Pipeline Publishing, Volume 4, Issue 4

	This Month's Issue:
	Maintaining Network Health

VRF - The Key to MPLS Performance Assurance

By Bruce Kelley

To satisfy demands for secure, cost-effective transport of converged voice and business applications, most telecommunications providers have introduced MPLS-based VPN service offerings. Enterprises are taking advantage to break out delivery for voice and video into high-priority classes and tiered choices for their more latency-tolerant business applications. In order to meet service quality expectations, carriers are facing some new challenges:

They need real-time, application-aware analysis of activity across the MPLS core
They must distinguish traffic between individual customers and locations
They have to track routing activity as an integral part of their traffic engineering tasks

In short, providers need to re-evaluate the ways in which they monitor performance as part of their assurance practices.

Network Considerations

Any performance monitoring approach needs to embrace awareness of how traffic is transported across the MPLS core. An MPLS network will adhere to the RFC 2547bis standard, which drives how VPN services are provided to customers. An MPLS network creates VPN tunnels based on MPLS routing and forwarding tables for each customer site connected to the service provider’s MPLS network.

A customer site is connected to a service provider network via one or more ports, and the service provider associates each port with a specific VPN routing and forwarding identifier known as a VRF. Since each customer likely uses internal IP addresses which may be duplicates of other customers’, a performance monitoring approach needs to examine and distinguish an individual customer’s traffic between the PE (provider edge) and P (provider core) routers in the MPLS network. The figure below provides an illustration of a simple MPLS network deployment.

Providers need to re-evaluate the ways in which they monitor performance as part of their assurance practices.

Two basic traffic flows occur in an MPLS VPN. The first is a control flow that is used for VPN route distribution and label switched path (LSP) establishment. The second is the actual data flow that is used to forward customer data traffic. Any MPLS monitoring approach will need to leverage information from both of these traffic flows to troubleshoot and plan capacity.

Possible Approaches

Many providers are recognizing the need for new methods of monitoring their MPLS services. First, they need to track dynamic MPLS labels in real-time, accommodate duplicate customer IP addresses, and recognize heterogeneous mixes of application traffic. This means that MPLS providers need approaches that provide detailed visibility by service and application, despite the fact that the viewpoint will be a network perspective. Also important is the ability to isolate and rapidly correct not just faults and failures, but a more challenging class of issues - service degradations. This is especially important with the advent and growth of latency-sensitive applications, such as IP voice and video conferencing.

Service performance assurance is a significant challenge for many operators because the sources of management data that they are accustomed to using do not provide the

article page | 1 | 2 | 3 |

© 2006, All information contained herein is the sole property of Pipeline Publishing, LLC. Pipeline Publishing LLC reserves all rights and privileges regarding
the use of this information. Any unauthorized use, such as copying, modifying, or reprinting, will be prosecuted under the fullest extent under the governing law.