A server response time intelligent optimization method and system

Abstract

The invention discloses a server response time intelligent optimization method and system. The method comprises analyzing factors influencing the performance of a thread pool, and determining how to evaluate the performance of one thread pool; from the perspective of the user, the user is more concerned about when the submitted task request can be processed and responded; therefore, the evaluation of the performance of one thread pool is the time spent from the submission of the task request by the user to the start of the task queue to the completion of the task in the thread pool, and this time is called the server response time of the user task; on the basis of deeply analyzing the operation of the thread pool, a mathematical model capable of reflecting the performance of the thread pool is established; according to the constraint relation of the characteristic quantity to the thread pool performance obtained through the scheme, the thread pool performance optimization model for adjusting the size of the thread pool according to the change of the three characteristic quantities of the throughput, the task operation time and the task blocking time is determined.

Description

technical field [0001] The invention relates to the field of smart grids, in particular to a method and system for intelligent optimization of server response time. Background technique [0002] With the in-depth construction and application of the smart grid, more and more users request that the grid ICT server can provide stable and reliable services. The ICT server application needs multiple concurrent threads for each user to process their requests. But this method has two disadvantages: first, in the face of a large number of user requests, the system overhead of creating a large number of threads is still very large; second, the system also consumes a lot of resources for maintaining active threads. The result of this is likely to be that when the ICT server application is faced with excessive user service requests, resources such as memory will be exhausted due to the creation, maintenance and destruction of so many corresponding threads, which will greatly affect th...

AI Technical Summary

Problems solved by technology

But this method has two disadvantages: first, in the face of a large number of user requests, the system overhead of creating a large number of threads is still very large; second, the system also consumes a lot of resources for maintaining active threads

The result of this is likely to be that when the ICT server application is faced with excessive user service requests, resources such as memory will be exhausted due to the creation, maintenance and destruction of so many corresponding threads, which will greatly affect the ICT server application. server performance

Therefore, the ICT server application needs a mechanism to solve the system performance degradation caused by frequent creation and destruction of threads and the exhaustion of system resources caused by excessive client requests

Images