Rejuvenation effects on the grid environment performance with response time delays using Monte Carlo simulation |
| |
Affiliation: | 1. School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, Sichuan, PR China;2. School of Aerospace, Mechanical and Manufacturing Engineering, RMIT University, Victoria 3083, Australia;3. Institute of Electronics, Chinese Academy of Sciences.;1. School of Software Engineering, Chengdu University of Information Technology, No.24 Block 1, Xuefu Road, Chengdu 610225, People''s Republic of China;2. School of Computing & Engineering, University of Huddersfield, Queensgate HD1 3DH, Huddersfield, UK;3. School of Computer Science, Chengdu University of Information Technology, No.24 Block 1, Xuefu Road, Chengdu 610225, People''s Republic of China;4. School of Optical Science and Engineering, Zhejiang University, Hangzhou 310027, People''s Republic of China;1. Dept. of Electrical Engineering, University of Taif, Al-Haweiah, 21974, Saudi Arabia;2. Samovar CNRS, Telecom SudParis, Institut Polytechnique de Paris, Evry Cedex 91011, France |
| |
Abstract: | Grid computing is an evolutionary technological achievement that takes advantages of wide area communication and large scale resource sharing. Although there are several advantages, grid does not guaranty stableness of resources due to their resources’ nature, diverse failures and error conditions that may appear. Consequently, grid dependability issues arise. In this paper, a grid computing environment with star topology and direct access to resources is considered, which consists of a Resource Management System (RMS) and distributed Root Nodes (RNs). The distributed RNs are considered either as operational or as failed, though the RMS is susceptible to resource exhaustion, which can lead to software failures. The response waiting time from a distributed RN is also taken under consideration. A software rejuvenation to counteract RMS resource exhaustion is adopted, and its effect on grid environment performance is studied extensively. A non-Markovian approach is considered to model the system’s evolution in time and to evaluate the proposed performance measures. Due to the complicated structure of the system, analytic formulas for the proposed measures are not available, thus Monte Carlo simulation methods are employed. |
| |
Keywords: | Grid environment with star topology Software rejuvenation Performance indicator Expected accumulated cost Monte Carlo simulation Estimation |
本文献已被 ScienceDirect 等数据库收录! |
|