Research On Running State Processing Of Business Workflow Based On Prediction

With the gradual development of cloud computing,large-scale time-constrained business processes are gradually being valued by many commercial organizations.Since the cloud computing environment can provide unlimited cloud computing resources,users can apply for the required resources at any time,so the cloud Instance-intensive business workflows in cloud computing environments are becoming an extremely important business application.However,the cloud computing environment itself has dynamics and uncertainty.It is difficult to ensure that the business workflow can be completed within the specified time.However,if the method of constructing the corresponding software and hardware platform is used to deal with the business workflow,it will cost a lot.Resource costs can even have security issues.In contrast,the cloud computing environment is more suitable for processing such time-constrained business workflows.Due to the shortcomings of cloud computing itself.,how to improve the service quality requirements of business workflows in cloud computing environments,and satisfy users' requirements for cloud computing services.Satisfaction has become an extremely important issue.The main research work and innovations of this thesis mainly include:Firstly,a predictive-based business workflow detection point selection strategy is proposed.Because instance-intensive business workflows can have timing delays at runtime,they can't be completed within the specified time.Therefore,when the business workflow is executed,it needs to be monitored.However,because the number of activities of the business workflow is too large,each activity is detected,which is costly,so setting a reasonable detection point is to improve the business workflow.An effective way to complete the success rate on time.However,if the timing exception occurs,and then the corresponding processing,for some business workflows,may result in unrecoverable results.Therefore,in the process of business workflow execution,the situation of the remaining activities is predicted by the execution of some activities to proactively predict the possible occurrence of the anomaly.This thesis finds the location of the detection point by determining the relationship between the scale of the activity of the business workflow and the detection points.The experimental results show that the strategy can well predict the implementation od business workflow and improve the service quality requirements of business workflow.Secondly,based on the selection of detection points,a timing exception processing strategy is proposed.During the execution of the business workflow,timing exceptions will occur.If it is not processed timely and effectively,the business workflow may not be completed on time within the specified time.Therefore,it is very important to adopt a reasonable exception handling strategy.In the cloud computing environment,the user can apply for the required resources at anytime and anywhere.Based on the selection of the detection points,this thesis applies for appropriate server resources by predicting the possible execution time and delay of the subsequent activities at the detection point location.However,due to the problem of dealing with the abnormal cost of timing,the thesis monitor the server for additional applications in real time,and continuously observe the execution time of the remaining activities and the current execution of the server to determine when to deactivate the server and the number of servers that are deactivated.This reduces the cost of unnecessary server uptime.The experimental results show that the strategy can deal with the anomalies that occur effectively,improve the service quality of business workflow,and achieve user satisfaction.Aiming at the monitoring of business workflow execution process,this thesis proposes a prediction-based detection point selection strategy to solve the problem that timing conflicts may occur during the execution of business workflows with many activities Firstly.This strategy can reduce the number of monitoring and cost,and can meet the demand for service quality of business workflow.In order to further solve the problem of the method when the anomaly occurs,this thesis proposes a timing exception handling strategy based on the selection of the detection point,dynamically requests additional server resources and dynamically disables it.The on-time completion rate of business workflows,and to some extent,reduces the additional cost of exception handling.