Using MPLS TE in Real Life

Three basic real-life applications for MPLS TE are

• Optimizing your network utilization

• Handling unexpected congestion

• Handling link and node failures

Optimizing your network utilization is sometimes called the strategic method of deploying MPLS TE. It's sometimes also called the full-mesh approach. The idea here is that you build a full mesh of MPLS TE-LSPs between a given set of routers, size those LSPs according to how much bandwidth is going between a pair of routers, and let the LSPs find the best path in your network that meets their bandwidth demands. Building this full mesh of TE-LSPs in your network allows you to avoid congestion as much as possible by spreading LSPs across your network along bandwidth-aware paths. Although a full mesh of TE-LSPs is no substitute for proper network planning, it allows you to get as much as you can out of the infrastructure you already have, which might let you delay upgrading a circuit for a period of time (weeks or months). This translates directly into money saved by not having to buy bandwidth.

Another valid way to deploy MPLS TE is to handle unexpected congestion. This is known as the tactical approach, or as needed. Rather than building a full mesh of TE-LSPs between a set of routers ahead of time, the tactical approach involves letting the IGP forward traffic as it will, and building TE-LSPs only after congestion is discovered. This allows you to keep most of your network on IGP routing only. This might be simpler than a full mesh of TE-LSPs, but it also lets you work around network congestion as it happens. If you have a major network event (a large outage, an unexpectedly popular new web site or service, or some other event that dramatically changes your traffic pattern) that congests some network links while leaving others empty, you can deploy MPLS TE tunnels as you see fit, to remove some of the traffic from the congested links and put it on uncongested paths that the IGP wouldn't have chosen.

A third major use of MPLS TE is for quick recovery from link and node failures. MPLS TE has a component called Fast Reroute (FRR) that allows you to drastically minimize packet loss when a link or node (router) fails on your network. You can deploy MPLS TE to do just FRR, and to not use MPLS TE to steer traffic along paths other than the ones your IGP would have chosen.

Chapters 9 and 10 discuss strategic and tactical MPLS TE deployments; Chapter 7 covers Fast Reroute.