Self-optimizing distributed trees (original) (raw)
Related papers
Distributed construction of resource-efficient overlay tree by approximating MST
2004 IEEE International Conference on Multimedia and Expo (ICME) (IEEE Cat. No.04TH8763), 2004
This paper presents a distributed protocol called RESMO for constructing overlay tree to support video streaming applications. RESMO reduces network resource usage by approximating minimum spanning tree and achieves low endto-end latency between the sender and each receiver at the same time. The resulting overlay is a compromise between overlay minimum spanning tree and shortest path tree. We evaluated the tree constructed by RESMO through simulations, and found that RESMO gives significant improvement over existing protocols in terms of link stress, relative delay penalty and resource usage.
Extensible optimization in overlay dissemination trees
Proceedings of the 2006 ACM SIGMOD international conference on Management of data - SIGMOD '06, 2006
We introduce XPORT, a profile-driven distributed data dissemination system that supports an extensible set of data types, profile types, and optimization metrics. XPORT efficiently implements a generic tree-based overlay network, which can be customized per application using a small number of methods that encapsulate application-specific data filtering, profile aggregation, and optimization logic. The clean separation between the "plumbing" and "application" enables the system to uniformly support disparate dissemination-based applications. We first provide an overview of the basic XPORT model and architecture. We then describe in detail an extensible optimization framework, based on a two-level aggregation model, that facilitates easy specification of a wide range of commonly used performance goals. We discuss distributed tree transformation protocols that allow XPORT to iteratively optimize its operation to achieve these goals under changing network and application conditions. Finally, we demonstrate the flexibility and the effectiveness of XPORT using real-world data and experimental results obtained from both prototype-based LAN emulation and deployment on PlanetLab.
Thicket: A Protocol for Building and Maintaining Multiple Trees in a P2P Overlay
One way to efficiently disseminate information in a P2P overlay is to rely on a spanning tree. However, in a tree, interior nodes support a much higher load than leaf nodes. Also, the failure of a single node can break the tree, impairing the reliability of the dissemination protocol. These problems can be addressed by using multiple trees, such that each node is interior in just a few trees and a leaf node in the remaining; the multiple trees approach allows to achieve load distribution and also to send redundant information for fault-tolerance. This paper proposes Thicket, a decentralized algorithm to efficiently build and maintain such multiple trees over a single unstructured overlay network. The algorithm has been implemented and is extensively evaluated using simulation in a P2P overlay with 10.000 nodes.
Failure-Tolerant Overlay Trees for Large-Scale Dynamic Networks
2008 Eighth International Conference on Peer-to-Peer Computing, 2008
Trees are fundamental structures for data dissemination in large-scale network scenarios. However, their inherent fragility has led researchers to rely on more redundant mesh topologies in the presence of churn or other highly dynamic settings. In this paper, instead, we outline a novel protocol that directly and efficiently maintains a tree overlay in the presence of churn. It simultaneously achieves other beneficial properties such as limiting the maximum node degree, minimizing the extent of the tree topology changes resulting from failures, and limiting the number of nodes affected by each topology change. Applicability to a range of distributed applications is discussed and results are evaluated through extensive simulation and a PlanetLab deployment.
OBST: A self-adjusting peer-to-peer overlay based on multiple BSTs
IEEE P2P 2013 Proceedings, 2013
The design of scalable and robust overlay topologies has been a main research subject since the very origins of peerto-peer (p2p) computing. Today, the corresponding optimization tradeoffs are fairly well-understood, at least in the static case and from a worst-case perspective.
PLATON: Peer-to-Peer load adjusting tree overlay networks
Peer-to-Peer Networking and Applications, 2011
Peer-to-Peer systems supporting multi attribute and range queries use a number of techniques to partition the multi dimensional data space among participating peers. Load-balancing of data accross peer partitions is necessary in order to avoid the presence of network hotspots which may cause performance degradation or failures within the distributed environment. In this paper, we introduce a novel framework, PLATON, that preserves load balancing accross peer partitions when the multi-dimensional data space is dynamic, without requiring up-to-date global load information, e.g. information about the most loaded or least loaded peers in the network. A theoretical analysis on the upper bounds (ie. worst case) of the proposed The work done by Chariklis was carried out while he was final year student at NTUA. L. Lymberopoulos (B) · M. Grammatikou · S. Papavassiliou · V. Maglaris algorithm is presented; its performance is evaluated in large-scale simulated networks and validated within in the PlanetLab emulation platform.
A Topologically-Aware Overlay Tree for Efficient and Low-Latency Media Streaming
Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2009
Streaming a live music concert over the Internet is a challenging task as it requires real-time, high-quality data delivery over a large number of geographically distributed nodes. In this paper we propose MusiCast, a real-time peer-to-peer multicast system for streaming midi events and compressed audio data. We present a scalable and distributed tree construction algorithm where nodes across the Internet self-organize into a low-latency tree. Our system is built ontop of the pastry DHT and takes advantage of the DHT's properties to construct an end-toend low-latency dissemination tree using topology oriented information. The benefit of our scheme is that it is completely decentralized, allowing nodes to connect to each other using local information only, and achieves good performance by considering latency information when constructing the tree. Our experimental results illustrate the benefits of our approach.
X-BOT: A Protocol for Resilient Optimization of Unstructured Overlays
Gossip, or epidemic, protocols have emerged as a highly scalable and resilient approach to implement several application level services such as reliable multicast, data aggregation, publish-subscribe, among others. All these protocols organize nodes in an unstructured random overlay network. In many cases, it is interesting to bias the random overlay in order to optimize some efficiency criteria, for instance, to reduce the stretch of the overlay routing. In this paper we propose X-Bot, a new protocol that allows to bias the topology of an unstructured gossip overlay network. X-Bot is completely decentralized and, unlike previous approaches, preserves several key properties of the original (non-biased) overlay (most notably, the node degree and consequently, the overlay connectivity). Experimental results show that X-Bot can generate more efficient overlays than previous approaches.
Computer Networks, 2011
Even though Application Level Multicast (ALM) protocols are proposed as an efficient alternative to the classical IP multicast model, they still suffer from the same rearrangement overhead problem. Indeed, when a member joins or leaves the multicast session, a part of the overlay tree should be rearranged. This rearrangement overhead can significantly disturb the fluidity of the application data stream received by other members. In addition, the efficiently of the constructed overlay tree depends entirely on the honesty and the cooperation of all participating members. However such behavior cannot be guaranteed and some selfish and non-cooperative nodes may take profit from the honesty of other members in the overlay. In this paper we present a new technique, called Membership Duration Aware ALM (MDA technique), that takes advantage of the announced membership duration of each new joining member in order to construct a more stable and robust overlay trees. Our technique can be used together with any existing centralized or distributed ALM protocol to reduce the overall rearrangement overhead. Simulation results confirm the efficiency of our technique when applied to both centralized and distributed ALM protocol. Then, we analyze in depth the impact of cheating nodes on the performances and the stability of overlay trees constructed by our MDA-ALM protocol.
D2-Tree: A New Overlay with Deterministic Bounds
Lecture Notes in Computer Science, 2010
We present a new overlay, called the Deterministic Decentralized tree (D 2 -tree). The D 2 -tree compares favourably to other overlays for the following reasons: (a) it provides matching and better complexities, which are deterministic for the supported operations; (b) the management of nodes (peers) and elements are completely decoupled from each other; and (c) an efficient deterministic load-balancing mechanism is presented for the uniform distribution of elements into nodes, while at the same time probabilistic optimal bounds are provided for the congestion of operations at the nodes. The load-balancing scheme of elements into nodes is deterministic and general enough to be applied to other hierarchical tree-based overlays. This load-balancing mechanism is based on an innovative lazy weight-balancing mechanism, which is interesting in its own right.