From: Victor Cappuccio (cvictor@protokolgroup.com)
Date: Tue Jul 25 2006 - 12:49:01 ART
Hmmmm So no simple solution :D
How can you configure this scenario using MPLS?
Thanks to both
Victor.-
-----Mensaje original-----
De: Brian McGahan [mailto:bmcgahan@internetworkexpert.com]
Enviado el: Martes, 25 de Julio de 2006 11:43 a.m.
Para: Victor Cappuccio; Ivan; ccielab@groupstudy.com
Asunto: RE: Transiting Non-BGP Speaking Devices
Yes you would have to modify R3's IGP routing policy in order to
reflect your desired BGP policy. Remember that the traffic flow of BGP
traffic is always at the mercy of how IGP recourses to the next-hop
value.
HTH,
Brian McGahan, CCIE #8593
bmcgahan@internetworkexpert.com
Internetwork Expert, Inc.
http://www.InternetworkExpert.com
Toll Free: 877-224-8987 x 705
Outside US: 775-826-4344 x 705
24/7 Support: http://forum.internetworkexpert.com
Live Chat: http://www.internetworkexpert.com/chat/
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf
Of
> Victor Cappuccio
> Sent: Tuesday, July 25, 2006 10:25 AM
> To: 'Ivan'; ccielab@groupstudy.com
> Subject: RE: Transiting Non-BGP Speaking Devices
>
> Ivan, I know that R3 is having that problem.
>
> The thing here is that you need to apply your Redistribution in a Way
to
> guarantee that Any BGP Local Policy is reflected also in the BGP-> IGP
> redistribution
>
> If you do normal redistribution (BGP -> IGP), R3 would load balance
> between
> the BGP routes, if running the same routing Protocols with R1 and R2.
> But Say that R1 - R3 is running a different Routing Protocol that R3 -
R2
> That makes things harder right?
>
> I hope a Routing Guru could chip in here.
>
> Thanks
> Victor.
>
>
> -----Mensaje original-----
> De: Ivan [mailto:ivan@iip.net]
> Enviado el: Martes, 25 de Julio de 2006 11:13 a.m.
> Para: ccielab@groupstudy.com; Victor Cappuccio
> Asunto: Re: Transiting Non-BGP Speaking Devices
>
> This issue appears because R3 don't know about external BGP routes. To
> force
>
> R3 learn BGP routes you can redistribute to IGP or set static route to
> some
> gateway.
>
> Also synchronisation may guarantee that all routers have consistent
FIB.
> If choose synchronisation, there must be a match for prefixes in IP-
> routing
> table to iBGP path consider as valid. Usually this prefixes extend
with
> IGP
> therefore all routers on path "know" about originating this prefixes.
In
> this
> case need to note about OSPF RID and BGP RID. I think in this case
> redistribution must take place.
>
> > Hi Brian, thanks for your reply
> >
> > What I'm trying to do here, is to obtain global rechability, but R3
in
> this
> > case is not running BGP.
> >
> > I know that I can peer the loopback of the remote-neighbor, but if I
> ping
> > from R5 to any BB attached to R1 or R2; R3 would drop all Packets,
sat
> that
> > R1 and R2 peer to BB1 and BB3 and this 2 routers belongs to AS 54
> > (something similar to your WB topology, but with BB1 also connected
to
> Sw2
> > in port f0/23)
> >
> > Yes: R1,R2,R4,R5 are in the same AS; and they peer to each other in
a
> full
> > mesh fashion using their loopbacks.
> >
> >
> > for example lets say that R2 -- BB3 is setting the local Preference
200
> for
> > all BGP routes and a weight of 1000; and R1 -- BB1 is setting the
Local
> > Preference for all BGP route to 100 and a Weight of 1000.
> >
> > If you see the BGP Routing Table at R4 or R5, if would prefer R2 to
exit
> > the AS, but if you trace the BGP route it would be droped at R3,
another
> > point here is that when you redistribute BGP routes into IGP, it
would
> not
> > respect the Local AS Policies, so simple redistribution could not be
> > accomplished (I think).
> >
> > I know this is disparate, but it would be nice to have something
like a
> > Multipoint Tunnel at each router.
> >
> > Please let me know if you need some more information
> > Saludos and Thanks
> > Victor.-
> >
> >
> >
> > -----Mensaje original-----
> > De: Brian McGahan [mailto:bmcgahan@internetworkexpert.com]
> > Enviado el: Martes, 25 de Julio de 2006 10:04 a.m.
> > Para: Victor Cappuccio; ccielab
> > Asunto: RE: Transiting Non-BGP Speaking Devices
> >
> > Victor,
> >
> > What exactly are you trying to accomplish, IP reachability
> > between R1, R2, R4, and R5? Are they all in the same AS? Where are
the
> > prefixes coming from?
> >
> >
> > Brian McGahan, CCIE #8593
> > bmcgahan@internetworkexpert.com
> >
> > Internetwork Expert, Inc.
> > http://www.InternetworkExpert.com
> > Toll Free: 877-224-8987 x 705
> > Outside US: 775-826-4344 x 705
> > 24/7 Support: http://forum.internetworkexpert.com
> > Live Chat: http://www.internetworkexpert.com/chat/
> >
> > > -----Original Message-----
> > > From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On
Behalf
> >
> > Of
> >
> > > Victor Cappuccio
> > > Sent: Monday, July 24, 2006 10:00 PM
> > > To: 'ccielab'
> > > Subject: Transiting Non-BGP Speaking Devices
> > >
> > > Hi Guys, I know this is a very newbie question, but it keeps
spinning
> >
> > in
> >
> > > my
> > > head.
> > >
> > > Ok this is the dilemma
> > >
> > > R1 --- R3 ---- R2
> > >
> > > R1-R2-R3 runs any IGP.
> > >
> > > R1 and R2 are running BGP in AS 12 and they peer via each other
> >
> > Loopback
> >
> > > Address (/32 BTW).
> > >
> > > So, I need to solve the Non-BGP Transitive Device Problem, I know
that
> >
> > I
> >
> > > can
> > > use tunnels or maybe redistribute BGP routes at R2 and R1.
> > >
> > > But the question is more difficult (for me at least); say that I
add
> > > another
> > > 2 BGP Devices connected to R3
> > >
> > > R5
> > > .
> > > .
> > > R1 ------ R3 ------ R2
> > > .
> > > .
> > > R4
> > >
> > > I need to create a full mesh BGP Session between R1; R2; R5; R4
using
> > > their
> > > loopbacks Address (/32 BTW).
> > >
> > > So creating tunnels here is out of the game, because you can not
add
> >
> > extra
> >
> > > Ip addressing.
> > >
> > > Now redistributing the BGP Routes to the current IGP, would NOT
help
> >
> > me if
> >
> > > I
> > > need to create some AS Policies. - Like Local Preference.
> > >
> > > Maybe MPLS would solve the problem (do not know how to configure,
and
> >
> > I
> >
> > > think that would be out of the scope of the CCIE Lab for now)
> > >
> > > Any recommendations for this particular problem?
> > >
> > > Thanks
> > > Victor.-
> >
> >
This archive was generated by hypermail 2.1.4 : Tue Aug 01 2006 - 07:13:48 ART