From: Howard C. Berkowitz (hcb@gettcomm.com)
Date: Sat Apr 19 2003 - 18:15:31 GMT-3
At 2:47 AM +0800 4/20/03, wsqccie@hotnail.com wrote:
>hi,group
>
> R4---R5---R3---R2
> /
> r6
>R4 and R5 are EBGP
>R5 and R6 ebgp
>R5 and R3 EBGP
>R6 and R2 Ebgpt
>there is a lo0 in R2 and distri into bgp,now
>how can we control that this route only is advised from r3, and r6
>will not broadcast it till the link between r3 and r2 is down,
Most people learn better if they understand how they are getting
somewhere, rather than just receiving the answer. So, here are some
questions, which begin with being able to state the problem clearly.
First, it's much harder to follow BGP topologies without AS numbers.
You are dealing with a matter of routing policy, which is about AS at
the higher level and routers within AS at the lower level. With your
drawings, I can't see the boundaries of the AS.
Is this what you are saying?
AS1-----AS2------AS3-----AS4
R4 R5 R3 R2
|
|
AS5
R6
To use the IOS feature that you will probably need, you will need to
identify the actual routes in question (e.g., 192.168.1.0/28), where
they originate,
and whether any attributes or filters are defined that affect route selection.
I'm only guessing that the route in question is of the loopback in R2.
It wouldn't be out of place to post the configs of your system that
is propagating the route everywhere. A show ip bgp on R2, R3, and R5
might also be helpful.
This archive was generated by hypermail 2.1.4 : Thu May 01 2003 - 13:35:58 GMT-3