An Adaptive Scheduling Mechanism for Computational Desktop Grid Using GridGain

Abstract

A computational grid essentially represents a dynamic and distributed environment. Unlike, a tightlycoupled parallel computing environment, high performance computing on a grid is a complex environment because of the heterogeneous computational performance of the nodes, possible unavailability of nodes, unpredictable node behavior, and unreliable network connectivity. Therefore, in a computational grid, scheduling of jobs is an exigent proposition considering its dynamic nature. In comparison to static scheduling, an adaptive scheduling mechanism is more attractive in a grid-computing environment, owing to its capability to adjust the scheduling policy according to the dynamically changing computational environment. This paper presents an adaptive scheduling method, which can be used for parallel applications whose total workload is unknown a priori. This method can deal with the unpredictable execution conditions commonly encountered on grids. To address this scheduling problem, parameters which quantify the dynamic nature of the execution conditions had to be defined. The proposed scheduling method is based on an online algorithm which is to be adaptable to the varying execution conditions.