We present Scriptroute, a system that allows ordinary Internet users to conduct network measurements from remote vantage points. We seek to combine the flex-ibility found in dedicated measurement testbeds such as NIMI with the general accessibility and popularityof Web-based public traceroute servers. To use Scriptroute, clients use DNS to discover measurement serversand then submit a measurement script for execution in a sandboxed, resource-limited environment. The serversensure that the script does not expose the network to attack by applying source- and destination-specific filtersand security checks, and by rate-limiting traffic. Scriptroute code is publicly available and has been de-ployed on the PlanetLab testbed of 42 sites. As proof-of-concept, we have used it both to create RPT, a toolfor measuring routing trees toward a destination, and to repeat the experiment used to evaluate GNP, a recently proposed Internet distance estimation technique. We find that our system is flexible enough to implement avariety of measurement tools despite its security restrictions, that access to many remote vantage points makesthe system valuable, and that scripting is an apt choice for expressing and combining measurement tasks.

Structured peer-to-peer overlays provide a natural infrastructure for resilient routing via efficient fault detection and precomputation of backup paths. These overlays can respond to faults in a few hundred milliseconds by rapidly shifting between alternate routes. In this paper, we present two adaptive mechanisms for structured overlays and illustrate their operation in the context of Tapestry, a fault-resilient overlay from Berkeley. We also describe a transparent, protocol-independent traffic redirection mechanism that tunnels legacy application traffic through overlays. Our measurements of a Tapestry prototype show it to be a highly responsive routing service, effective at circumventing a range of failures while incurring reasonable cost in maintenance bandwidth and additional routing latency.

Overlay networks have made it easy to implement multicast functionality in wireless ad hoc networks. Their flexibility to adapt to different environments has helped in their steady growth. In MANETs, the position of nodes constantly changes; as a result, overlay multicast trees that are built using location information to account for node movement would certainly have a low latency. However, the performance gains of such a tree are offset by the overhead involved in maintaining precise location information. As the degree of (location) accuracy increases, the performance improves but the overhead required to store and broadcast this information also increases. In this paper, we present SOLO-Net�, a design to build a sub-optimal location aided overlay multicast tree, where location updates of each member node are event based. Our simulation results indicate that such a sub-optimal tree does not compromise the performance gains of a location aided overlay multicast tree. 1.

ABSTRACT We describe a new scalable application-layer multicast protocol, specif-ically designed for low-bandwidth, data streaming applications with large receiver sets. Our scheme is based upon a hierarchical cluster-ing of the application-layer multicast peers and can support a number of different data delivery trees with desirable properties.We present extensive simulations of both our protocol and the Narada application-layer multicast protocol over Internet-like topolo-gies. Our results show that for groups of size 32 or more, our protocol has lower link stress (by about 25%), improved or similar end-to-end latencies and similar failure recovery properties. More importantly, it is able to achieve these results by using orders of mag-nitude lower control traffic. Finally, we present results from our wide-area testbed in whichwe experimented with 32-100 member groups distributed over 8 different sites. In our experiments, averagegroup members establishedand maintained low-latency paths and incurred a maximum packet loss rate of less than 1 % as members randomly joined and left themulticast group. The average control overhead during our experiments was less than 1 Kbps for groups of size 100.

This paper investigates the properties of an edge-based ow control framework that could make ita viable data-plane building block for quality of service (QoS) architectures. We consider two broad categories of owcontrol: end host-based (e.g., TCP), and network-based (e.g., Fair Queuing, CSFQ) that di er in the choice of nodes that cooperate and where functionality is placed. We propose an edge-based closedloop (EC) ow control framework in the network-based category in which only network nodes are expected to cooperate. The novelty of the EC-framework lies in its entire placement of ow control functionality at edge nodes, except interior core routers provide one or two isolated FIFO queues for the overall framework. The framework is divided into logical components such as congestion estimation and congestion response. Component instances can be combined to form schemes that do not depend upon, but accommodate, packet dropping or marking for congestion detection at core routers. We show that the Vegas is one possible scheme in the EC-framework, and propose two new schemes, \Monaco " and \Riviera". A uid model analysis is used to develop key concepts, to provide a reference for packet system implementations, and to demonstrate stability, fairness and bounded queue behavior. Simulations illustrate scheme performance, robustness comparisons, potential solutions to pitfalls of existing mechanisms. Architectural mapping of the EC-framework to Di Serv and commercial scenarios like cross-ISP data VPN is discussed. 1.