From: Mahmud, Yasser (YMahmud@Solutions.UK.ATT.com)
Date: Sun Sep 29 2002 - 21:58:19 GMT-3
Tom,
The equation has 2 conditions because of the point you mentioned
1) If K5=0 then
Metric = [ K1 * Bandwidth + (K2 * Bandwidth)/(256-load) + K3*Delay ]
2) If K5 is not equal to zero and hence want to take reliability into
consideration normally by setting it to 1 then the equation
Metric = [ K1 * Bandwidth + (K2 * Bandwidth)/(256-load) + K3*Delay
]*[K5/(reliability + K4)]
Yasser
> -----Original Message-----
> From: baganini@attbi.com [SMTP:baganini@attbi.com]
> Sent: Sunday, September 29, 2002 4:45 PM
> To: ccielab@groupstudy.com
> Subject: IGRP metric
>
> Hi all,
>
> Something very basic I have Noticed with the IGRP metric
> equation I found in Doyle's book and on the link
> http://www.cisco.com/warp/public/103/3.html I wanted to
> share with you. The metric equation is shown as
>
> Metric = [ K1 * Bandwidth + (K2 * Bandwidth)/(256-load) + K3*Delay
> ]*[K5/(reliability + K4)]
>
> Now this question is for all you mathematicians here.
> If K5 is 0 all what number is in the second square
> brackets will be zero. So, the result of
> [ K1 * Bandwidth + (K2 * Bandwidth)/(256-load) +
> K3*Delay ] multiplied by zero, will be zero, unless the
> * sign means something else other that multiplication.
> In order to have the second square bracket to be a term,
> you sign between them should be a (+), for example.
>
> What do think??
>
> --
> Tom Thouma
> CCIE# 7003
> Director of Research & Development
> thouma@cyscoexpert.com
>
> CyscoExpert, Inc.
> 4433 W. Touhy Ave.
> LincolnWood, IL 60712
> info@cyscoexpert.com
> Ph:(847) 674-3392
> FX:(847) 674-2625
This archive was generated by hypermail 2.1.4 : Mon Oct 07 2002 - 07:44:07 GMT-3