Summary

MPLS TE can be useful in networks where links are running close to capacity. Using MPLS TE can help utilize the bandwidth available in nonshortest paths. However, MPLS TE might not be needed if the links have low utilization or if bandwidth can be easily added by lighting up more fiber in the ground (adding more links). Even in networks in which some links are running at capacity, though, MPLS TE can be useful. As illustrated in this chapter's first example, MPLS TE can help redirect traffic in the network, resulting in better utilization and lower traffic loss. It need not always be used in full mesh fashion. For example, MPLS TE can be used just around choke points to redirect traffic in a small portion of the network without adding more bandwidth capacity.

MPLS FRR is the most important application of MPLS TE and is useful in protecting against failures at Layer 3, eliminating the cost of redundant and protected circuits. Many service providers are deploying MPLS FRR, but they are doing so without deploying end-to-end tunnels to exactly achieve a protection solution with one-hop primary tunnels for link protection.

Node and path protection are also useful techniques, depending on the topology of the network. Traffic can also be protected against node failures in distributed platforms using nonstop forwarding of traffic while the control plane is recovering from the failure. Finally, although MPLS TE is not a must-have in the network in which you want to run other MPLS services, it certainly has its benefits. Before deciding on MPLS TE, a cost-benefit analysis must be done to determine whether the benefit it provides is more than the cost it might incur. For smaller networks, metric manipulation can be sufficient for better traffic engineering. However, as the network size grows, metric manipulation can have a number of undesired effects. MPLS TE can be much easier to manage and responds much more dynamically to changing topologies than metric readjustment in large networks.