From: Brian McGahan (bmcgahan@internetworkexpert.com)
Date: Fri Jan 26 2007 - 11:22:46 ART
It's used to prevent the case where penultimate hop popping
occurs too soon and a device without VPNv4 routes sees the BGP bound VPN
label instead of the LDP/TDP bound IGP label.
For example suppose you have the following topology:
CE1--PE1--NetX--P2--P3--NetY--PE4--CE5
PE1 and PE2 peer VPNv4 using their NetX and NetY interfaces
instead of /32 loopbacks. For PE4 to get to CE1 routes it goes to PE1
via P3 via label L1. For P3 to get to PE1 it goes to P2. Since the
peering using NetX P2 advertises implicit-null to P3, and P3 performs
PHP. When the unlabeled packet gets to P2 it is dropped because P2 does
not have the VPNv4 routes. When /32 interfaces on PE1 and PE4 are used
for the peerings P2 performs PHP towards PE1 and P3 performs PHP towards
PE4, which is as it should be.
HTH,
Brian McGahan, CCIE #8593 (R&S/SP)
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
John Mistichelli
Sent: Friday, January 26, 2007 3:48 AM
To: CCIE Lab Group
Subject: MPLS VPN question
I am curios - does anyone know why MPLS VPNs require a
host route at head end and tail end?
John
7536
This archive was generated by hypermail 2.1.4 : Thu Feb 08 2007 - 23:46:57 ART