Multi-Objective Optimization for Workflow Scheduling Under Task Selection Policies in Clouds

Abstract

Cloud computing provides infrastructure for executing workflows that require high processing and storage capacity. Although there are several algorithms for scheduling workflows, few consider security criterion. Algorithms that cover security usually optimize either cost or makespan. However, there are cases where the user would like to choose or evaluate among different solutions that present a trade-off between monetary cost and execution time (makespan) of the workflow. The selection of the tasks, which involve confidential/sensitive data, has to prioritize the safe execution of the workflow. In this paper, we propose a multi-objective optimization for scheduling of workflow tasks in cloud environments by considering cost and makespan under different task selection policies. Extensive experiments in real-world workflows with different policies show that our approach returns several solutions in the Pareto frontier for both cost and makespan. The results revealed a reasonable ability to find Pareto frontiers during the optimization process.