The Research On Workflow Scheduling In Cloud Environment

In recent years, cloud computing technology has made great progress. By usingvirtualization technology, it can delivery various of IT resources to users in the formof service via the Internet. Like water and electricity, computing resources, storageresources, software resources and all kinds of application service can be easily used.Users use a "pay as you go" model to consumer services. With the cloud computingtechnology becoming mature and perfect, many big businesses launch their own cloudcomputing products, such as Google’s Google App Engine, Amazon’s EC2and so on.With the advent of these products, some issues began to emerge, such as performance,safety and reliability. Task scheduling is one of the key problems to be solved. It runsthrough the entire process of cloud computing architecture.In this paper, we studied some workflow task scheduling algorithms used incloud computing, most of them are derived from grid computing. Due to the cloudcomputing has a feature of service-oriented, so the original scheduling algorithmsused in cloud computing will make for inefficiency. On this basis, by studying thescheduling process of workflow under the cloud computing environment, this paperproposes a workflow scheduling algorithm called CSR based on control structurereduction. It converts all of the control structure into equivalent sequence controlstructure by such means as mergers and reduction. Then easily identify the criticalpath of the workflow process. By using Time Float Distribution Algorithm, the totaltime float is allocated to each task based on critical tasks in critical path, eventually toenlarge the cost optimization intervals of all tasks.Finally, this paper extends the Cloudsim simulation environment and makessimulation experiments for data and algorithms simulation in the environment.Simulation results show that, compared to other algorithms, CSR has a promisingperformance in decreasing the execution cost, especially for large workflows.