We describe the design and implementation of SWORD, a scalable resource discovery service for wide-area distributed systems. In contrast to previous systems, SWORD allows users to describe desired resources as a topology of interconnected groups with required intra-group, inter-group, and per-node characteristics, along with the utility that the application derives from specified ranges of metric values. This design gives users the flexibility to find geographically distributed resources for applications that are sensitive to both node and network characteristics, and allows the system to rank acceptable configurations based on their quality for that application. Rather than evaluating a single implementation of SWORD, we explore a variety of architectural designs that deliver the required functionality in a scalable and highly-available manner. We discuss the tradeoffs of using a centralized architecture as compared to a fully decentralized design to perform wide-area resource discovery. To summarize our results, we found that a centralized architecture based on 4-node server cluster sites at network peering facilities outperforms a decentralized DHT-based resource discovery infrastructure with respect to query latency for all but the smallest number of sites. However, although a centralized architecture shows significant promise in stable environments, we find that our decentralized implementation has acceptable performance and also benefits from the DHT’s self-healing properties in more volatile environments. We evaluate the advantages and disadvantages of centralized and distributed resource discovery architectures on 1000 hosts in emulation and on approximately 200 PlanetLab nodes spread across the Internet.

Range query, which is defined as to find all the keys in a certain range over the underlying P2P network, has received a lot of research attentions recently. However, cover query, which is to find all the ranges currently in the system that cover a given key, is rarely touched. In this paper, we first identify that cover query is a highly desired functionality by some popular P2P applications, and then propose distributed segment tree (DST), a layered DHT structure that incorporates the concept of segment tree. Due to the intrinsic capability of segment tree in maintaining the sturcture of ranges, DST is shown to be very efficient for supporting both range query and cover query in a uniform way. It also possesses excellent parallelizability in query operations and can achieve O(1) complexity for moderate query ranges. To balance the load among DHT nodes, we design a downward load stripping mechanism that controls tradeoffs between load and performance. We implemented DST on publicly available OpenDHT service and performed extensive real experiments. All the results and comparisons demonstrate the effectiveness of DST for several important metrics. 1.

Abstract not found

This paper introduces a new consistency metric, Network Imprecision (NI), to address a central challenge in largescale monitoring systems: safeguarding correctness despite node and network failures. To implement NI, an overlay that monitors a set of attributes also monitors its own state so that queries return not only attribute values but also information about the stability of the overlay—the number of nodes whose recent updates may be missing and the number of nodes whose inputs may be double counted due to overlay reconfigurations. When NI indicates that the network is stable, query results reflect the true state of the system, but when the network is unstable, NI puts applications on notice that query results should not be trusted, allowing them to take corrective action such as filtering out inconsistent results. To implement NI’s introspection scalably, our prototype introduces a key optimization, dual-tree prefix aggregation, which exploits overlay symmetry to reduce overheads by more than an order of magnitude. Evaluation of three monitoring applications demonstrates that NI flags inaccurate results while incurring low overheads, and monitoring applications that use NI to select good information can reduce their inaccuracy by nearly a factor of five.

Many Peer-to-peer (P2P) applications such as media broadcasting and content distribution require a high performance overlay structure in order to deliver satisfying quality of service (QoS). Previous approaches to building such overlays either involve a shared contact point, which results in non-scalable solutions, or rely on gossip style membership dissemination, which lacks QoS awareness. In this paper, we present a distributed membership service called RandPeer, which manages membership information on behalf of P2P applications, and allows peers to locate good neighbors based on their QoS characteristics. Using this service, P2P applications can easily construct their overlays in a scalable and QoS aware fashion. We have implemented RandPeer and experimented in both local and wide area environments. Our results show that (1) RandPeer is scalable and robust to highly dynamic P2P memberships; (2) RandPeer has good lookup performance, both in terms of response time and the randomness of peer selection. The latter improves load balancing and failure resilience of P2P applications; (3) when used to improve the performance of a mesh based P2P overlay, RandPeer achieves 10 % improvement in just 2 protocol rounds, which is more than 5 times faster than pure random neighbor selections. 1

We consider using traceroute-like end-to-end measurement to infer the underlay topology for a group of hosts. One major issue is the measurement cost. Given hosts in an asymmetric network without anonymous routers, traditionally full @ IA traceroutes are needed to determine the underlay topology. We investigate how to efficiently infer an underlay topology with low measurement cost, and propose a heuristic called Max-Delta. In the heuristic, a server selects appropriate host-pairs to measure in each iteration so as to reveal the most undiscovered information on the underlay. We further observe that the presence of anonymous routers significantly distorts and inflates the inferred topology. Previous research has shown that obtaining both exact and approximate topology in the presence of anonymous routers under certain consistency constraints is intractable. We hence propose fast algorithms on how to practically construct an approximate topology by relaxing some constraints. We investigate and compare two algorithms to merge anonymous routers. The first one uses Isomap to map routers into a multidimensional space and merges anonymous routers according to their interdistances. The second algorithm is based on neighbor router information, which trades off some accuracy with speed. We evaluate our inference algorithms on Internet-like and real Internet topologies. Our results show that almost full measurement is needed to fully discover the underlay topology. However, substantial reduction in measurements can be achieved if a little accuracy, say 5%, can be compromised. Moreover, our merging algorithms in the presence of anonymous routers can efficiently infer an underlay topology with good accuracy.

Detecting coordinated attacks on Internet resources requires a distributed network monitoring infrastructure. Such an infrastructure will have two logically distinct elements: distributed monitors that continuously collect traffic information, and a distributed query system that allows network operators to efficiently correlate information from different monitors in order to detect anomalous traffic patterns. In this paper, we discuss the design and implementation of MIND, a distributed index management system that supports the creation and querying of multiple distributed indices. We validate MIND using traffic traces from two large backbone networks, then examine the performance of a MIND prototype on more than 100 PlanetLab machines. Our experiments show that MIND can detect and report network anomalies in about one second on an inter-continental backbone. We also analyze the efficiency of our load balancing mechanism and evaluate the robustness of MIND to node failure. I.

We address the problem of providing scalable support for subscriptions with personalized value-based notification conditions in widearea publish/subscribe systems. Notification conditions can be finetuned by subscribers, allowing precise and flexible control of when events are delivered to the subscribers. For example, a user may specify that she should be notified if and only if the price of a particular stock moves outside a “radius ” around her last notified value. Naive techniques for handling notification conditions are not scalable. It is challenging to share subscription processing and notification dissemination of subscriptions with personalized valuebased notification conditions, because two subscriptions may see two completely different sequences of notifications even if they specify the same radius. We develop and experimentally evaluate scalable processing and dissemination techniques for these subscriptions. Our approach uses standard network substrates for notification dissemination, and avoids pushing complex application processing into the network. Compared with other alternatives, our approach generates orders of magnitude lower network traffic, and incurs lower server processing cost. 1

Abstract — Peer-to-peer (P2P) systems have been widely used for exchange of voluminous information and resources among thousands or even millions of users. Since shared data are normally identified by multiple attributes, a fundamental issue in P2P systems is to efficiently support complex queries on multidimensional data. Prior works suffer from some fundamental limitations, such as being constrained to support certain types of queries, excessive maintenance overheads, and etc. In this study, we propose a framework, called distributed peer tree (DPTree), which efficiently supports various types of queries on multidimensional data in P2P systems based on balanced tree indexes. DPTree achieves the efficiency through the following designs: 1) distributing the tree structure among peers in a way preserving the nice properties of balanced tree structures yet avoiding single points of failure and performance bottlenecks; 2) organizing peers into an overlay structure that enables efficient navigation yet is easy to maintain; 3) an efficient navigation algorithm; 4) an innovative wavelet-based load balancing mechanism. Through extensive performance evaluation, we verify the superiority of DPTree over existing works. I.

Abstract—Provisioning random access functions in peer-topeer on-demand video streaming is challenging, due to not only the asynchronous user interactivity but also the unpredictability of group dynamics. In this paper, we propose VMesh, a distributed peer-to-peer video-on-demand (VoD) streaming scheme which efficiently supports random seeking functionality. In VMesh, videos are divided into segments and stored at peers’ local storage in a distributed manner. An overlay mesh is built upon peers to support random forward/backward seek, pause and restart during playback. Our scheme takes advantage of the large aggregate storage capacity of peers to improve the segment supply so as to support efficient interactive commands in a scalable manner. Unlike previous work based on “cacheand-relay” mechanism, in our scheme, user interactivity such as random seeking performed by a peer does not break the connections between it and its children, and hence our scheme achieves better playback continuity. Through simulation, we show that our system achieves low startup and seeking latency under random user interactivity and peer join/leave which is a crucial requirement in an interactive VoD system. Index Terms—Peer-to-peer, media streaming, random seeking, distributed storage, locality-aware, popularity-based, distributed consensus. I.