Abstract :
A desktop grid, which is a computing grid composed of idle computing resources in a large network of desktop computers, is a promising platform for compute-intensive distributed computing applications. However, due to volatility of computing resources, effective scheduling for reliable execution of parallel computing applications on such a platform is a difficult problem. This paper proposes a new scheduling method aimed at reducing cases of task suspension and failure for more reliable execution of tasks as well as improving the total execution time of a parallel application on a desktop grid. The proposed method is based on utilizing the histories of execution behavior of individual computing nodes in the scheduling algorithm. In order to test out the feasibility of this idea, execution trace data were collected from several desktops and workstations. Then, based on this data, the execution of parallel applications consisting of independent tasks was simulated using trace-driven simulation. The simulation results showed that the proposed method reduced instances of application suspension and failure significantly when compared to FCFS by 52% and 78% on average, respectively. In addition, the total execution time of the target applications was also improved in most simulations when compared to previous desktop grid scheduling methods.
Keywords :
grid computing; parallel processing; scheduling; task analysis; compute intensive distributed computing; desktop computer; effective scheduling method; execution trace data; grid computing; idle computing resource; parallel computing; reliable execution; task suspension; trace driven simulation; desktop grid; task failure; task scheduling; task suspension;
