MPLS Fundamentals

In this lesson, we will focus MPLS in Networking and we will learn the answer of what is MPLS? So, What is MPLS? MPLS is a Packet-Forwarding technology which uses MPLS Labels for data forwarding decisions. It is the abbreviation of Multi Protocol Label Switching. Over time, users in networks have increased and this big traffic has brought high bandwidth and forwarding speed need. The existing routing protocols are looking to the routing table for forwarding decision on all nodes and this was overloading the routers beside causing a lot of time lose. Here Multi Protocol Label Switching brings also CEF and this bypass the slow part of this layer 3 decisions. Beside that it brings the speed of OSI layer 2 switching. In brief, MPLS Networking is combining the beneficial parts of both Layer 2 and Layer 3 of OSI reference model.

If you would like to learn more about MPLS, you can follow the lessons of Nokia MPLS Training.

In addition to these negative facts, there were some other problems about ATM technology, while working together with Ethernet and Frame Relay. To solve all these problems IETF developed MPLS in 1997. This development could be done with existing routing protocols. But changing all the existing routing protocols was very difficult, so IETF has chosen developing a new protocol.

This new protocol is using Labels for routing decisions. These labels are router specific and local significant to that router.

MPLS Header

We have learned what is MPLS. Now, let’s focus MPLS Header. For MPLS Networking, a header is added between Layer 2 and Layer 3 header in packets. This MPLS Header is 32 bits long. This header consist of the below fields:

• Label (20 bits),
• Experimental Bits (3 bits),
• Bottom of Stack (1bit),
• Time to Live (8 bits)

As an example, you can check the below packet capture that includes also an MPLS Header. Here, MPLS Header is shown with blue rectangle. As you can see Multi Protocol Label Switching header is between the Layer 2 header (ethernet II header) and Layer 3 header.

You can test your knowledge with MPLS Questions

Firstly, I want to talk about Router terms used in this technology. There are different routers used in MPLS Technology. These routers are given below:

• Customer Router (C)
• Customer Edge Router (CE)
• Provider Router (P)
• Provider Edge Router (PE)

Customer Router is the router of customer, so it is not related to MPLS network.

Customer Edge Router is the router that is connected to MPLS PE Router of the Service Provider Network.

Provider Router is the router used in Service Provider Networks for MPLS Label Switching and Data Forwarding.

Provider Edge Router is the router that is located at the edges of the Service Provider Networks. They are the routers that VPN Services are defined. They are the borders between customer and Service Provider networks.

The below shape clearly explains the location of these routers.

Instead of Provider Router or Provider Edge Router, we can also use LSR (Label Switching Router) and Edge LSR (Edge Label Switching Router) terms for the routers capable of label switching. Specifically, Edge LSRs can be divided into two as Ingress Edge LSR and Egress Edge LSR. The below shape shows this clearly:

Label Distribution Protocols

The labels that will be used in MPLS network are distributed with some Label Distribution Protocols. And with these labels forwarding decision is done. As you remember from ip routing, this decision is done through IP addresses in normal routing. MPLS technology is basically a concept that changes a layer 3 routed network to a switched network.

LDP (Label Distribution Protocol)  is the simplest way for Label Distribution. Basically, LDP trusts Interior Gateway Protocol and selects the next hop that has already selected by IGP. Then, it takes label from this next hop. In other words, this type of label distribution is dependent to Routing Protocol decision.

RSVP-TE (Resource Reservation Protocol-Traffic Engineering), is another way used for Label Distribution. With this way, the next hop selection can be determined by IGP or can be specified manually. With this protocol, Traffic Engineering activities can be done that manages the flow of different types of traffic.

There are four type table that is related to Multi Protocol Label Switching. This tables are formed with CEF (Cisco Express Forwarding) mechanism and the Label Distribution Protocols. So what are the tables used in MPLS? You can find Multi Protocol Label Switching tables below:

• RIB (Routing Information Base): It is the classical routing table of the router. This table contains the classic routing information about the destination network.

• FIB (Forwarding Information Base): It is formed from RIB by CEF to accelerate route forwarding process.

• LIB (Label Information Base): It is the table that is formed by the label distribution protocol and it contains all the labels known by an LSR. It contains the label, LSR and network information.

• LFIB (Label Forwarding Information Base): It is the routing database for labeled packets. It determines how the incoming label will be forwarded to the next-hop.

Control Plane ve Data Plane

MPLS works in two different systems in Routers. These systems are Control Plane and Data Plane. So, what is MPLS Control Plane? What is MPLS Data Plane?

• Control Plane: Control plane is the system in which the Layer 3 Routing Protocols are work and the routing information is store. Beside protocols like RSVP and LDP works here.

• Data Plane: Data planeis the system that forwards the Label Based data. Data plane does this with the table (LIB) stored in control plane.

The following pictures of LSR and Edge LSR will explain the Control Plane and Data Plane more clearly. Here both Control Plane and Data Plane are shown beside the tables related with them.

• It combines the beneficial parts of Layer 2 and Layer 3 of OSI model,
• It decreases the router load, provide speed and performance increment,
• It provides easy managable VPN definitions between edge devices,
• With traffic engineering, it provides management of the traffic,
• It simplifies the IP forwarding over ATM devices,
• And more…