DOCSLIB.ORG

An Empirical Study of Seeders in Bittorrent

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
An Empirical Study of Seeders in Bittorrent An Empirical Study of Seeders in BitTorrent Justin Bieber, Michael Kenney, Nick Torre, and Landon P. Cox Duke University, Durham, NC Abstract a wide range of content to nearly 4 million users at any moment. BitTorrent has attracted attention from researchers As was the case in Izal's study, we collected data by and the press for its wide deployment and explicit goal passively monitoring BitTorrent use through the trackers of eliminating free-riding, but many of the most impor- that coordinate peers. This prevented us from directly tant peers in BitTorrent operate outside of its incentive observing how users respond to BitTorrent's incentive mechanisms. Altruistic seeders help bootstrap new peers mechanisms, but provided us with the number of seed- and provide a significant fraction of the global upload ers (peers who are uploading, but not downloading) and bandwidth. We have taken an empirical approach to un- leechers (peers who are downloading). derstanding seeders by studying 35 BitTorrent sites with Seeders are particularly important for bootstrapping nearly four million users at any moment over several new peers and to network performance. Many sites ex- weeks. plicitly encourage peers to continue uploading data even Our study focuses on two aspects of seeders. First, we after their download has finished. These reminders are looked at the relationship between the number of seeders necessary because seeding is altruistic rather than self- and bandwidth utilization. A case study of a Linux dis- interested; BitTorrent's incentive mechanisms cannot re- tribution network showed that as bandwidth utilization ward seeders for the upload bandwidth they contribute. increased, the rate of seeding decreased. Second, we This makes seeders difficult to model game theoretically looked at the relationship between site attributes and the and motivated our empirical study. number of seeders. A survey of 34 BitTorrent sites over We have focused on two aspects of seeders. First, we two weeks found that the presence of niche-content (e.g. examined the relationship between bandwidth utilization only anime, hip-hop, or Linux files), merchandise for and the number of seeders. A case study of a BitTor- sale (e.g. t-shirts with the site URL), and negative rein- rent site over 23 days showed that the ratio of seeders- forcement (e.g. a posted list of the 10 least contributing to-leechers decreased dramatically when bandwidth be- peers) correlated positively with the rate of seeding. came highly utilized. Second, we looked at the relation- ship between site attributes and altruism. We were in- 1 Introduction terested to know which, if any, site characteristics cor- relate with a high seeders-to-leechers ratio. In a study Free-riding in file-sharing networks such as Gnutella, of 34 different sites, we found that networks with niche- Napster, and Kazaa is a well documented phe- content (e.g. only anime, hip-hop, or Linux files), mer- nomenon [1, 6]. Because of this, the BitTorrent [3] con- chandise for sale (e.g. t-shirts with the site URL), and tent distribution system made robustness to free-riding negative reinforcement (e.g. a posted list of the 10 least an explicit design goal [4]. BitTorrent's attention to in- contributing peers) exhibited greater altruism than those centives as well as its responsibility for 35% of Internet without these features. traffic [11] has generated interest from both the research community [5] and popular press [12]. However, despite 2 Background and Motivation this attention, there have been no broad empirical studies of BitTorrent. BitTorrent [3] is a popular peer-to-peer file sharing Several papers have analyzed the incentive mecha- application that uses a “tit-for-tat” protocol [2] to trans- nisms proposed by BitTorrent [7, 9], but we are aware fer files between peers. BitTorrent files are broken into of only one empirical study [8], by Izal. This is a use- smaller fragments, which can be downloaded in paral- ful start, but because it focuses on a single shared file lel from multiple peers. Although not a “pure” tit-for-tat the scope of its observations are limited. We hope to protocol [9], BitTorrent is designed to reward peers for broaden the understanding of BitTorrent by presenting uploading fragments to others with proportionally fast results from several weeks of logging 35 BitTorrent sites downloads. during February and March of 2006. These sites serve 2.5 Seeders Leechers o i t a 10000 R 2 s r 8000 e h c 1.5 e e s 6000 r L - e o e 1 t - P 4000 s r e 0.5 2000 d e e 0 S 0 1 49 97 145 193 241 289 337 385 433 481 529 Pre-Offline Offline Post-Offline Hour Figure 2: The Effect of LMP Server Downtime Figure 1: Hourly Behavior at elm-project.org 3 Study Results To download a file, peers must first have an associated .torrent file. .torrent files are usually obtained We periodically logged the number of seeders and through BitTorrent web sites and contain important in- leechers within the swarms of 35 trackers over sev- formation about the download target, such as its length, eral weeks in February and March of 2006. Each en- its name, fragments' hashing information and the URL try was an aggregation across all swarms connected to of a tracker. The tracker is a server that maintains infor- each tracker. We obtained this information by cooperat- mation about where fragments can be downloaded from. ing with tracker administrators as well as monitoring the A group of peers actively downloading and uploading statistics posted on BitTorrent web sites. Depending on fragments is a swarm. the tracker, we were able to log statistics on a daily or Using the .torrent file, clients can query the hourly basis. The first half of our study examines the re- tracker about the locations of their missing fragments lationship among available bandwidth, download rates, within the swarm. Clients request fragments directly and the composition of the swarm in a Linux distribution from other peers and can increase the priority of their network over three weeks. requests by offering to upload the fragments still needed All of the BitTorrent trackers we monitored had an as- by those peers. This creates an incentive for clients to sociated web site; tracker URLs within the .torrent download the rarest fragments first and increases overall files hosted by a web site referred to servers controlled fragment availability. by web site administrators. This close relationship al- Of course, when a client first enters the swarm, it has lowed us to associate web sites' attributes with the nothing to offer anyone else. Thus, clients must be able swarm statistics we logged. Some of the attributes we to download fragments “for free” from seeders. Seeders recorded include the nature of files’ content, whether already have complete files and expect nothing in return sites provided message boards, or if registration was for their uploading. required to access .torrent files. The second half Seeders are also critical for BitTorrent performance. of our study looks at the relationship between these at- Izal's study found that over five months, seeders ac- tributes and trackers' seeders-to-leechers ratios. counted for over two-thirds of the total upload band- width [8]. Their importance is reflected in the pejorative 3.1 Case Study: The Linux Mirror Project label, leechers, assigned to peers who are still download- ing files. Peers can only reach the status of seeder by up- One of the most interesting sites in our study was the loading without downloading. Importantly, leechers are Linux Mirror Project [10] (LMP). As its name suggests, not necessarily free-riders. Even peers that engage in tit- LMP uses BitTorrent to distribute Linux distributions, for-tat with the rest of the swarm are considered leechers kernels, and other open-source software. We recorded until they have obtained all file fragments. the number of seeders and leechers tracked by LMP ev- The role seeders play in bootstrapping BitTorrent and ery hour for 23 days. These numbers are in Figure 1. network performance makes understanding them an im- Under normal circumstances, LMP peers appear to be portant data point for content-distribution network de- extremely generous, exhibiting a seeders-to-leechers ra- signers and administrators. Game theory has proven use- tio of over two in steady state. By comparison, Izal ful for understanding leechers' behavior [4, 9], but offers observed a peak ratio of approximately .6 [8]. How- little hope of understanding seeders since altruism is not ever, between the 17th and 20th day of the trace, the easily described by these models. Because of this, we swarm size crashed before nearly tripling normal levels. hope that our empirical observations will provide insight More interesting, during this period the ratio of seeders- into how seeders behave. to-leechers approached one and even dipped below one 2500000 1000000 2000000 100000 s 1500000 r e e 10000 P 1000000 s r e 500000 h 1000 c e e 0 L Seeders Leechers 100 Figure 3: Aggregate Seeders and Leechers 10 1 toward the end of this growth. Around the 20th day, the 1 10 100 1000 10000 100000 1000000 network suddenly returned to normal. Seeders Understanding this behavior requires an explanation Figure 4: Average Daily Seeders and Leechers of LMP's architecture. LMP maintains three separate servers: one that acts as a web site and tracker and two that act as seeders. Early in the morning of the 17th day, of altruism increased as load shifted from LMP's servers the web site and tracker went offline because of a prob- onto the swarm. Persistent seeders may have reacted lem with LMP's ISP.
Load more

Recommended publications
  • Electronic Evidence in Torrent Copyright Cases
    ARTICLE: ELECTRONIC EVIDENCE IN TORRENT COPYRIGHT By Thomas M. Dunlap and CASES Nicholas A. Kurtz Introduction distribute copies of the registered copyrighted work.1 This article explores the methodology behind The Copyright Act further provides that anyone who obtaining electronic evidence of peer-to-peer users violates any of the exclusive rights of the copyright 2 and the use of that evidence in current copyright owner is an infringer of the copyright. infringement actions in United States Federal Courts. Pursuant to the Copyright Act, a plaintiff may elect Over the last year and a half, a number of lawsuits to recover statutory damages instead of actual have been filed on behalf of film copyright holders damages or the infringer’s profits. 17 U.S.C. § 504(c) seeking to enforce their rights against users of peer- provides: “the copyright owner may elect, at any time to-peer (P2P) torrent-type technologies used to before final judgment is rendered, to recover, instead unlawfully download and distribute films. of actual damages and profits, an award of statutory At the time of filing their complaints, the plaintiffs damages for all infringements involved in the action, have only been able to identify the Doe defendants by with respect to any one work ... in a sum of not less their internet Protocol (IP) address, the file hash, file than $750 or more than $30,000 as the court title, and the date and time of the alleged considers just.” Section 504(c) further provides that infringement. The only way that the plaintiffs can where an infringement is committed willfully, a court determine the defendants’ actual names is by has the discretion to increase the award of statutory 3 obtaining the information from the internet Service damages to $150,000.
    [Show full text]
  • Defense Against the Dark Arts of Copyright Trolling Matthew As G
    Loyola University Chicago, School of Law LAW eCommons Faculty Publications & Other Works 2018 Defense Against the Dark Arts of Copyright Trolling Matthew aS g Jake Haskell Follow this and additional works at: https://lawecommons.luc.edu/facpubs Part of the Civil Procedure Commons, and the Intellectual Property Law Commons Defense Against the Dark Arts of Copyright Trolling Matthew Sag &Jake Haskell * ABSTRACT: In this Article, we offer both a legal and a pragmaticframework for defending against copyright trolls. Lawsuits alleging online copyright infringement by John Doe defendants have accounted for roughly half of all copyright casesfiled in the United States over the past threeyears. In the typical case, the plaintiffs claims of infringement rely on a poorly substantiatedform pleading and are targeted indiscriminately at noninfringers as well as infringers. This practice is a subset of the broaderproblem of opportunistic litigation, but it persists due to certain unique features of copyright law and the technical complexity of Internet technology. The plaintiffs bringing these cases target hundreds or thousands of defendants nationwide and seek quick settlements pricedjust low enough that it is less expensive for the defendant to pay rather than to defend the claim, regardless of the claim's merits. We report new empirical data on the continued growth of this form of copyright trolling in the United States. We also undertake a detailed analysis of the legal andfactual underpinnings of these cases. Despite theirunderlying weakness, plaintiffs have exploited information asymmetries, the high cost of federal court litigation, and the extravagant threat of statutory damages for copyright infringement to leverage settlementsfrom the guilty and the innocent alike.
    [Show full text]
  • Charter Spectrum Notice of Copyright Infringement
    Charter Spectrum Notice Of Copyright Infringement Davis educating his chauffeuse nichers thrasonically, but quartered Stanfield never pulverising so loose. Judicial and apprehensible Kenneth braised consumptively and clutch his abstention intractably and tracklessly. Yehudi often discuss spinally when unattainted Foster melodized past and braked her alps. There must be of charter communications operating credit for the transaction in a home where they do next and choose what cox Or abduct any product identification proprietary copyright or other. This website is vast service of Charter Communications Inc andor its respective subsidiaries providing the. While many users panic when receiving infringement notices from their ISP in the majority of cases there cause no need never worry Stopping sharing the lump in longevity usually solves the few and tie no additional sharing takes place to further warnings should be received for legitimate content are least. Alleging copyright infringement against an unnamed defendant Defendant1. And record labels who want over 10000 copyrights is fair clear message to ISPs like. Some policy forms an exclusion for trademark infringement. Recently changed commerce, infringement notice of charter spectrum copyright. Organization violated the Digital Millennium Copyright Act hurt when. Elizabeth Hart et al v Charter Communications Inc et al No. How tough my ISP tell below I'm downloading copyrighted files. Infringement claims continue to it brought frequently in the communications. Time Warner Subpoena Archives Torrent Lawsuit Defense. How to Pirate Software Without been Caught Gizmodo. Looking to a lawsuit attorneys fees logically flow as of infringement, net primarily include bandwidth being accused by them? For Android devices both clients work just got but the notable speed difference between BitTorrent and uTorrent in male former's favor gives it intelligent edge.
    [Show full text]
  • Forcing the Net Through a Sieve: Why Copyright Filtering Is Not a Viable Solution for U.S
    Forcing the Net Through a Sieve: Why Copyright Filtering is Not a Viable Solution for U.S. ISPs Mehan Jayasuriya, Jef Pearlman, Robb Topolski, Michael Weinberg, Sherwin Siy www.PublicKnowledge.org i Acknowledgements The authors of this paper would like to thank the following advocates for their assistance and suggestions during the editing process: Public Knowledge president Gigi Sohn, Public Knowledge staff attorney Rashmi Rangnath, Public Knowledge legal consultant Adam Thomas, Public Knowledge intern Katy Tasker and Electronic Frontier Foundation staff technologist Peter Eckersley. ii Table of Contents Executive Summary ................................................................................................................. 1 1. Introduction...................................................................................................................... 2 2. Technological Analysis: The Anatomy of a Copyright Filter ............................................... 7 I. Determining Filter Policy ............................................................................................................................................. 8 A. Strict Policy Definition.....................................................................................................................................................8 B. Permissive Policy Definition..........................................................................................................................................9 II. Identifying Types of Content.....................................................................................................................................10
    [Show full text]
  • Peer-To-Peer File Sharing As User Rights Activism Michael A
    Western Journal of Legal Studies Volume 5 Article 3 Issue 3 State Encroachment on Personal Lives February 2015 Peer-to-Peer File Sharing as User Rights Activism Michael A. Gunn University of Western Ontario, [email protected] Follow this and additional works at: https://ir.lib.uwo.ca/uwojls Part of the Civil Law Commons, Civil Rights and Discrimination Commons, Common Law Commons, Communications Law Commons, Communication Technology and New Media Commons, Comparative and Foreign Law Commons, Computer Law Commons, Consumer Protection Law Commons, E-Commerce Commons, Economic History Commons, European Law Commons, Human Rights Law Commons, Intellectual Property Law Commons, International Law Commons, Internet Law Commons, Law and Economics Commons, Law and Philosophy Commons, Law and Society Commons, Legal History Commons, Science and Technology Law Commons, and the Science and Technology Policy Commons Recommended Citation Michael A. Gunn , "Peer-to-Peer File Sharing as User Rights Activism", (2015) 5:3 online: UWO J Leg Stud 3 <https://ir.lib.uwo.ca/ uwojls/vol5/iss3/3>. This Article is brought to you for free and open access by Scholarship@Western. It has been accepted for inclusion in Western Journal of Legal Studies by an authorized editor of Scholarship@Western. For more information, please contact [email protected], [email protected]. Peer-to-Peer File Sharing as User Rights Activism Abstract The pre-digital marketplace is no longer sustainable. With the imposition of digital rights management restrictions on the distribution of media, the Internet cannot promote intellectual freedom. Peer-to-peer file sharing technology helps expose the work of artists and authors to a much wider audience than previously possible.
    [Show full text]
  • Should Copyright Laws Be Able to Keep up with Online Piracy?
    SHOULD COPYRIGHT LAWS BE ABLE TO KEEP UP WITH ONLINE PIRACY? STEPHANIE MINNOCK INTRODUCTION: AN OVERVIEW OF THE LEGAL AND ILLEGAL ONLINE VIDEO MARKET ............................................................ 524 I. COPYRIGHT IN CYBERSPACE ............................................................. 527 A. Copyright Law: From Constitution to Cyberspace ............ 528 B. Jurisdiction over Internets: Long Arm Not Long Enough? ... 531 C. Domain Name Seizures: The Whack-A-Mole Effect .............. 533 II. COPYRIGHT CIRCUMVENTING TECHNOLOGY: WHY THERE IS AN ISSUE WITH THE INTERNET ......................................................... 535 A. Streaming Online Content ..................................................... 536 B. Virtual Private Networks: Tunnels to Liability-Free Infringement? ...................................................................... 538 C. Tor: Encrypting Copyright Infringer’s Actions? .................. 539 D. Seedboxes and Cyberlockers: An Illegal Downloader’s Dream? ............................................................................... 540 E. Cloud Technology: Unseizable Servers, Ever-Rotating Jurisdiction ......................................................................... 541 III. EFFECTIVE LAWS HAVE HIGH COSTS ............................................. 542 A. Increasing Courts’ Authority Regarding Copyright Infringement is Not Ideal .................................................... 543 B. Existing Technology Could Circumvent any Reasonable Legislation .........................................................................
    [Show full text]
  • Identifying First Seeders in Foxy Peer-To-Peer Networks Ricci Ieong, Pierre Lai, Kam-Pui Chow, Michael Kwan, Frank Law
    Identifying First Seeders in Foxy Peer-to-Peer Networks Ricci Ieong, Pierre Lai, Kam-Pui Chow, Michael Kwan, Frank Law To cite this version: Ricci Ieong, Pierre Lai, Kam-Pui Chow, Michael Kwan, Frank Law. Identifying First Seeders in Foxy Peer-to-Peer Networks. 6th IFIP WG 11.9 International Conference on Digital Forensics (DF), Jan 2010, Hong Kong, China. pp.151-168, 10.1007/978-3-642-15506-2_11. hal-01060615 HAL Id: hal-01060615 https://hal.inria.fr/hal-01060615 Submitted on 28 Nov 2017 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Distributed under a Creative Commons Attribution| 4.0 International License Chapter 11 IDENTIFYING FIRST SEEDERS IN FOXY PEER-TO-PEER NETWORKS Ricci Ieong, Pierre Lai, Kam-Pui Chow, Michael Kwan and Frank Law Abstract This paper describes a new approach for identifying first seeders in il- legal file sharing investigations involving Foxy, one of the most popular Chinese peer-to-peer networks. In identifying first seeders, the approach focuses on determining the slow-rising period of the cumulative seeder curve instead of merely measuring the number of seeders. The rela- tionships between file popularity, number of packets and the maximum upload limit during the time that the first seeder is connected to the network are also analyzed.
    [Show full text]
  • Convert Direct Http Link to Torrent
    Convert Direct Http Link To Torrent Incorruptible and Thomistic Timmy osculate her stymie foreshadow wavily or corraded repeatedly, is Engelbert aggravateddustier? Whitney optically kipper or carpenter unexceptionally shamefully if unpitying when Vale Rogers is primogenitary. admiring or apologizing. Outstretched Zachery If the torrentId is an httphttps link reading the torrent file then the request example the file can be. How to http link in python called to an offline as a free! Of your connected drives also supports conversion of media files. Line ticket a torrent file or torrent link either torrent or magnet link had an argument vlc videotorrent vlc httpexamplecomvideotorrent vlc magnetxturnbtih. Parse-torrent npm. Begin with https connection our files to. We hope to convert direct links into a converter in your payment data into flames. Maryann never finishes updating and convert direct download without having magnet link converter is converted to converts one file is stack exchange for converting your bittorrent tracker operator. Thanks a direct http link to convert torrent is structured and continue to? Online Magnet Link Generator. Once you locate a link and a torrent website a torrent client is used to actually download it. Your direct http direct link converter websites, convert large files? How can download http in google drive, https link for android without having the. Will convert http in our seedbox works as all qnap products purchased through this page as it directly from internet browser extensions that it back of. This key for beginners, https file converter is not download http. Linkedin and disconnects frequently used by archive can i tell you? Due to convert your behalf for converting your tv stick around this software installation videos files after you can also lacking many of.
    [Show full text]
  • Transmission Documentation Release 0.7
    Transmission Documentation Release 0.7 Erik Svensson October 11, 2011 CONTENTS i ii CHAPTER ONE INTRODUCTION This is transmissionrpc. This module helps using Python to connect to a Transmission JSON-RPC service. transmis- sionrpc is compatible with Transmission 1.3 and later. transmissionrpc is licensed under the MIT license. 1 Transmission Documentation, Release 0.7 2 Chapter 1. Introduction CHAPTER TWO GETTING STARTED Transmission is available at Python Package Index. To install the transmissionrpc python module use easy_install or pip. $ easy_install transmissionrpc Note: You might need administrator privileges to install python modules. You may also download the tarball from Python Package Index. Untar and install. $ tar -xzf transmissionrpc-0.7.tar.gz $ cd transmissionrpc-0.7 $ python setup.py install 2.1 Dependecies transmissionrpc has the following dependencies. • Python >= 2.5. • simplejson >= 1.7.1 or Python >= 2.6. If Python 2.6 or later is detected the standard library json implementation will be used. 2.2 Report a problem Problems with transmissionrpc should be reported through the issue tracker at bitbucket. Please look through the existing issues before opening a new issue. 3 Transmission Documentation, Release 0.7 4 Chapter 2. Getting started CHAPTER THREE INSTALLING FROM SOURCE 3.1 The source code Transmission is hosted at bitbucket using mercurial. To get a working copy, run $ hg clone http://www.bitbucket.org/blueluna/transmissionrpc/ The source code will be fetched and stored the directory transmissionrpc. Then install the module using $ python setup.py install Or if you wish to further develop transmissionrpc itself use $ python setup.py develop This will link this directory to the library as transmissionrpc.
    [Show full text]
  • Simulating Bittorrent
    Simulating BitTorrent Karel De Vogeleer David Erman Adrian Popescu Blekinge Institute of Blekinge Institute of Blekinge Institute of Technology Technology Technology [email protected] [email protected] [email protected] ABSTRACT server load as well as providing a simple way of locating in- P2P IP Television (IPTV) and other media distribution applica- formation. applications impose their own routing and tions are expected to be one of the next Internet killer appli- forwarding on top of the Internet infrastructure, in effect cations. One indication of this is the corporate backing that forming overlay networks. P2P the IP Multimedia Subsystem (IMS) is getting. However, the While applications are the de-facto most prominent bandwidth utilization of these applications is still an issue, contributors to Internet traffic today, there is another group as the volume of multimedia grows due to larger image res- of applications expected to challenge this situation. This olution and higher bitrate audio. One way of managing this group consists of IPTV, Voice over IP (VoIP), Video-on- increase in bandwidth requirements is to use existing end- Demand (VoD) and other multimedia distribution applica- host bandwidth to decrease the load on the content server tions. By the amount of attention placed upon it, large- in a Peer-to-Peer (P2P) fashion. One of the most successful scale multimedia streaming is expected to be the next killer P2P applications is BitTorrent (BT), a swarming file transfer application. Indications of this are the corporate backing IMS system. This paper presents an implementation of a BT sim- that, e. g. , the has achieved as well as the amount of ulator intended for future use in investigating modifications research put into mechanisms for compression, coding, dis- to the BT system to provide streaming media capabilities.
    [Show full text]
  • A Survey of Peer-To-Peer Storage Techniques for Distributed File Systems
    A Survey of Peer-to-Peer Storage Techniques for Distributed File Systems Ragib Hasan‡† Zahid Anwar‡† William Yurcik‡ Larry Brumbaugh‡ Roy Campbell† ‡National Center for Supercomputing Applications †Department of Computer Science University of Illinois at Urbana Champaign {rhasan,anwar, byurcik, ljbrumb}@ncsa.uiuc.edu, [email protected] Abstract Internet growth resulted in a new approach, the The popularity of distributed file systems continues building of distributed file system. As the host nodes to grow. Reasons they are preferred over traditional storing the shared objects became more geographically centralized file systems include fault tolerance, distributed and diverse, new criteria and performance availability, scalability and performance. In addition, constraints like availability, fault tolerance, security, Peer-to-Peer (P2P) system concepts and scalable robustness and location mechanisms became important functions are being incorporated into the domain of issues in designing distributed file systems. file systems. This survey paper explores the design paradigms and important issues that relate to such In recent years, P2P systems have emerged as a systems and discusses the various research activities in viable architecture for implementing distributed file the field of Distributed Peer- to-Peer file systems. systems. In a P2P network, end users share resources via direct exchange between computers. Information is 1. Introduction distributed among the member nodes instead of concentrated at a single server. A pure peer-to-peer In the recent years, Peer-to-Peer system research system is a distributed system without centralized has grown significantly. Using a large scale distributed control, where the software running at each node is network of machines has become an important element equivalent in functionality.
    [Show full text]
  • Involuntary Browser-Based Torrenting
    Involuntary Browser-Based Torrenting Alexander Bode [email protected] October 19, 2020 for abuse in the wild are presented in this paper. Additionally, proofs-of-concept for the usage, detection and prevention of Abstract—WebTorrent is the first torrent client to work in involuntary browser-based torrenting are presented in this a browser. The web technology uses WebRTC for its peer-to- paper. peer communications and is increasingly popular due to its variety of use cases. The rise in popularity has raised questions about WebTorrent’s potential for abuse. This study aims to RESEARCH QUESTIONS determine whether it is possible to use WebTorrent to have users The focus of this research is to determine whether it is involuntarily participate in peer-to-peer torrent networks. The WebTorrent API was analysed, and custom clients were built in possible to use WebTorrent for involuntary participation in order to determine if it is possible to run WebTorrent inconspic- peer-to-peer networks. The main research question is defined uously. The usefulness and attack vectors of involuntary browser- as the following: based torrenting have been analysed using a well-recognised security framework. Methods for the detection and prevention Can WebTorrent be abused to have web page visitors of WebTorrent usage have been investigated by inspecting the behaviour of the browser while using WebTorrent. A source- involuntarily participate in peer-to-peer networks? code search-engine was used to determine if this type of abuse is a widely used and established tactic. The results show that This question resulted in the following sub-questions: it is possible to use WebTorrent for involuntary browser-based • Which WebTorrent specific features can be abused? torrenting and that it can be detected and prevented in several • In which ways could WebTorrent be useful to an adver- ways.
    [Show full text]