From: Brent D. Stewart (brent@stewart.hickory.nc.us)
Date: Wed May 28 2003 - 18:50:16 GMT-3
IGRP ends out periodic updates and can adapt to changes in load and
reliability. EIGRP only sends updates when the topology changes, and so
cannot notify its neighbors of changes in load or reliability. Although
you can change the relevant k values (2,4 and 5), changing the values
could have unpredictable results and calculation won't update.
Brent
-----Original Message-----
From: Brian McGahan [mailto:brian@cyscoexpert.com]
Sent: Wednesday, May 28, 2003 5:37 PM
To: Brent D. Stewart; Danny.Andaluz@triaton-na.com;
ccielab@groupstudy.com
Subject: RE: EIGRP metric (just to be sure)
Brent,
What do you mean by "Load and reliability should not be included
in the calculation (by changing k values) because EIGRP does not send
out updates if they change." ?
Brian McGahan, CCIE #8593
Director of Design and Implementation
brian@cyscoexpert.com
CyscoExpert Corporation
Internetwork Consulting & Training
Toll Free: 866.CyscoXP
Fax: 847.674.2625
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf
Of
> Brent D. Stewart
> Sent: Wednesday, May 28, 2003 3:37 PM
> To: Danny.Andaluz@triaton-na.com; ccielab@groupstudy.com
> Subject: RE: EIGRP metric (just to be sure)
>
> No. For an excellent discussion of this, refer to Alvaro Retana's
book
> "EIGRP for IP".
>
> EIGRP builds a metric using the equation:
>
> Metric= (k1*bw + k2*bw/(256-load) + k3*delay) (k5/(reliability+k4)
>
> By default, k1=1, k2=0, k3=1, k4=0, k5=0 and k2, k4, and k5 really
> shouldn't ever be changed. K1 and k3 are very seldom changed. If
k5=0,
> the last part is ignored, so this reduces to:
>
> Metric= bw + delay
>
> Each of these deserves a discussion.
>
> "bw" is actually 10^7/min(bw). In other words, the bw part uses the
> slowest link along the path and is inversely proportional to 1
million.
>
> "delay" is an accumulation of all interface delays along the path.
>
> Load and reliability should not be included in the calculation (by
> changing k values) because EIGRP does not send out updates if they
> change. MTU isn't even an option, but the idea was that it might be
> used for minimum MTU discovery along a path. As far as I know, this
was
> never implemented.
>
> EIGRP actually also tracks hops. The default max is 100. EIGRP
doesn't
> use this in the metric, it's there as a loop-detection mechanism.
>
> Regards,
>
> Brent D. Stewart
>
> -----Original Message-----
> From: Danny.Andaluz@triaton-na.com
[mailto:Danny.Andaluz@triaton-na.com]
>
> Sent: Wednesday, May 28, 2003 2:41 PM
> To: ccielab@groupstudy.com
> Subject: EIGRP metric (just to be sure)
>
> Hello, Group. Is the below correct with regards to how eigrp picks a
> best
> route? Does it do it like the below or does each value get taken into
> account for the composite metric?
>
> BW
> Delay (in case of tie with BW) ?
> relia (in case of tie with delay) ?
> load (in case of tie with relia) ?
> MTU (in case of tie with load) ?
>
> Thanks,
> Danny
>
> ---
> Incoming mail is certified Virus Free.
> Checked by AVG anti-virus system (http://www.grisoft.com).
> Version: 6.0.484 / Virus Database: 282 - Release Date: 5/27/2003
--- Incoming mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.484 / Virus Database: 282 - Release Date: 5/27/2003
This archive was generated by hypermail 2.1.4 : Mon Jun 02 2003 - 15:13:49 GMT-3