{"title":"Architecture of a p2p distributed adaptive directory","authors":"G. Cordasco, V. Scarano, Cristiano Vitolo","doi":"10.1145/1013367.1013436","DOIUrl":null,"url":null,"abstract":"Bookmarks are, nowadays, an important aid to navigation since they represent an easy way to reduce the cognitive load of managing and typing URLs. All the browsers have always provided, since the very beginning of the WWW, friendly ways of managing bookmarks. In this paper we deal with the problem of enriching this supportive framework for bookmarks (as provided by the browsers) by adding collaboration and (group) adaptation with a P2P system. In this paper, we describe a system that offers a distributed, cooperative and adaptive environment for bookmark sharing. DAD (Distributed Adaptive Directory) offers an adaptive environment since it provides suggestions about the navigation based on (a) the bookmarks, (b) the feedback implicitly provided by users and (c) the structure of the Web. DAD is fully scalable because of its peerto-peer architecture and provides, also, an infrastructure to build easily P2P overlay networks. Our system is structured on three separated layers (see Fig. 1): on the bottom we have an overlay network named CHILD (CHord with Identification and Lonely Discovery). Because of scalability and fault tolerance, our architecture is a pure Peer to Peer system: we realized a Distributed Hash Table (DHT) to store information about bookmarks. The DHT is based on the well-known Chord [5] protocol, which we extended with a boot mechanism through caching (i.e., no server is required), a mechanism that allows registration and authentication on the system and a tool that allows to save and reload the DHT data after the system shuts itself down or is started again. We would like to emphasize that the infrastructure provided by CHILD can be used as a base to develop a DHT for any pure P2P system. In fact, CHILD is implemented by a Java package that allows to build and manage any generic DHT, acting as middleware platform for P2P systems. The programmer can, therefore, implement the business logic of his application on top of our layer. CHILD will be soon freely available under the GNU public license at the project home page (http://isis.dia.unisa.it/projects/DAD).","PeriodicalId":409891,"journal":{"name":"WWW Alt. '04","volume":"21 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2004-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"8","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"WWW Alt. '04","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/1013367.1013436","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 8
Abstract
Bookmarks are, nowadays, an important aid to navigation since they represent an easy way to reduce the cognitive load of managing and typing URLs. All the browsers have always provided, since the very beginning of the WWW, friendly ways of managing bookmarks. In this paper we deal with the problem of enriching this supportive framework for bookmarks (as provided by the browsers) by adding collaboration and (group) adaptation with a P2P system. In this paper, we describe a system that offers a distributed, cooperative and adaptive environment for bookmark sharing. DAD (Distributed Adaptive Directory) offers an adaptive environment since it provides suggestions about the navigation based on (a) the bookmarks, (b) the feedback implicitly provided by users and (c) the structure of the Web. DAD is fully scalable because of its peerto-peer architecture and provides, also, an infrastructure to build easily P2P overlay networks. Our system is structured on three separated layers (see Fig. 1): on the bottom we have an overlay network named CHILD (CHord with Identification and Lonely Discovery). Because of scalability and fault tolerance, our architecture is a pure Peer to Peer system: we realized a Distributed Hash Table (DHT) to store information about bookmarks. The DHT is based on the well-known Chord [5] protocol, which we extended with a boot mechanism through caching (i.e., no server is required), a mechanism that allows registration and authentication on the system and a tool that allows to save and reload the DHT data after the system shuts itself down or is started again. We would like to emphasize that the infrastructure provided by CHILD can be used as a base to develop a DHT for any pure P2P system. In fact, CHILD is implemented by a Java package that allows to build and manage any generic DHT, acting as middleware platform for P2P systems. The programmer can, therefore, implement the business logic of his application on top of our layer. CHILD will be soon freely available under the GNU public license at the project home page (http://isis.dia.unisa.it/projects/DAD).
如今,书签是导航的重要辅助工具,因为它们代表了一种减少管理和输入url的认知负荷的简单方法。自从万维网诞生以来,所有的浏览器都提供了友好的管理书签的方式。在本文中,我们通过在P2P系统中添加协作和(组)适应来丰富书签支持框架(由浏览器提供)的问题。本文描述了一个为书签共享提供分布式、协作和自适应环境的系统。DAD(分布式自适应目录)提供了一个自适应环境,因为它提供了基于(a)书签、(b)用户隐式提供的反馈和(c) Web结构的导航建议。由于其点对点架构,DAD是完全可扩展的,并且还提供了一个基础设施来轻松构建P2P覆盖网络。我们的系统由三个分离的层构成(见图1):在底层,我们有一个名为CHILD (CHord with Identification and Lonely Discovery)的覆盖网络。由于可伸缩性和容错性,我们的体系结构是一个纯粹的点对点系统:我们实现了分布式哈希表(DHT)来存储关于书签的信息。DHT基于众所周知的Chord[5]协议,我们通过缓存(即不需要服务器)扩展了该协议的引导机制,该机制允许在系统上注册和身份验证,并提供了一个工具,允许在系统关闭或重新启动后保存和重新加载DHT数据。我们想强调的是,CHILD提供的基础设施可以作为开发任何纯P2P系统的DHT的基础。实际上,CHILD是由一个Java包实现的,它允许构建和管理任何通用的DHT,充当P2P系统的中间件平台。因此,程序员可以在我们的层之上实现他的应用程序的业务逻辑。CHILD将很快在GNU公共许可下在项目主页(http://isis.dia.unisa.it/projects/DAD)上免费提供。