Research And Application Of High Concurrency Performance Based On Mobile App

With the rapid development of mobile Internet and the wide rise of mobile intelligence,many Internet Companies have shifted the focus of business to the mobile terminal.A large number of mobile terminal traffic continues to rise.According to statistics,in 2017,92% of Tmall business comes from mobile terminals,The peak paid 255,000 per second,database processing peak reached 42 million,trading peak reached 325,000 times per second.In the face of such high concurrent flow,how to build a stable efficient and high concurrent mobile architecture becomes particularly important.High concurrency refers to a large number of users send requests at the same time,due to resource constraints,when the request reaches the threshold,the performance of the server will plummet and may even cause server to be down.According to the characteristics of high concurrency,this paper mainly studies the application of load-balancing,cache technology,distributed technology,database technology and message queue middleware technology in the backstage architecture of mobile application App high concurrent system,combined with these technologies,this paper build a high concurrent system architecture model.And because of the importance of load-balancing in high concurrent system architecture,this paper proposes an improved adaptive weighted round-robin random load-balancing algorithm,and the improved load-balancing algorithm is applied in the background of the App high concurrency system architecture,the clustering performance of App background system architecture to achieve optimal in dynamic operation and some of the strategies and techniques at present in the face of high concurrency scenarios(such as service demotion,current limiting,message queue)are analyzed.The main research contents of the paper are as follows:(1)Introduce the background and significance of the study.This paper expounds the research status of high concurrency performance of mobile applications and analyzes the Nginx technology and cache technology currently used in high concurrency scenarios.(2)The common classical load-balancing algorithm is described.The weight based load-balancing algorithm and the improvement of predecessors are analyzed.Based on that,an improved dynamic adaptive weighted round-robin random load balancing algorithm is proposed.(3)The technical selection of some high concurrency technologies commonly used in high concurrency scenarios is given,and the design and implementation scheme is given.and then combined with some strategies of high concurrency commonly used to build a App background of high concurrency system architecture model,and use the improved load-balancing algorithm to balance the load.At the same time,through the combination of Lua script and Nginx technology,a variety of business logic is implemented at the gateway layer to reduce the pressure on the background server.(4)Finally,we apply the high concurrency system model to the background order system architecture of the three thousand passenger mobile terminal,test and analyze the high concurrent business scenario in the system.