• /  12
  • 下载费用: 19.90积分  

Using Data Stream Management Systems for Traffic Analysis

'Using Data Stream Management Systems for Traffic Analysis'
Using Data Stream Management Systems for Traffic Analysis – A Case Study – Thomas Plagemann2, 1, Vera Goebel2, 1, Andrea Bergamini1, Giacomo Tolu1, Guillaume Urvoy-Keller1, Ernst W. Biersack1 1Institut Eurecom, Corporate Communications, 2229 Route des Crêtes BP 193 F-06904 Sophia Antipolis Cedex, France {bergamin, tolu, urvoy, erbi}@eurecom.fr 2University of Oslo, Department of Informatics, Postbox 1080 Blindern, 0316 Oslo, Norway {plageman, goebel}@ifi.uio.no Abstract. Many traffic analysis tasks are solved with tools that are developed in an ad-hoc, incremental, and cumbersome way instead of seeking systematic solutions that are easy to reuse and understand. The huge amount of data that has to be managed and analyzed together with the fact that many different analysis tasks are performed over a small set of different network trace formats, motivates us to study whether Data Stream Management Systems (DSMSs) might be useful to develop traffic analysis tools. We have performed an ex- perimental study to analyze the advantages and limitations of using DSMS in practice. We study how simple and complex analysis tasks can be solved with TelegraphCQ, a public domain DSMS, and present a preliminary performance analysis. 1 Introduction and Motivation The number of tools for analyzing data traffic in the Internet is continuously increas- ing, because there is an increasing need in many different application domains. For example, network operators and service providers need to monitor data traffic to analyze the provided service level, to identify bottlenecks, and initiate appropriate counter measures, if possible. This is especially important in the Internet, because the amount of data traffic continuously increases and the behavior and requirements of end-users are changing over time, like accepted response time from web servers. Another application domain is the development and improvement of new protocols and applications, like overlay networks and peer-to-peer (P2P) file sharing applica- tions. The complexity of these protocols and applications, as well as the complexity of the environment they are used in, often impose that a meaningful analysis can only be done during their operation in the Internet. The typical coarse grain architecture of these tools consist of two components, first, a packet capturing or flow statistic com- ponent like TCPdump or NetFlow, and second, an analysis component to examine the resulting traces and draw certain conclusions. Performing traffic analysis to gain new knowledge is normally an iterative process. Traffic analysis tools are used to get a better understanding of network dynamics, protocol behavior, etc. Based on these new insights and influenced by changes in the Internet, e.g., traffic mix and end-user behavior, new analysis goals are defined for the next iteration step. For example, in our recent BitTorrent work, we measured the average throughput of leechers (i.e., clients that have not completed the down
关 键 词:
UsingDataStreamManagementSystemsforTrafficAnalysis
 天天文库所有资源均是用户自行上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作他用。
关于本文
本文标题:Using Data Stream Management Systems for Traffic Analysis
链接地址: https://www.wenku365.com/p-40984894.html
关于我们 - 网站声明 - 网站地图 - 资源地图 - 友情链接 - 网站客服点击这里,给天天文库发消息,QQ:1290478887 - 联系我们

本站为“文档C2C交易模式”,即用户上传的文档直接卖给(下载)用户,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有【成交的100%(原创)】。本站是网络服务平台方,若您的权利被侵害,侵权客服QQ:1290478887 欢迎举报。

1290478887@qq.com 2017-2027 https://www.wenku365.com 网站版权所有

粤ICP备19057495号 

收起
展开