The Research Of Cost And Energy Aware Scientific Workflow Scheduling In Clouds

Cloud computing emerged as an attractive distributed computing paradigm can provide resources for various scientific workflow applications dynamically.Users have to care about the economical cost incurred by renting resources from cloud data centers,which charged by hourly-based pricing model.Moreover,the problem of energy consumption has become one of the major concerns in clouds.Although workflow scheduling problems have been extensively studied,few works take both cost and energy efficiency into account,especially in clouds.In this paper,we focus on how to reduce cost and energy consumption of secheduiling concurrently while meeting the particular deadline.First of all,this paper introduces the basic knowledge of cloud computing and workflow,describes the related concepts of workflow scheduling in clouds,and expounds the cloud workflow scheduling algorithms,classification,as well as the advantages and disadvantages of all kinds of algorithms.Then,we make the process of modeling workflow,cloud data center,energy,and on the basis of this,we present a cost and energy aware scheduling(CEAS)algorithm for scientific workflows.The proposed CEAS algorithm covers five sub-algorithms to optimize the cost and energy consumption step by step.Among the sub-algoriths,sequence task merging method and parallel task merging method can not only reduce the scheduling cost and energy,but also decrease the scheduling complexity of the workflow without affecting the precedence relationship of tasks.Further more,the concept of cost utility is proposed to map each task to an optimal VM according to assigned sub-makespan,and the VM reuse sub-algorithm is also applied to further reduce monetary cost and energy.After that,slack time reclamation for non-critical task by Dynamic Voltage and Frequency Scaling(DVFS)technique is utilized to save energy of leased virtual machine(VM)as long as the makespan of workflow is no more than the specified deadline.The time complexity of CEAS is polynomial time.Finally,the CEAS algorithm is evaluated using Cloudsim and various real-world scientific workflow applications,which demonstrates that the proposed algorithm performs better than the related state-of-the-art approaches for reducing cost and energy.