Previous

Content

Next 


1.0.- Introduction about MPLS  

Having an IP domain like this:

R1 and R6 are called edge routers; they are located just in the frontier of the IP domain. R2, R3, R4 and R5 are called core routers; they do not communicate directly with the external world but through the edge routers.
How to convert this IP domain to a MPLS domain?
Let's start naming the edge routers as LER (Label Edge Router) and the core routers as LSR (Label Switch Router); for now is just a way of naming. Later on we will find naturally and very descriptive these funny names.
What work each of these routers has to do? LER converts IP packets into MPLS packets when they are entering the domain, and when leaving, it converts them back from MPLS packets to IP packets. LSR should forward MPLS packets following some instructions contained in one table installed on it. Based on some information contained in the MPLS packet, called Label, it selects a register from its installed table, and following instructions given in this register, it forwards ahead the packet. Next figures explain this better:
 

 
LER receives an IP packet; then applying some internal process, it converts the packet in one MPLS packet and forward it into the MPLS domain.
 

 

LSR seems to have a more complicated work. It receives the MPLS packet, inspecting it to extract the information it requires to enter in the internal table. Having this information, which is called a label, it enters in the table with this value (imagine a numbered key that open a box with instruction of what to do next). With the value it can acceses the instructions required to know what to do with the packet and how to forward it to the next selected interface.

   


Previous

Content

Next