A novel monitoring mechanism by event trigger for Hadoop system performance analysis. (3rd October 2013)

• Record Type: Journal Article
• Authors: Liao, Chih‐Shan; Chuang, Chen‐Po; Chang, Ruay‐Shiung; Ranjan, Rajiv; Buyya, Rajkumar; Leitner, Philipp; Haller, Armin; Tai, Stefan
• Is Part Of: Software, practice & experience. Volume 44:Number 7(2014)
• Page Start: 823
• Page End: 834
• Publication Date: 2013-10-03
• Subjects: Computer software -- Periodicals; Computer programming -- Periodicals; Computer programs -- Periodicals; 005.3
• DOI: 10.1002/spe.2230
• Languages: English
• ISSNs: 0038-0644

Abstract

<abstract abstract-type="main" id="spe2230-abs-0001"> <title>SUMMARY</title> <p id="spe2230-para-0001">Cloud computing has become more and more popular nowadays since it possesses various conveniences such as pay as you go, dynamically deployment, and high efficiency. A numbers of cloud computing platforms are now in the market. Apache Hadoop, an open‐source platform, is one of these platforms which has been widely used by companies and organizations, because it is free for both research and production. From the perspective of service providers, it is better to see infrastructure running at the lowest possible power consumption and in the same time satisfying the QoS. From the viewpoint of the service clients, it is also the best situation that all functions they desire are running well at the lowest service cost. By putting ourselves in their shoes, in this paper, we propose the event trigger, which provides an automatic event recording mechanism. The event trigger applies the recorded parameters onto an easy‐understanding visual graph. By doing so, users will be able to adjust their deployment of hardware resource through a simple and straightforward understanding of their system performance records. Copyright © 2013 John Wiley &amp; Sons, Ltd.</p> </abstract>