Reliability of grid service systems

Abstract

Grid computing system is different from conventional distributed computing systems by its focus on large-scale resource sharing, where remote accessing and information communication have great influence on grid computing. As an important metric for services, the grid service reliability is studied in this paper. There are some earlier studies on the reliability analysis of conventional small-scale distributed systems, which ignored the communication time and processing time when studying the reliability, but it is not practical in the reliability analysis of grid services due to the wide-area property of the network. Thus, this paper presents a model for the grid services and develops an approach to analyze the grid service reliability. Detailed algorithms to compute the grid service reliability are presented and a self-sensing technology is proposed for parameterization and monitoring. A numerical example is used for the purpose of illustration. Finally, the new model is compared with different types of conventional models, through which this model is verified more suitable for grid service reliability. © 2006 Elsevier Ltd. All rights reserved.