From: Joachim Baer (baerj@gmx.de)
Date: Fri Sep 24 2004 - 16:38:11 GMT-3
Hi,
sory may with my discription not clear.
==============
quoted
R3 and R1 next-hop-self for the internal AS
==============
Why next-hop-self on R3? Is it not enough on R1? or the link R1---R3 is not
part of IGP?
>>> It is the router R6 and R1 with next hop self.
=================
quoted
In the routing table of R3 the routes from AS30 rechable via loopback R1 and
loopback R6. The networks are reachable from R6 and R1. But not from R6,
because the next hop isn't an directly connected network (the loopback).
=================
Please correct me: the networks from AS30 are on the routing table of R3,
but when you check the entry with next-hop R6 it points out a reachability
problem. Is it that?
>>> Reachability Problem for both entrys R6 and R1, it is the loopback IP
for both routes (update-source loopback 0, on both router). So the next hop
for the routes are not in an direct connecteted network.
===============
quoted
1. Soulution is to redistribute the AS30 routes into igp at R1 and R6.
==============
ok, because R4 does not know how to deal with external routes
>>> I think in this scenario R4 don't reach the networks only the BGP
speeking Router. Redistribution isn' t permited.
==============
quoted
2. Soulution is to make gre tunnel and peer R1 to R3 and R6 to R3 via
gre-tunnel with source loopback and destination loopback and any network
over them.
==============
Why tunnel between R1 and R3? Looking at your diagram they seem to be direct
connected. The problem seems to be just R3-R4-R6
>>> The update source of R1 is the loopback interface and this isn't direct
connected.
==============
quoted
what is the best solution for this problem or there any other solutions.
==============
On real world, redistributing BGP into IGP would not be a best solution, if
you receive full Internet route.
On the lab, it would be aceptable.
>>> When it isn't permitted to redistribute GRE tunnels are ok?
> Hy, trying to understand your scenario.
>
> ==============
> quoted
> R3 and R1 next-hop-self for the internal AS
> ==============
> Why next-hop-self on R3? Is it not enough on R1? or the link R1---R3 is
> not part of IGP?
>
> =================
> quoted
> In the routing table of R3 the routes from AS30 rechable via loopback R1
> an
> loopback R6. The networks are reachable from R6 and R1. But not from R6,
> because the next hop isn't an directly connected network (the loopback).
> =================
> Please correct me: the networks from AS30 are on the routing table of R3,
> but when you check the entry with next-hop R6 it points out a reachability
> problem. Is it that?
>
> ===============
> quoted
> 1. Soulution is to redistribute the AS30 routes into igp at R1 and R6.
> ==============
> ok, because R4 does not know how to deal with external routes
>
> ==============
> quoted
> 2. Soulution is to make gre tunnel and peer R1 to R3 and R6 to R3 via
> gre-tunnel with source loopback and destination loopback and any network
> over them.
> ==============
> Why tunnel between R1 and R3? Looking at your diagram they seem to be
> direct connected. The problem seems to be just R3-R4-R6
>
> ==============
> quoted
> what is the best solution for this problem or there any other solutions.
> ==============
> On real world, redistributing BGP into IGP would not be a best solution,
> if you receive full Internet route.
> On the lab, it would be aceptable.
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Fri Oct 01 2004 - 15:00:48 GMT-3