Hello Yogijain!
For MPLS TE you can think like this:
- MPLS - tunneling mechanism -
- TE - Traffic engineering features for traffic which is inside MPLS Tunnel.
Application cases:
- For one specific Application/Service you can assign MPLS TE tunnel to be used for bearing the application/service payload.
- For Each MPLS TE Tunnel you can define the specific/desired behavior - path to be taken by application, specific Bandwidth requirements - like you App need to get 1Gbps BW, and you tunnel will be setup with this requirements and should be supported by all intermediate nodes.
Current status that it cab be achieved with separate protocol RSVP and also RSVP relies in IGP (OSPF/ISIS) to flood the real status about BW and next-hop availability.
Currently, the big discussion and implementation is going on about adapting Segment-Routing, the idea behind is to use
already existing database of IGP to advertise the desired parameters for TE tunnel, but it can be issue with the depth of the label-stack. as a better solution is to use IPv6 with Segment-Routing header - pure IPv6 forwarding without MPLS :)
Take care and have a great day,
VIktor.