Research On Quality Of Service About Business Workflow Based On Empirical Study

With the rapid development of e-commerce,the applications of large-scale parallel business processes increases,and instance-intensive business workflow has become an important and widely used commercial application.Due to the huge amount of data in instance-intensive business workflows,to ensure the quality of service of this kind of business process,the traditional distributed computing and grid computing has been difficult to meet the requirements,while cloud computing can handle massive data through the cloud.Users can pay for the high-performance computing resources according to their requirements and obtain the services.However,because the dynamic cloud computing environment has the natural characteristic of uncertainty,when dealing with instance intensive business workflow by using cloud computing resources,ensuring that the business process is completed within the prescribed time and reducing users' expenses,can improve the service quality of the cloud environment,so as to improve the users of satisfaction.Therefore,how to guarantee the quality of business workflow in cloud environment is a key problem in the research of instance-intensive business workflow.The main research works and major contributions of this thesis include:First,an empirical research method is introduced to propose the QoS factors of business workflow in cloud environment.In the traditional environment,the QoS of business workflow system is related to many factors,but when the business workflow is combined with cloud computing,there will be new changes.According to the method of systematic literature review,this thesis makes a literature retrieval of five databases,and finally determines 310 documents that meet the requirements.After rigorous systematic review and analysis,30 experimental literatures were identified.According to these 30 literatures,the QoS factors of business workflow in the cloud environment are determined by the method of data analysis and synthesis from the system literature review:execution time and execution cost.Second,a cloud workflow server selection strategy is proposed to minimize the cost of using cloud server while satisfying the quality of user service.Business workflow can process massive data through cloud,but users need to pay a certain cost while using cloud computing resources.At present,the performance of all kinds of resources in the cloud computing environment is not unique,and the fees charged by cloud service providers are different for different performance.Users usually want to complete the business workflow by paying the least amount of money on the premise of ensuring the QoS.However,cloud service providers only give some parameters and fees for different resources.According to this information,users can not choose resources that suit their requirements.For this reason,this paper proposes a cost-effective cloud computing resource selection strategy,which provides an effective solution for users to select cloud resources.This strategy can ensure that the cloud workflow can be completed in deadline and reduce the cost of cloud resources.Third,for the timing exception that occurs during the execution of a business workflow,a throughput based instance-intensive business workflow temporal violation handling strategy is proposed to minimize the user's cost.Because of the dynamics and uncertainty of the cloud resources,when the performance of the cloud resources changes during the execution of the workflow,the actual execution time does not meet the required completion time of the user and produces the timing anomaly.Based on this problem,this thesis proposes a timing exception selection strategy based on throughput,and selects the timing exception point to handle the timing exception,which ensures that the workflow can be completed on time and minimize user costs.This strategy can dynamically change the number of cloud servers to restore the temporal anomalies of business workflow,so then improve the completion rate of business workflow on time,and reduce the user cost.This thesis first analyzes the two QoS factors:user execution time and execution cost which are most concerned by users about instance-intensive business workflows in cloud according to a result of a comprehensive systematic literature review.Focusing on the two QoS factors,this thesis further proposes a strategy for selecting cloud workflow resources based on QoS constraints,which provides a solution for users when selecting cloud resources.The strategy can ensure the cloud workflow completed within the stipulated time,and reduce the user's cost.In order to further improve the workflow service quality and reduce the cost of users,this thesis then creates a sequential consistency model of business workflow from the perspective of throughput,puts forward a throughput-based instance-intensive business workflow timing anomaly processing strategy and proposes a throughput based instance-intensive business workflow temporal violation handling strategy.This strategy fulfills the functions of timing constraint setting,timing monitoring and timing anomaly processing in the execution of business workflow,which significantly improves the service quality of business workflow and saves the execution cost of business workflow.