A Method of Improving the Service Quality of Massive Parallel Business Processes

Abstract

The invention discloses a method for improving the service quality of a large batch of parallel business processes, which comprises the following steps: obtaining the historical average values ​​of various activities in the entire batch of parallel business processes through historical big data, wherein the execution of each type of activity is independent of each other; Select a type of activity in the entire batch of parallel business processes, select some activities in this type of activity, define the time when the execution of this part of the activity is completed as the timing detection point, and count the average value of some activities before the timing detection point; calculate according to the analogy relationship The predicted mean value of other types of activities in the entire batch of parallel business processes and the predicted completion time of the entire batch of parallel business processes; The efficiency of the server determines the amount of added resources. According to the method of the present invention, the number of detection points is small, which can satisfy a certain quality of service, and greatly reduces the overhead of the sequence exception processing process.

Description

A Method for Improving Service Quality of Mass Parallel Business Process technical field The present invention belongs to the category of parallel business process detection points, and is specifically related to a kind of improving large-scale parallel business process service. method of service quality. Background technique In recent years, e-commerce and e-government have developed rapidly, and government departments and enterprises often have a large number of business The task needs to be processed in a short time. The processing of these business processes has a certain time limit, such as securities transactions The business needs to be liquidated before 22:00 on the same day. If such massively parallel business processes cannot be It will reduce customer satisfaction and even economic losses. To complete such massively parallel business processes For business processing, a large amount of capital is required for infrastructure construction and operation ...

AI Technical Summary

Problems solved by technology

[0008] Second, the amount of additional server resources needed to handle the abnormalities that have occurred. If too many server resources are added, it will cause some unnecessary waste of resources and funds. If the increase of server resources is too small, it may cause The business process also cannot be completed on time and cannot meet the user's demand for service quality

But in real life, some situations do not allow delays, such as weather forecast, if the forecast result cannot be calculated within the specified time, then the result will lose the significance of the forecast

Images