Customer Net A Customer Net D Customer Net C Customer Net

As IP technologies providing both tremendous capacity and the ability to establish dynamic secure associations between endpoints emerge, Virtual Private Networks (VPNs) are going through dramatic growth. The number of endpoints per VPN is growing and the communication pattern between endpoints is becoming increasingly hard to forecast. Consequently , users are demanding dependable, dynamic con-nectivity between endpoints, with the network expected to accommodate any traac matrix, as long as the traac to the endpoints does not overwhelm the rates of the respective ingress and egress links. We propose a new service interface , termed a hose, to provide the appropriate performance abstraction. A hose is characterized by the aggregate traac to and from one endpoint in the VPN to the set of other endpoints in the VPN, and by an associated performance guarantee. Hoses provide important advantages to a VPN customer: (i) exibility to send traac to a set of endpoints without having to specify the detailed traac matrix, and (ii) reduction in the size of access links through multiplexing gains obtained from the natural aggregation of the ows between endpoints. As compared with the conventional point to point (or customer-pipe) model for managing QoS, hoses provide reduction in the state information a customer must maintain. On the other hand, hoses would appear to increase the complexity of the already diicult problem of resource management to support QoS. To manage network resources in the face of this increased uncertainty, we consider both conventional statistical multiplexing techniques, and a new resizing technique based on online measurements. To study these performance issues, we run trace driven simulations, using traac derived from AT&T's voice network , and from a large corporate data network. From the customer's perspective, we nd that aggregation of traac at the hose level provides signiicant multiplexing gains. From the provider's perspective, we nd that the statistical multi-plexing and resizing techniques deal eeectively with uncertainties about the traac, providing signiicant gains over the conventional alternative of a mesh of statically sized customer-pipes between endpoints. 1 Introduction Virtual Private Network services have been ooered in various forms over an extended period of time and have recently received considerable attention within the IP, frame-relay, MPLS, and ATM networking communities 1, 2, 3, 4, 5]. VPNs are likely to be used by customers as a replacement for networks constructed using private lines and should therefore , at the very least, provide a comparable …