RE: BGP Reachability through non bgp routers - help

From: Robert Watson (watson.robert@gmail.com)
Date: Mon Nov 27 2006 - 10:37:35 ART

 This does not help. If I learn the 201.200.1.0/24 network from as 54 and
it gets sent to r1 then R5 will learn of this route through IBGP. Problem
arises when R5 tries to send packets to destination 201.200.1.0 network
first hop to get to said network is r4 which is not running BGP so we have a
problem. I am trying to canvas the community as to what a lab best practice
would be to resolve. Sync (igp to bgp redistribution) or tunnels or some
other solution

-----Original Message-----
From: Heiko Liedtke [mailto:heiko.liedtke@gmx.net]
Sent: Sunday, November 26, 2006 1:24 PM
To: Robert Watson
Cc: ccielab@groupstudy.com
Subject: Re: BGP Reachability through non bgp routers - help

I am not sure if you have the possibilty
to test this in a lab...

There will be no problem for R1 to receive AS254 routes. This is
BGP as usual.
The routes are sent to R1 via iBGP with the BB2 address as the next-hop.
(iBGP does not change any bgp attributes on received routes (if not told
by you))

If the segment between R5 and BB2 is not routed in your IGP, you might
have to use the command neighbor xxxxx next-hop-self on R5.
The AS254 routes on R1 should now have R5 as next-hop and everything should
be fine.
Redistributing BGP into IGP is never a good solution. Think about 200k
routes
redistributing into OSPF... this will melt down your network in seconds...

Hope this helps

Heiko

Robert Watson schrieb:

>Scenario
>
>R1 and R5 are ibgp peers in as 100
>
>Bb1 in as 54 connected to R1
>BB2 in as 254 connected to R5
>
>There are 3 non bgp routers between R1 and R5
>
>R1 - R2 - sw1 - r4 - R5
>
>How many different ways can R1 reach as 254 routes behind R5 and what is
>"lab best practice"
>
>Tunnels
>Sync (redistribute bgp to igp and back)
>
>_______________________________________________________________________
>Subscription information may be found at:
>http://www.groupstudy.com/list/CCIELab.html

This archive was generated by hypermail 2.1.4 : Fri Dec 01 2006 - 08:05:48 ART