Large-scale Internet applications can benefit from an ability to predict round-trip times to other hosts without having to contact them first. Explicit measurements are often unattractive because the cost of measurement can outweigh the benefits of exploiting proximity information. Vivaldi is a simple, light-weight algorithm that assigns synthetic coordinates to hosts such that the distance between the coordinates of two hosts accurately predicts the communication latency between the hosts.

Peer-to-peer (P2P) file sharing accounts for an astonishing volume of current Internet tra#c. This paper probes deeply into modern P2P file sharing systems and the forces that drive them. By doing so, we seek to increase our understanding of P2P file sharing workloads and their implications for future multimedia workloads. Our research uses a three-tiered approach. First, we analyze a 200-day trace of over 20 terabytes of Kazaa P2P tra#c collected at the University of Washington. Second, we develop a model of multimedia workloads that lets us isolate, vary, and explore the impact of key system parameters. Our model, which we parameterize with statistics from our trace, lets us confirm various hypotheses about file-sharing behavior observed in the trace. Third, we explore the potential impact of localityawareness in Kazaa.

Abstract — In this paper, we present the design, implementation, and evaluation of the iPlane, a scalable service providing accurate predictions of Internet path performance for emerging overlay services. Unlike the more common black box latency prediction techniques in use today, the iPlane builds an explanatory model of the Internet. We predict end-to-end performance by composing measured performance of segments of known Internet paths. This method allows us to accurately and efficiently predict latency, bandwidth, capacity and loss rates between arbitrary Internet hosts. We demonstrate the feasibility and utility of the iPlane service by applying it to several representative overlay services in use today: content distribution, swarming peer-to-peer filesharing, and voice-over-IP. In each case, we observe that using iPlane’s predictions leads to a significant improvement in end user performance. 1

While the advantages of multicast delivery over multiple unicast deliveries is undeniable, the deployment of the IP multicast protocol has been limited to "islands" of network domains under single administrative control. Deployment of inter-domain multicast delivery has been slow due to both technical and administrative reasons. In this paper we propose a Host Multicast Tree Protocol (HMTP) that (1) automates the interconnection of IP-multicast enabled islands and (2) provides multicast delivery to end hosts where IP multicast is not available. With HMTP, end-hosts and proxy gateways of IP multicast-enabled islands can dynamically create shared multicast trees across different islands. Members of an HMTP multicast group self-organize into an efficient, scalable and robust multicast tree. The tree structure is adjusted periodically to accommodate changes in group membership and network topology. Simulation results show that the multicast tree has low cost, and data delivered over it experiences moderately low latency. I.

Internet coordinate schemes have been proposed as a method for estimating minimum round trip time between hosts without direct measurement. In such a scheme, each host is assigned a set of coordinates, and Euclidean distance is used to form the desired estimate. Two key questions are: How accurate are coordinate schemes across the Internet as a whole? And: are coordinate assignment schemes fast enough, and scalable enough, for large scale use? In this paper we make contributions toward answering both those questions. Whereas the coordinate assignment problem has in the past been approached by nonlinear optimization, we develop a faster method based on dimensionality reduction of the Lipschitz embedding. We show that this method is reasonably accurate, even when applied to measurements spanning the Internet, and that it naturally leads to a scalable measurement strategy based on the notion of virtual landmarks.

This paper introduces Lighthouse, a scalable location mechanism for wide-area networks. Unlike existing vector-based systems such as GNP, we show how network-location can be established without using a xed set of reference points. This lets us avoid the communication bottlenecks and single-points-of-failure that otherwise limit the practicality of such systems.

mechanism to estimate Internet network distance (i.e., round-trip delay or network hops). Network distance estimation is important in many applications, for example, network-aware overlay construction and server selection. There are several proposals for distance estimation in the Internet but they all suffer from problems that limit their benefit. Most rely on a small set of infrastructure nodes that are a single point of failure and limit scalability. Others use sets of peers to compute coordinates but these coordinates can be arbitrarily wrong if one of these peers is malicious. While it may be reasonable to secure a small set of infrastructure nodes, it is unreasonable to secure all peers. PIC addresses these problems: it does not rely on infrastructure nodes and it can compute accurate coordinates even when some peers are malicious. We present PIC's design, experimental evaluation, and an application to network-aware overlay construction and maintenance.

Pangaea is a wide-area file system that supports data sharing among a community of widely distributed users. It is built on a symmetrically decentralized infrastructure that consists of commodity computers provided by the end users. Computers act autonomously to serve data to their local users. When possible, they exchange data with nearby peers to improve the system's overall performance, availability, and network economy. This approach is realized by aggressively creating a replica of a file whenever and wherever it is accessed. This paper presents

This paper introduces a lightweight, scalable and accurate framework, called Meridian, for performing node selection based on network location. The framework consists of an overlay network structured around multi-resolution rings, query routing with direct measurements, and gossip protocols for dissemination. We show how this framework can be used to address three commonly encountered problems, namely, closest node discovery, central leader election, and locating nodes that satisfy target latency constraints in large-scale distributed systems without having to compute absolute coordinates. We show analytically that the framework is scalable with logarithmic convergence when Internet latencies are modeled as a growthconstrained metric, a low-dimensional Euclidean metric, or a metric of low doubling dimension. Large scale simulations, based on latency measurements from 6.25 million node-pairs as well as an implementation deployed on PlanetLab show that the framework is accurate and effective.

Embedding of a graph metric in Euclidean space efficiently and accurately is an important problem in general with applications in topology aggregation, closest mirror selection, and application level routing. We propose a new graph embedding scheme called Big-Bang Simulation (BBS), which simulates an explosion of particles under force field derived from embedding error. BBS is shown to be significantly more accurate, compared to all other embedding methods including GNP. We report an extensive simulation study of BBS compared with several known embedding schemes and show its advantage for distance estimation (as in the IDMaps project), mirror selection and topology aggregation.