This is Video #10 of our MPLS Traffic Engineering series, which is designed to help viewers gain basic understanding of MPLS Traffic Engineering and how to configure TE tunnels in Cisco routers.

In this video, we will do a lab that demonstrates how to configure Per VRF MPLS TE Tunnels.



  • All VRFs/ MPLS L3VPNs routes announced by a PE will have the PE peering IP address as the next-hop IP address.
  • This means all routes in all VRFs will take the same TE tunnel/s if that next-hop IP address is routed through the tunnel.
  • To have granular control, configure a different BGP next-hop address in each of the VRF. Only then the VRFs can be routed to different TE tunnels simply by configuring a static route of the respective BGP next-hop addresses having the corresponding tunnels as the exit interface.
  • The configured BGP next-hop addresses need to be announced to IGP.
  • MPLS TE tunnels which will have now different VRFs passing through them can still use the same tunnel destination.


!Change BGP next hop at the advertising PE
ip vrf CLIENT
 rd 2:2
 bgp next-hop Loopback100
 route-target export 2:2
 route-target import 2:2
!Add static routes at the head-end PE of the MPLS TE tunnel.
ip route tun200
ip route tun100