Resource reconfiguration policy based on Temporal quorum status estimation scheme in computational grids

Abstract

Computational Grids are emerging as a new computing paradigm for solving grand challenge application in science, engineering, and economics. Grid development involves the efficient management of heterogeneous, geographically distributed, and dynamically available resources. Intrinsic properties of the Computational and Data Grid can be defined as dynamicity and unpredictability. Those properties lead the degradation of the performance and failure of system. Quorum based resource allocation scheme is one of the prominent management policies for reducing complexity and increasing reliability. Resource reconfiguration means that reshuffling the current quorum resource set for maintaining the quality level of the resources. In this paper, we propose the resource reconfiguration algorithm based on temporal quorum status estimation method. Algorithm has two phases. One is the status estimation using temporal execution time deviation. The other is resource reconfiguration based on estimated status. After triggering based on estimation, reconfiguration algorithm tries to optimize the current resource allocation status by reshuffling the selected quorum and re-allocating resource into the application. Our approach shows the increase of the stability of execution environment as well as decrease the completion time compare to the method that is not adapted the resource reconfiguration.