Title: Fault tolerance in grid computing by resource clustering

Authors: Miloud Khaldi; Mohammed Rebbah; Boudjelal Meftah; Mohammed Debakla

Addresses: Computer Science Department, Faculty of Exact Sciences, University Mustapha Stambouli, Mascara, Algeria ' Computer Science Department, Faculty of Exact Sciences, University Mustapha Stambouli, Mascara, Algeria ' Computer Science Department, Faculty of Exact Sciences, University Mustapha Stambouli, Mascara, Algeria ' Computer Science Department, Faculty of Exact Sciences, University Mustapha Stambouli, Mascara, Algeria

Abstract: Computational grids have the potential for solving large-scale scientific problems using heterogeneous and geographically distributed resources. At this scale, the characteristics of dynamicity, resource heterogeneity and scalability have made fault tolerance more complex. In this paper, we propose FT-GRC a fault tolerance model that seeks to find the best substitute for the failed node by the clustering of the grid resources. This model is based on dynamic coloured graphs without replication of computer resources. The proposed fault tolerance mechanism uses scoring function to determine the appropriate substitute for each failed node by calculating the performance level of each node, and later exploits clustering to determine optimally the choice of substitute. Experimental results show the efficiency of the scoring method and the gain obtained by looking for the substitutes in the same cluster and then by the research for the nearest substitutes.

Keywords: grid computing; dynamic coloured graph; fault tolerance; clustering; scoring.

DOI: 10.1504/IJITST.2020.104577

International Journal of Internet Technology and Secured Transactions, 2020 Vol.10 No.1/2, pp.120 - 142

Received: 17 Feb 2018
Accepted: 16 May 2018
Published online: 21 Jan 2020 *

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