jmeter-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Deepak Goel <deic...@gmail.com>
Subject Re: Load I can generate?
Date Tue, 16 May 2017 18:59:57 GMT
Sergio

How many total slave machines did you manage in your test bed? Or how many
total number of concurrent users did you test with?

Thanks
Deepak



Deepak
"Please stop cruelty to Animals, help by becoming a Vegan"
+91 73500 12833
deicool@gmail.com

LinkedIn: www.linkedin.com/in/deicool
https://www.facebook.com/deicool

"Contribute to the world, environment and more : http://www.gridrepublic.org
"

On Tue, May 16, 2017 at 9:23 PM, Sergio Boso <sergio@bosoconsulting.it>
wrote:

> Hi,
>
> IMHO, scaling depends on a lot of factors.
> Some of them are pretty difficult to predict in advance, but I will
> include:
> - size and complexity of the request and response (do you retrieve
> embedded objects? How many threads?)
>    Generally speaking,  REST interfaces are lighter than complex web pages
> - type and complexity of assertion applied on responses
> - complexity of the script: some samplers can be heavier than expected
> (e.g. OS sampler)
> - amount of data to use (CSV files)
>
> I suggest you to do some test with your script on a single slave
> configuration and check you mileage.
> That said, I managed a test bed where each slave machine simulated 2200
> users, with constant throughput timer at 30seconds.
> JVM had 4 G Bytes and 4 fast speed core (AWS infrastructure).
> There was even room for more load.
> Some key factors you have to keep under control are:
> - bandwidth toward the system-under-test
> - bandwidth toward the master (result data)
> - amount of RAM for the JVM and GC configuration.
>
> HTH
>
> Sergio
>
> Il 16/05/2017 15:56, Pravesh Prajapati ha scritto:
>
>> Hi Deepak.
>> One correction its 16 gb not 8 gb.. for ram.
>>
>> On Tuesday, May 16, 2017, Pravesh Prajapati <prajapati.pravesh@gmail.com>
>> wrote:
>>
>> Hi Sir,
>>> Please find the attachment for scaling environment in Jmeter.
>>>
>>> I have mentioned the things as follow.
>>> If we have 4-6 scenario per application with 10-15 samples,
>>> Then we need M\C having configuration of 8 GB RAM, I-3/I-5
>>> processor(2.5-3
>>> GHZ)  for executing approx 1800 VU on slave m\c.
>>>
>>>
>>>
>>>
>>>
>>> On Tue, May 16, 2017 at 9:55 PM, Pravesh Prajapati <
>>> prajapati.pravesh@gmail.com
>>> <javascript:_e(%7B%7D,'cvml','prajapati.pravesh@gmail.com');>> wrote:
>>>
>>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
>

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