Why labels in MPLS?

[jrobert]

Am very new to MPLS concepts. So kindly help me to get my doubt clarified.

In routers, we have table to maintain the ip, in-label, out-going interface and out-label.

Instead of using the labels, just cant we use the out-going interface (exit interface) to direct the packets to reach the destination?

kindly help me out.....

[Giuseppe Larosa · Accepted solution]

Hello Robert,

the objective is to be able to foward MPLS frame without examining the payload (that can be IPv4 or other stuff) the label is needed to make possible for transit routers to swap labels: the information of outgoing interface is useful but it is not enough.

Different LSPs can have part of their path in common but at some place in the network they can take different paths to reach different Edge LSR / PE nodes.

For this reason each end to end path (unidirectional) needs its own set of labels to be used at each router hop in the path.

This is similar to what happens in WAN technologies like frame relay or ATM where a logical circuit identifier is used at each switch hop.

With FR an end to end path uses for example DLCI 50 then DLCI 88,  ... and reach the other device with DLCI 201 for example.

the label values can be reused in different places of the network exactly as it happens for FR DLCI values.

The labels have local meaning as it happens for FR DLCI values.

Hope to help

Giuseppe

[jrobert]

Hi Giuslar,

     Thank you so much for you clarification about my question. kindly help me in due course of time as am new to MPLS.