Large scale Grinder test configuration

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Large scale Grinder test configuration

Tom Vitez

Hi Grinder Users!

 

I have been using Grinder successfully on a number of projects over the past few years. My past tests have been relatively small in terms of vUser or thread count (i.e. < 500) but with high transaction rates (TPS). I have a need to run a test in the order of 30-40,000 vUsers so I need to spread across many Grinder agent VM’s and tie them back to a single Grinder Console instance to consolidate the results.

 

My questions are:

 

Is are there any physical limits to the number of agents that can be configured to a single Grinder console?

 

What is the largest number of Grinder Agent’s you have configured successfully with a single Console instance?

 

Any recommendations on value for grinder.reportToConsole.interval to minimize network bandwidth without sacrificing performance metrics (don’t want to rob network bandwidth from the application under test)?

 

Kind Regards,

 

Tom


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Large scale Grinder test configuration

Darren Ball

Why not segregate your networks?  Do you have the option of multiple networks?  Have your agents drive tests on a private network, and your console and agent communication with the console on another?

On Mar 24, 2016 2:56 PM, "Tom Vitez" <[hidden email]> wrote:

Hi Grinder Users!

 

I have been using Grinder successfully on a number of projects over the past few years. My past tests have been relatively small in terms of vUser or thread count (i.e. < 500) but with high transaction rates (TPS). I have a need to run a test in the order of 30-40,000 vUsers so I need to spread across many Grinder agent VM’s and tie them back to a single Grinder Console instance to consolidate the results.

 

My questions are:

 

Is are there any physical limits to the number of agents that can be configured to a single Grinder console?

 

What is the largest number of Grinder Agent’s you have configured successfully with a single Console instance?

 

Any recommendations on value for grinder.reportToConsole.interval to minimize network bandwidth without sacrificing performance metrics (don’t want to rob network bandwidth from the application under test)?

 

Kind Regards,

 

Tom


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Large scale Grinder test configuration

Marc Holden
Hi Tom,

We have run over 40K VUs across 25 grinder agents successfully w/o adjusting reportToConsole interval.  

Depending on the size of the responses in the transactions, it is worth while to get some good network/systems monitoring in place and coordinate with the appropriate groups.

Thanks,

-Marc



On Thu, Mar 24, 2016 at 8:57 PM Darren Ball <[hidden email]> wrote:

Why not segregate your networks?  Do you have the option of multiple networks?  Have your agents drive tests on a private network, and your console and agent communication with the console on another?

On Mar 24, 2016 2:56 PM, "Tom Vitez" <[hidden email]> wrote:

Hi Grinder Users!

 

I have been using Grinder successfully on a number of projects over the past few years. My past tests have been relatively small in terms of vUser or thread count (i.e. < 500) but with high transaction rates (TPS). I have a need to run a test in the order of 30-40,000 vUsers so I need to spread across many Grinder agent VM’s and tie them back to a single Grinder Console instance to consolidate the results.

 

My questions are:

 

Is are there any physical limits to the number of agents that can be configured to a single Grinder console?

 

What is the largest number of Grinder Agent’s you have configured successfully with a single Console instance?

 

Any recommendations on value for grinder.reportToConsole.interval to minimize network bandwidth without sacrificing performance metrics (don’t want to rob network bandwidth from the application under test)?

 

Kind Regards,

 

Tom


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use

------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use

------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Large scale Grinder test configuration

Tom Vitez

Thanks Marc,

 

This is good to know… I will report back once I have things fully configured and working to let all know how many VU and agents I am using…

 

Tom

 

From: Marc Holden [mailto:[hidden email]]
Sent: Friday, March 25, 2016 8:42 AM
To: grinder-use <[hidden email]>
Subject: Re: [Grinder-use] Large scale Grinder test configuration

 

Hi Tom,

 

We have run over 40K VUs across 25 grinder agents successfully w/o adjusting reportToConsole interval.  

 

Depending on the size of the responses in the transactions, it is worth while to get some good network/systems monitoring in place and coordinate with the appropriate groups.

 

Thanks,

 

-Marc

 

 

 

On Thu, Mar 24, 2016 at 8:57 PM Darren Ball <[hidden email]> wrote:

Why not segregate your networks?  Do you have the option of multiple networks?  Have your agents drive tests on a private network, and your console and agent communication with the console on another?

On Mar 24, 2016 2:56 PM, "Tom Vitez" <[hidden email]> wrote:

Hi Grinder Users!

 

I have been using Grinder successfully on a number of projects over the past few years. My past tests have been relatively small in terms of vUser or thread count (i.e. < 500) but with high transaction rates (TPS). I have a need to run a test in the order of 30-40,000 vUsers so I need to spread across many Grinder agent VM’s and tie them back to a single Grinder Console instance to consolidate the results.

 

My questions are:

 

Is are there any physical limits to the number of agents that can be configured to a single Grinder console?

 

What is the largest number of Grinder Agent’s you have configured successfully with a single Console instance?

 

Any recommendations on value for grinder.reportToConsole.interval to minimize network bandwidth without sacrificing performance metrics (don’t want to rob network bandwidth from the application under test)?

 

Kind Regards,

 

Tom


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use

------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785471&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Large scale Grinder test configuration

Tom Vitez
In reply to this post by Darren Ball

Hi Darren,

 

This is a good recommendation, I will see if I can request separate networks for this configuration.

 

Tom

 

From: Darren Ball [mailto:[hidden email]]
Sent: Thursday, March 24, 2016 8:49 PM
To: grinder-use <[hidden email]>
Subject: Re: [Grinder-use] Large scale Grinder test configuration

 

Why not segregate your networks?  Do you have the option of multiple networks?  Have your agents drive tests on a private network, and your console and agent communication with the console on another?

On Mar 24, 2016 2:56 PM, "Tom Vitez" <[hidden email]> wrote:

Hi Grinder Users!

 

I have been using Grinder successfully on a number of projects over the past few years. My past tests have been relatively small in terms of vUser or thread count (i.e. < 500) but with high transaction rates (TPS). I have a need to run a test in the order of 30-40,000 vUsers so I need to spread across many Grinder agent VM’s and tie them back to a single Grinder Console instance to consolidate the results.

 

My questions are:

 

Is are there any physical limits to the number of agents that can be configured to a single Grinder console?

 

What is the largest number of Grinder Agent’s you have configured successfully with a single Console instance?

 

Any recommendations on value for grinder.reportToConsole.interval to minimize network bandwidth without sacrificing performance metrics (don’t want to rob network bandwidth from the application under test)?

 

Kind Regards,

 

Tom


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785471&iu=/4140
_______________________________________________
grinder-use mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/grinder-use
Loading...