No abstract

We present RCM, a system to monitor link delays on a network using a single measurement host. RCM is a combination of a new measurement system and a new network tomography technique. The measurement system employs tunnels to connect to border routers where it can source and sink measurements across the network. RCM uses network tomography to calculate the delays across individual network links from these measurements. The network tomography technique expands on previous linear algebra techniques to deal with the limitations of the resulting data without assuming either link delay symmetry or a particular topology. The network tomographic technique is compared against direct measurements in simulation to ensure accuracy. RCM is deployed on a large ISP's network to diagnose the cause of end-to-end delays, from which additional results are presented. The results are compared against known behaviors of the network to ensure the results are consistent with those behaviors. The system is analyzed for its ability to pin-point the cause of changes in end-to-end delay.

No abstract

Virtual Private Networks provide secure and reliable communication between customer sites. With increase in number and size of VPNs, providers need efficient provisioning techniques that adapt to customer demand by leveraging a good understanding of VPN properties.In this paper we analyze two important properties of VPNs that impact provisioning -(a) structure of customer endpoint (CE) interactions and (b) temporal characteristics of CE-CE traffic. We deduce these properties by computing traffic matrices from SNMP measurements. We find that existing traffic matrix estimation techniques are not readily applicable to the VPN scenario due to the scale of the problem and limited measurement information. We begin by formulating a scalable technique that makes the most out of existing measurement information and provides good estimates for common VPN structures.We then use this technique to analyze SNMP measurement from a large IP VPN service provider. We find that even with limited measurement information we can realize adaptive provisioning for a significant fraction of VPNs, namely, those constituting the "Hub-and-Spoke" category. In addition, the ability to infer the structure of VPNs holds special significance for provisioning tasks arising from topology changes, link failures and maintenance. We are able to provide a classification of VPNs by structure and identify CEs that act as hubs of communication and hence require prioritized treatment during restoration and provisioning.

The growth of one-to-many applications in enterprise networks is fueling the demand for VPNs to support multicast applications. The deployment of such a Multicast VPN service creates the need for appropriate management tools and techniques including performance monitoring, problem isolation and troubleshooting. In this paper we present the MVPN monitoring problem, identify requirements for solutions to the problem, and describe VMScope, a virtual MVPN performance monitor designed to facilitate provider monitoring and debugging of the MVPN service.