PREVIOUSLY: Configuring MPLS
This tip is sixth in a series designed to help you learn skills required for the MPLS portion of the Cisco CCIP certification. These skills are tested in the CCIP 642-611 - Implementing Cisco MPLS Exam.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
In the last tip on MPLS configuration, I discussed how to turn on MPLS service on the router and how to bind the labels to the IP packets. Here, I am going to focus on the creation and configuration of the Virtual Routing and Forwarding (VRF) delivered by Layer 3 MPLS VPNs.
The concept of virtual routing is an important aspect of MPLS. It enables PE routers to appear to be many routers to the customer edge (CE) routers. The provider edge (PE) router will maintain separate and distinct routing tables for each customer. Each PE builds these unique routing tables with their own routing table mechanisms for each customer that is connected to the PE. This unique separation of routing tables allows PE routers to store routes and forward packets even if the customers are using identical addressing.
The Cisco components of the VRF are as follows:
- A routing table specific to each VRF
- The associated customer interfaces for each VRF
- A CEF table
- Routing protocols for exchange of routing information per customer
The key components of the VRF configuration are the VRF name and the route distinguisher. Remember that the route distinguisher is used to distinguish between overlapping addresses in the VRF. The route distinguisher can be of two forms:
When designing your MPLS backbone, it is very important to have a well thought-out VPN naming scheme and route distinguisher numbering. A good way to do the route distinguisher name is to utilise Autonomous System Numbers (ASN) for the 16 bit address and the second 32 bits something specific to customers (customer ASN numbers will work if they are using BGP). Think of the operational support teams when assigning names and route distinguishers. Use a nomenclature that allows identification of the customer or the customer's entry point into the network.
Let's assume the customers have the following network addressing scheme:
Customer 1: 192.168.1.0/24
Customer 2: 192.168.1.0/24
The following commands will illustrate how to create the VRF on the PE routers for two customers, "Customer A" and "Customer B."
P1# config t
P1(config)# ip vrf VPN_1
P1(config-vrf)# rd 1:1
P1(config)# ip vrf VPN_2
P1(config-vrf)# rd 1:2
The VRF commands must be configured on each PE that interconnects Customer 1 and Customer 2 CE routers.
What this configuration does is create a VRF table that the PEs utilise to exchange the customer routes. The customer routes are identical except for the route distinguisher. This allows the PE to distinguish the 192.168.1.0/24, Customer 1 routes from the 192.168.1.0/24, Customer 2 routes. These are called IP-VPN routes and are exchanged between the PE routers using Multiprotocol BGP (MP-IBGP). The next article is the last MPLS CCIP study article. In it I will discuss the configuration of MP-IBGP to exchange IP-VPN routes between PEs.
About the author: Robbie Harrell (CCIE#3873) is the National Practice Lead for Advanced Infrastructure Solutions for SBC Communications. He has more than 10 years of experience providing strategic, business, and technical consulting services to clients. Robbie is a graduate of Clemson University. His background includes positions as a Principal Architect at International Network Services, Lucent, Frontway and Callisma.