Research On High Concurrent Access Mechanism Based On Load Balancing And Service Flow Control

With the popularity and rapid development of the Internet,software applications have higher and higher requirements for concurrency and service quality,which promotes the continuous evolution of the Internet's architecture.The rapidly growing user scale and increasingly complex business systems have led to an explosive growth in concurrent access traffic on the network.A single server architecture is limited by hardware and network bandwidth,and it is difficult to cope with massive user access.Clustering and load balancing technologies have emerged,which can provide more powerful task processing performance and fault tolerance.Among them,the microservice architecture has received widespread attention due to its excellent organizational structure and development performance.It can reduce the degree of system coupling by splitting a complex system into multiple independent service architectures,and can be combined with cluster technology to achieve traffic distribution type processing.However,when the total amount of user access traffic is too large,no matter what load balancing technology is used,the traffic allocated to each server in the cluster will still exceed the maximum processing capacity of the machine.Especially for burst traffic,it is necessary to apply flow control technology.solve.To this end,this paper will conduct research from two aspects of cluster load balancing and microservice flow control,and propose an optimized dynamic load balancing model and a multi-level dynamic service flow control model based on the sr TCM algorithm,and design and The realization scheme is verified,and the main research contents of this paper are as follows.First,based on the processing time of the request task as a performance evaluation parameter,a basic model of load balancing based on the ant colony algorithm is constructed.Secondly,the basic model is comprehensively optimized,and an optimized dynamic load balancing model is proposed.Its main optimization work is: comprehensive optimization of the algorithm rules of annotated ant colony algorithm;periodic collection of data on the load information parameters of microservice nodes,The path search method of the ant colony algorithm is optimized;the pheromone adjustment factor is introduced,and the pheromone update method is optimized according to the request response time.Third,based on the Open Resty platform,an optimized dynamic load balancing model implementation scheme is designed.Research on the theories related to flow control technology under the microservice architecture,and propose a multi-level dynamic service flow control model based on the srTCM algorithm,which can dynamically control and differentiate the tasks entering the service node according to the type of request and the running state of the machine And designed a dynamic flow control model implementation scheme based on the Gateway gateway.Through the performance comparison experiment,the performance of the research results of this paper is verified,and it is proved that it is superior to the commonly used load balancing and current limiting schemes.