Scheduling Delay-Sensitive Bag-of-Task Jobs with Imprecise Computations on Distributed Resources

The evolution of the distributed computing paradigm had as a result new computing models such as grid and cloud computing. Furthermore, in these environments it is common to run complex parallel applications thus making the employment of appropriate scheduling strategies of paramount importance. This research examines the scheduling of multiple-task applications of bag-of-tasks type that have different characteristics, utilizing imprecise computations. Two scheduling schemes are utilized in a distributed system. Their performance is evaluated and compared using simulation. The experimental results demonstrate that the proposed scheduling technique that considers the delay sensitivity of the applications and allows partial task executions, can provide significant performance improvement at the expense of a low decrease in the results precision.