This article is the second in a series tailored towards certification skills required for the MPLS portion of the Cisco CCIP certification. These skills are tested in the CCIP exam 642-611 - Implementing Cisco MPLS.
There are multiple concepts you must understand regarding the overall MPLS architecture. The MPLS architecture is defined as how an IP backbone builds label forwarding tables and actually provides the service of forwarding labeled traffic. Key concepts include the following:
- Forward Equivalence Class
- MPLS Control Plane
- MPLS Forwarding Plane
- Label Distribution Protocol
Basically, an MPLS router makes a forwarding decision for labeled packets in the same manner that it does for an IP packet. It will look at the label on the incoming packet, then consult the forwarding table to identify the interface to forward the labeled packet out of, then it will forward the packet.
The key to making MPLS work is that routers must build the necessary label forwarding tables upon which MPLS rests. Just as routing protocols build forwarding tables by exchanging information between themselves using dynamic routing protocols such as OSPF and BGP, so also must routers build label forwarding tables. This involves the MPLS control plane, which binds labels to the networking routes in the IP routing table, and which manages the distribution of those bindings to other MPLS routers. To make this happen, the control plane uses the label distribution protocol to exchange labels between MPLS routes. Each MPLS router assigns a local label and an outgoing label for such routes that it advertises to its neighbors.
MPLS uses the concept of a forward equivalency class (FEC) to designate a group of IP packets to which a specific label is bound. The group of IP packets is a loose term as this really means an IP prefix or a route. The FEC can be more granular that just a network prefix as it can also specify an IP prefix with other criteria such as IP TOS bits.
So the router identifies an FEC out of the IP routing table and assigns a label to it. Then the router advertises to its peers that the FEC has this label assigned to it. The list of FECs with labels assigned to them is called the label information base (LIB). The router makes forwarding decisions based on the information in the LIB. The labels that are actually in use for forwarding are placed in the label forwarding information base (LFIB). This is the MPLS forwarding plane.
So let's put it all together. The control plane assigns labels to IP routes/prefixes in the IP routing table. The MPLS forwarding plane builds the label forwarding table that indicates which interface to forward the labeled packet out of and the label distribution exchanges label information.
The next step that the router performs is to build the labeled switch path through the network. This is a virtual path across which traffic will be forwarded from one end of the network to the other.
This is the main premise of MPLS label switching in terms of label distribution and the building of the label forwarding tables. Be sure to understand these concepts well. In the next article we will discuss the actual label distribution between the MPLS peers and traffic flow using labeled packets across the network.
About the author:
Robbie Harrell (CCIE#3873) is the National Practice Lead for Advanced Infrastructure Solutions for SBC Communications. He has over 10 years of experience providing strategic, business, and technical consulting services to clients. Robbie resides in Atlanta, and is a graduate of Clemson University. His background includes positions as a Principal Architect at International Network Services, Lucent, Frontway and Callisma.