Secure Selection of Multiple Resources Based on Virtual Private Network for Computational Grids
G. Kavitha and V. Sankaranarayanan
DOI : 10.3844/jcssp.2011.1881.1887
Journal of Computer Science
Volume 7, Issue 12
Problem statement: Grid computing provides a virtual framework for controlled sharing of resources across institutional boundaries. In computational grids, a client application is executed on the available set of resources that satisfy the user QoS requirements. Some applications require exhaustive computation power for execution of its tasks. In general, these tasks are assigned to a single available resource on the grid that has the required computation power. Therefore, the client application waits indefinitely until a suitable resource is found. Approach: In this study a novel multiple resource selection strategy is presented, which selects multiple resources based on trust and QoS parameters and the tasks are mapped to the appropriate resources for parallel execution. Selection of resources is based on the trust value of the resource, the available computation power at the time of job submission, the speed of the connectivity link, the time deadline and the budget constraints. The proposed method performs task grouping and selects the optimum number of resources for task execution. The tasks are executed in parallel among the multiple resources and the results are aggregated and transferred to the client within the specified time deadline. Security for the user tasks is strengthened by creating a Virtual Private Network (VPN) to the selected resources and tasks are further mapped to the resources through the secured VPN channel. Results: Simulations results show that there is a significant improvement in the overall resource utilization of the grid with a high success rate of jobs and reduction in the total execution time of submitted jobs. Conclusion: The tasks are scheduled to available multiple resources with VPN security. As optimum number of resources is selected for parallel execution, the resources are utilized to a maximum and there is a reduction in the percentage of pending jobs on the grid.
© 2011 G. Kavitha and V. Sankaranarayanan. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.