Friday, January 13, 2017

Approach to Service Provider domain : MPLS Basics ( RD and VRF concept )

Approach to Service Provider domain : MPLS  ( RD and VRF concept )

Well when we are talking about the service providers domain, the first thing came into mind is MPLS- Multi Protocol labelled switching

So MPLS is the technology used by service providers to connect the customer sites across the globe. To understand the MPLS we should know the basic routing, the concept of CEF, what is VRF, how data plane and control plane and also why we are using the various protocol like LDP/TDP, MP-iBGP protocol and what is label switching, what is PUSH/POP operations.

Audience : CCIE Service Providers, Network Engineers, Cisco enterprise engineers and TAC/NOC engineers and the systems engineer.

Lets start with the basic concept of MPLS, in short we can say that MPLS where routing at the edge and label switching in the core operates.The Edge devices in the MPLS network is PE router, which connected to CE ( Customer Edge router ) via routing protocol, it can be BGP or static routing that totally depends upon the number of the subnets routed into the domain of service provider.

We have two kind of MPLS services taken by customer, it can be L2MPLS and L3MPLS, now question what is L2MPLS and L3MPLS and when and where they can be opt for the operation.

What is the difference between L2MPLS and L3MPLS

L2MPLS is taken by the customer where customer don't want to share the L3 routing information with the service provider, it means the instead of pushing the packets from CE routers, we are pushing the frames from L2 device ( Frame relay switch or ATM Switch ) to the customer. The L2MPLS tunnel created end to end via the L2 media switches.

L3MPLS is taken by the customer where customer ready to share the L3 routing information with the customer and want to connect to the remote site. So the packets are being transferred from CE router to PE routers via routing protocol ( Static route or BGP or any other, depends upon the network design which customer actually requires )

So hope you understand the basic difference between L2MPLS and L3MPLS features and approach.

For L3MPLS, a single service provider PE router is connected to the various CE routers of different customer having the same IP subnet at local network. The subnets can be segregated  via RD ( Route distinguisher ) with VRF. Let me take an example

Fig 1.1- MPLS basic operations

Customer -1 : Subnet
Customer -2 : Subnet
Both the customers have the same subnet in their local domain and send the subnets to PE routers at service provider end, now it can only be distinguished via VRF+RD at the PE end. Now the situation is like
Customer -1 : Network-1 ( subnet ) + VRF-1 + RD100---- VPN 1 now
Customer -2 : Network-2 ( Subnet ) + VRF-2 + RD200---- VPN 2 now

So now the network is different for both the customers. I hope you guys understand the concept of RD and VRF here.

RD : Route Distinguisher - used to distinguished the same subnet getting from two different customer
VRF: Virtual route Forwarding - Its the Logical name to distinguish the network in the MPLS domain.

In next session we will come with the more concepts on RT, MP-iBGP, LDP/TDP, Label Swapping and other concepts used in the MPLS to understand the whole picture of the Service provider domain.

Networks Baseline

Author & Editor

Networks Baseline is a group of Network Engineers having the huge experience in Cisco network and architectural domain.

Popular Posts