Re: EIGRP - Split Horizon

>For the sake of it, lets forget about this other rule for now: "The
>split-horizon rule prohibits a router from advertising a route
through an
>interface that the router itself uses to reach the destination."

Why forget about this rule? Especially if it explains the behavior
perfectly?

Did you read Dave's most recent post carefully? He stated that when
he increased the metric on R3 to reach R2, R3 ceased advertising R1
Lo0 to R4. How does that square with your concept of an EIGRP path
(would love to see some supporting documentation on this if you have
any)? R3 is learning of R1 Lo0 from both R2 and R4 - in your scenario
these would remain unique "paths," so why, according to your logic,
would R3 suddenly cease advertising it to R4 due to a simple metric
increase? Would not the "path" via R2 remain unique, even if it
became a little more expensive to use? Is not the "path" received
from R2 coming in on a different INTERFACE than the one connecting R3
to R4? Your quote: "R3 is able to advertise the route that it
received from R2 because that route was received on a different
interface - same prefix, but different path and interface." So my
question would be:

Why does this behavior suddenly change with a simple metric increase
along the R3->R2->R1 "path?" My answer would be:

See the above rule you for some reason wanted us to forget about.

Blogs and organic groups at http://www.ccie.net
Received on Mon Jan 10 2011 - 10:40:55 ART