Re: VPNV4 Next Hop path manipulation querry

From: Vinu <vinupeter_at_gmail.com>
Date: Tue, 21 Jul 2009 18:24:21 +0400

VPNV4 backdoor ? wow is there something like that ? i couldn't even find it
on Google ?

On Tue, Jul 21, 2009 at 4:06 PM, Con Spathas <ccie19226_at_googlemail.com>wrote:

> I took Mahesh's topology as being
>
> IGP (OSPF in this case) within AS12 and AS34 for loopbacks and
> internal-AS networks
>
> ipv4 eBGP between R1-R3 and R2-R4 using directly connected + send-label
> ipv4 iBGP between R1-R2 and R3-R4 using loopbacks and set
> next-hop-self + send-label
>
> vpnv4 eBGP between R1-R3 and R2-R4 using loopbacks and ebgp-multihop
> vpnv4 iBGP between R1-R2 and R3-R4 using loopbacks and set next-hop-self
>
> LDP between R1-R2 and R3-R4
>
> To satisfy requirement 1 - I set local-pref 10000 to ipv4 routes
> inbound from AS12 (R1) on R3
> To try and satisfy requirement 2 - I set weight 10000 for only R2
> Loopback inbound from AS12 (R2) on R4
>
>
> On Tue, Jul 21, 2009 at 12:46 PM, Rich Collins<nilsi2002_at_gmail.com> wrote:
> > There is no intra-AS vpnv4 peering?
> >
> > On Tue, Jul 21, 2009 at 1:28 PM, Con Spathas<ccie19226_at_googlemail.com>
> wrote:
> >> Gday Mahesh,
> >>
> >> I've been toying with your question this morning.
> >>
> >> This might break your first requirement, however on R4 I manipulated
> >> the weight for R2 loopback learned from R2 so that it prefers the
> >> R4-R2 path instead of the R4-R3-R1-R2 path. R4 is still honoring
> >> getting to AS12 via R3 for everything except the loopback for R2. I
> >> figure by using weight - this is locally significant to R4 and won't
> >> influence the rest of AS34.
> >>
> >> That's where I'm not sure if doing this on R4 breaks your first
> >> requirement or actually addresses your second requirement and still
> >> satisfies the first requirement
> >>
> >> I'd be interested in reading other people's thoughts on this.
> >>
> >> Cheers.
> >> Con.
> >>
> >> On Tue, Jul 21, 2009 at 8:11 AM, mahesh s<maheshsachitanandan_at_gmail.com>
> wrote:
> >>> Hi,
> >>>
> >>> I am looking for a solution for VPV4 route path control
> >>>
> >>> Scenario,
> >>>
> >>> R1- -(AS12) - - -R2
> >>> | |
> >>> | |
> >>> | |
> >>> R3- --(AS34) - - R4
> >>>
> >>>
> >>> R1- R3 & R2- R4 ( IPV4 EBGP Peering on the the directly connected link)
> >>>
> >>> R1-R2 & R3-R4 ( IPV4 BGP peering is on the Loopback)
> >>>
> >>>
> >>> Inter AS, VPNV4 peering is between the loopback ( R1-R3 & R2-R4) and
> not on
> >>> the directly connected link.
> >>>
> >>> I am able to achieve , as well as able to bring up the LSP this using
> IPV4
> >>> BGP + Send lables
> >>> ----------
> >>> *Requirement*
> >>>
> >>> 1. Primary path for the IPV4 BGP traffic flow between from AS 34 to
> AS12
> >>> through R3
> >>>
> >>> 2. VPNV4 Customer Site is on R2 and any customer site in AS 34 to use
> R4
> >>> as primary exit
> >>> ----------
> >>> Unable to find a way to achieve the requirement 2 as the underlying
> IPV4 BGP
> >>> ( Carrying the transport label uses R3 as primary exit)
> >>>
> >>> Any suggestions VPNV4 backdoor or changing/manipulating the path for a
> VPNV4
> >>> neighbor
> >>>
> >>> Thanks for your help
> >>>
> >>> Mahesh
> >>>
> >>>
> >>> Blogs and organic groups at http://www.ccie.net
> >>>
> >>> _______________________________________________________________________
> >>> Subscription information may be found at:
> >>> http://www.groupstudy.com/list/CCIELab.html
> >>
> >>
> >> Blogs and organic groups at http://www.ccie.net
> >>
> >> _______________________________________________________________________
> >> Subscription information may be found at:
> >> http://www.groupstudy.com/list/CCIELab.html
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
>
>
>
>
>
>
>
>

-- 
Regards,
Vinu Peter
Blogs and organic groups at http://www.ccie.net
Received on Tue Jul 21 2009 - 18:24:21 ART

This archive was generated by hypermail 2.2.0 : Sat Aug 01 2009 - 13:10:22 ART