Research On Scheduling Algorithm Of Instance-Intensive Workflows On The Cloud Computing Invironment

Along with the rapid development of cloud computing and the Internet, instance-intensive workflow under the cloud computing environment, which is an important commercial applications, becomes widely used in people’s work and life, such as online payment and online shopping in e-commerce. Therefore, the study of its scheduling problems has important practical significance. However, how to utilize the characteristics of instance-intensive workflow and cloud computing, such as using in need and paying in use, and design a suitable algorithm to achieve the target for finishing the workflow instances as much as possible in the shortest time, is an problem which need to be solved.In this paper, we research the instance-intensive workflow scheduling algorithm under the cloud computing environment. Firstly, it introduces research background, significance and the research status both at home and abroad of the instance-intensive workflow scheduling algorithm. Secondly, it describes the relevant concepts of cloud computing and cloud workflow. Then workflow scheduling algorithms are classified and described, and the existing examples of instance-intensive workflow scheduling algorithm are emphatically discussed. Thirdly, according to the characteristics of cloud computing environment, a system framework of instance-intensive workflow is designed. On the basis of the system framework, this paper presents a instance-intensive workflow scheduling algorithm under the cloud computing through improving and optimizing the deadline for assignment algorithm and Min-Min algorithm:Minimum Total Cost Under User-designed Deadline(MTCUD)algorithm. Finally, the simulation experiment about the algorithm and comparative analysis are doing on the platform which is extended from the simulation platform CloudSim.The experimental results show that the MTCUD algorithm can improve the throughput of the system, and can finish workflow instances as much as possible in the shortest time. Therefore it is more suitable for multiple instances of the cloud computing environment. At the same time, the algorithm can effectively reduce execution time and execution cost of instance-intensive workflow in the cloud computing environment.