jmeter-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Deepak Shetty <shet...@gmail.com>
Subject Re: Questions on Capacity and Limitations of JMeter
Date Wed, 10 Mar 2010 03:22:16 GMT
A lot of your questions are environment specific. JMeter can be run in
client-slave or can be run separately on different machines and you can then
aggregate the results, so there is no theoretical maximum for separate
instances. Master Slave probably has a maximum but should be more than 1000
concurrent requests.

>Can a good laptop submit 1,000 HTTP requests in a short period (< 1 minute)
using JMeter
Yes but you havent specified concurrent(so it could be 100 concurrent * 10
requests each) .
If you actually mean 1000 concurrent requests then Normal laptops probably
cant run 1000 threads/ sockets concurrently and give you any valid results.
You probably would need about 4 laptops. Desktops perform better .

>Can we test long lasting HTTP session with multiple request/response (e.g.,
>Login, AddToShopingCart, CheckOut, SubmitPayment, Confirmation, ...) with
>JMeter automatically?
Yes. But what do you mean by automatically. For most non trivial
applications , record /replay is overrated because the replay wont work
because of dynamic data. You almost always need to tweak the script.

>What if Javascript and AJAX are used to process session data?
Jmeter isnt a browser and doesnt execute javascript. However any interaction
with the server (using AJAX or otherwise is still a HTTP request) can be
simulated by JMeter. So AJAX yes , javascript no. Im not sure what you mean
by javascript being used to process session data.

>Can we use multiple laptops at different locations (different IPs) to
concurrently load test the same website with JMeter?
Yes. As before you can use master-slaves or you run them separately at the
same time and just aggregate the results.

>Anything that commercial load test tools  (e.g., LoadRunner) can do, but
JMeter?
Out of the box Reports in Load Runner are better. I believe Load Runner also
does some server monitoring and can correlate the test run to these values
(You would need separate server monitoring tools for Jmeter). Its also
easier to do some sort of increasing load scenarios on load runner than it
is in jmeter .



regards
deepak


On Tue, Mar 9, 2010 at 6:51 PM, JIM B. <jzbao@yahoo.com> wrote:

>
> I'm new to JMeter and in the process to select a load testing tool.
>
> I got a requirement to load test a website running in a cluster of two
> Window 2008 servers.
> The website was built upon Microsoft technologies.
>
> The load test tool must be able to generate 1,000 concurrent users. Can
> JMeter do it?
>
> Can a good laptop submit 1,000 HTTP requests in a short period (< 1 minute)
> using JMeter? What is the max HTTP requests can be submitted within one
> second with JMeter?
>
> Is there any limitations on concurrent HTTP sessions on JMeter?
>
> Can we test long lasting HTTP session with multiple request/response (e.g.,
> Login, AddToShopingCart, CheckOut, SubmitPayment, Confirmation, ...) with
> JMeter automatically?
> What if Javascript and AJAX are used to process session data?
>
> Is there any bottleneck due to the client computer where JMeter is running?
>
> Can we use multiple laptops at different locations (different IPs) to
> concurrently load test
> the same website with JMeter?
>
> Any suggestions on client computers?
>
> Anything that commercial load test tools  (e.g., LoadRunner) can do, but
> JMeter?
> --
> View this message in context:
> http://old.nabble.com/Questions-on-Capacity-and-Limitations-of-JMeter-tp27844939p27844939.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
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message