Anzeige
Erschienen in:
2016 | OriginalPaper | Buchkapitel
Eavesdropper: A Framework for Detecting the Location of the Processed Result in Hadoop
verfasst von : Chuntao Dong, Qingni Shen, Wenting Li, Yahui Yang, Zhonghai Wu, Xiang Wan
Erschienen in: Information and Communications Security
Verlag: Springer International Publishing
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
Hadoop has become increasingly popular as it rapidly processes big data in parallel, while security mechanisms have been introduced or studied for Hadoop. In addition, other security issues that should not be neglected still exist. Data leakage is one of the major security challenges. This paper studies the vulnerability of authorization mechanism of services in Hadoop and the threat of information leakage. Some authorization mechanism allow all users to access services by default, an adversary can utilize these services to collect information of other users. We design and implement Eavesdropper, a framework which utilizes k-means clustering to address the nodes that store the processed results. We conduct a comprehensive of experiments, which clearly demonstrate that our detection framework is capable of detecting the nodes that store the results.