Design And Analysis Of Performance Test Of Family Medicine Back-end Management System Based On Load Runner

With the continuous development of the software industry,the focus of enterprises will gradually shift from the number of functions implemented by the software system to the quality of the functions implemented by the software.For the target customers of software systems,higher quality and stronger systems are more attractive to the market.How to shorten the test work time under the premise of ensuring software quality has always been a problem that testers need to face.Therefore,by analyzing the test set data of small sample influencing parameters,directly and accurately predicting the performance of a function,which will greatly help shorten the man-hours required for testing.How to perform high-quality performance tests and quickly build efficient and accurate performance prediction models based on small sample performance test data is the focus of this paper.This paper applies machine learning ideas to the performance prediction of system software,and proposes a hybrid k NN performance prediction module construction method.It is hoped that a performance prediction module that can be applied to multiple types of software performance testing applications can be constructed,thereby reducing the time spent by developers on performance testing.This paper takes the performance test of the family doctor platform background management system as the background,and combines the design and research purpose of the system performance prediction module.It mainly researches two aspects of software testing technology and software performance prediction methods.First set up a test environment according to the actual situation of the system,and develop and debug the performance test script used by Loadrunner according to the performance test cases.Then start with a single business load scenario,analyze the performance of each module in the performance test and performance defects,and put forward corresponding tuning opinions after research.Finally,starting from the comprehensive business load scenario,taking the residents management module as anexample,the working conditions and prediction accuracy of the hybrid k NN prediction model are studied,and the conclusion is drawn whether the prediction module can complete the performance prediction task.It also briefly analyzes the system's performance in the integrated business scenario and the tuning plan.