Post by Toni Menendez Lopez
response time in requests.
The thing is that in my application has a response time of 80mseg and
jmeter is not able to follow the throghput I set up in the constant
throughput timer !
Can anyone explain me why ?

---------------------------------------------------------------------

[Felix Frank]

Post by Toni Menendez Lopez
Look Andrei,
I have following scenario : Figure1.png and Figure2.png
With to scenarios 1st ) with average response time of 9 mseg 2nd ) with
average of 100mseg
The 1st case : Figure3.png, I am able to manage the 100reqxsec as is
specified in the constant throughput timer, but 2nd case ) Figure4.png I
am only able to send 10 reqxsec. The only difference in the scenario is
the response time.
Do you find any explanation ?

You run with a single Thread, right?

Let's do the math:

Max throughput with 0.009sec/access:
(1 second) / (0.009 seconds/access ) ~ 100 accesses.

Max throughput with 0.1sec/access:
(1 second) / (0.1 seconds/access ) = 10 accesses

Use at least 10 Threads to achieve 100 req/s for both transactions.

HTH,
Felix

[Deepak Goel]

Hey

Namaskara~Nalama~Guten Tag

Throughput = Number of Users or Threads in Jmeter's case

--------------------------------------------------------
Think Time + Response Time

So if your response time varies as in the second case for the same number of
threads (or users), your throughput will vary (request/sec). More the
response time, less the throughput and vice versa.

Deepak

--
Keigu

Deepak
[phone]
***@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
"

Post by Felix Frank

Post by Toni Menendez Lopez
Look Andrei,
I have following scenario : Figure1.png and Figure2.png
With to scenarios 1st ) with average response time of 9 mseg 2nd ) with
average of 100mseg
The 1st case : Figure3.png, I am able to manage the 100reqxsec as is
specified in the constant throughput timer, but 2nd case ) Figure4.png I
am only able to send 10 reqxsec. The only difference in the scenario is
the response time.
Do you find any explanation ?

You run with a single Thread, right?
(1 second) / (0.009 seconds/access ) ~ 100 accesses.
(1 second) / (0.1 seconds/access ) = 10 accesses
Use at least 10 Threads to achieve 100 req/s for both transactions.
HTH,
Felix
---------------------------------------------------------------------

[Toni Menendez Lopez]

Thanks a lot,

Toni.

Post by Deepak Goel
Hey
Namaskara~Nalama~Guten Tag
Throughput = Number of Users or Threads in Jmeter's case
--------------------------------------------------------
Think Time + Response Time
So if your response time varies as in the second case for the same number of
threads (or users), your throughput will vary (request/sec). More the
response time, less the throughput and vice versa.
Deepak
--
Keigu
Deepak
[phone]
http://www.simtree.net
Skype: thumsupdeicool
Google talk: deicool
Blog: http://loveandfearless.wordpress.com
Facebook: http://www.facebook.com/deicool
http://www.gridrepublic.org
"

Post by Felix Frank

Post by Toni Menendez Lopez
Look Andrei,
I have following scenario : Figure1.png and Figure2.png
With to scenarios 1st ) with average response time of 9 mseg 2nd ) with
average of 100mseg
The 1st case : Figure3.png, I am able to manage the 100reqxsec as is
specified in the constant throughput timer, but 2nd case ) Figure4.png I
am only able to send 10 reqxsec. The only difference in the scenario is
the response time.
Do you find any explanation ?

You run with a single Thread, right?
(1 second) / (0.009 seconds/access ) ~ 100 accesses.
(1 second) / (0.1 seconds/access ) = 10 accesses
Use at least 10 Threads to achieve 100 req/s for both transactions.
HTH,
Felix
---------------------------------------------------------------------