Title: STSHC: secure and trusted scheme for Hadoop cluster

Authors: Quan Zhou; Daixian Wu; Chunming Tang; Chunming Rong

Addresses: Key Laboratory of Mathematics and Interdisciplinary Sciences of Guangdong, Higher Education Institutes, Guangzhou University, 510006, Guangzhou, China ' College of Computer and Information Science and College of Software, Southwest University, 400715, Chongqing, China ' School of Mathematics and Information Science, Guangzhou University, 510006, Guangzhou, China ' Department of Science and Technology, University of Stavanger, 4036, Stavanger, Norway

Abstract: Hadoop is a distributed big data storage and processing framework hugely adopted in different sectors from online media, education, government and social media to handle the enormous growth of information in their respective domains. However, the core architecture of the solution is based on a trusted cluster. It lacks native methods for protecting sensitive data that cross over enterprises and are exposed or accessed illegally. In the paper, this work aims to construct a secure and trusted scheme for Hadoop cluster (STSHC). Firstly, we review the architecture of Hadoop cluster, and discuss the security risks and security requirements. And then we propose a secure and trusted architecture framework for Hadoop cluster with trusted computing base (TCB). Moreover, some protocols, which are the key management, the node trusted management protocol, the client access management protocol, the node access management protocol and job executing management protocol, are designed in detail. Finally, the security analysis shows that the scheme can efficiently meet the security requirement of Hadoop cluster.

Keywords: trusted computing base; Hadoop cluster; security requirements; TCB; security risks; big data security.

DOI: 10.1504/IJHPSA.2014.059874

International Journal of High Performance Systems Architecture, 2014 Vol.5 No.1, pp.63 - 69

Received: 05 Oct 2013
Accepted: 17 Nov 2013

Published online: 12 Jul 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article