Re: [ns] Many agents wrt large scale simulations

["Guo, Liang" <guol@cs.bu.edu>]

On Fri, 14 Dec 2001, Stanley M Bielski wrote:

> 
> Hi Folks,
> 
> I'm beginning to run fairly large simulations involving an 
> exponentially-increasing # of TCP and LossMonitor agents. Each agent only
> sends a finite amount of traffic, and after that they don't do anything.
> So I'm wondering if I will notice a significant speed-up if I 'destroy'
> each agent and associated sink after the data is sent. Does anyone have
> any experience in this area?
> 

It surely will and could avoid using too much memory, take a look at the
Web Traffic model by Haobo and Polly
(ns/webcache/webtraf.* and ns/tcl/webcache/webtraf.tcl)
to see how these agents could be "recycled".

> Holiday Wishes,
> -Stan
> 
>