Execution cost minimization scheduling algorithms for deadline-constrained parallel applications on heterogeneous clouds

摘要

The problem of minimizing the execution monetary cost of applications on cloud computing platforms has been studied recently, and satisfying the deadline constraint of an application is one of the most important quality of service requirements. Previous method of minimizing the execution monetary cost of deadline-constrained applications was the "upward" approach (i.e., fromexittoentrytasks) rather than combining the "upward" and "downward" approaches. In this study, we propose monetary cost optimization algorithm (DCO/DUCO) by employing "downward" and "upward" approaches together to solve the problem of execution cost minimization. "Downward" cost optimization is implemented by introducing the concept of the variable deadline-span and transferring the deadline of an application to each task. On the basis of DCO, the slack time is utilized to implement "upward" cost optimization without violating the precedence constraints among tasks and the deadline constraint of the application. Experimental results illustrate that the proposed approach is more effective than the existing method under various conditions.