Research On Load Balancing Strategy Based On Server Cluster

With the rapid development of Internet,traditional industries and the Internet are combined closely in modern society.When the Internet brings convenience for people's basic necessities of life,the growth of its explosive access also brings a huge challenge to the computing strength and processing power of the network core equipment.One of the more prominent thing is the server load capacity test of high concurrency access.Using load balancing cluster system is a common strategy to solve the above problems.The concept of load balancing system,background,current research has been described in this paper,and it analyzed the implementation of the principle of load balancing and the common load balancing algorithms.It discussed other Load Balancing strategy lately,and it made an exposition on how to get load information under the linux system and made an evaluation of selecting load information.On the basis of the previous research results of load-balanced cluster,the following research is done in this paper: The basic principles,features and problems of genetic algorithm and simulated annealing algorithm are analyzed in detail in this paper,and presents improvement measures with load balancing cluster features.It presents a improvement strategy for load information evaluation:by means of improved genetic algorithm to determine the weight of each server group for processing related tasks.Real-time load information is detected by means of the combination of the weights and the actual load parameters.In task allocation,genetic algorithm and simulated annealing algorithm are combined as a new load balancing algorithm: genetic-simulated annealing algorithm.Genetic algorithm is responsible for the global search of all the elements of the cluster in the early stage to overcome the shortcomings of slow convergence of simulated annealing algorithm,and make the target solution to converge to the global optimal region quickly.In the later stage,the local search ability is optimized by using simulated annealing algorithm and the global optimal solution is obtained to complete the optimization of load information algorithm.The genetic annealing algorithm is implemented on the Linux system based on Nginx,and the performance of the genetic algorithm is tested.A large number of research results show that the genetic-simulated annealing algorithm improves the efficiency of the load balancing cluster and improves the performance and efficiency of the cluster system.