Research On Load Balancing Technology Of Web Server Cluster Based On Nginx

Nowadays,the Internet is highly popular,and many industries are inseparable from the rapid development of Internet technology support.With the double growth of the number of Internet users and the use of devices,the number of website business visits has grown rapidly,which is a severe test for the concurrent response capabilities of various companies' Web servers,and people urgently need to explore new technologies to improve the overall performance of the Web server.The most common solution at present is to use multiple servers to form a cluster system to provide users with unified network services.Whether the cluster can properly distribute users' concurrent access requests to each server in the cluster system through a certain strategy is the key to measuring the performance of a server cluster.Throughout the process,load balancing is the process of distributing access requests.This paper improves an algorithm based on Nginx's original weighted round-robin by studying Nginx's load balancing strategy.The algorithm selects CPU,memory,disk I/O and network bandwidth as load indicators,uses the load information to calculate the weights of each node in the cluster,and distributes the access request according to the weight of the node.Moreover,the algorithm also monitors the load status of each node in the cluster.If the cluster load is found to be unbalanced,the weight of the server node with heavy load is reduced,so that the cluster system is re-balanced.In addition,the test tool is used to test the three load balancing strategies of Nginx's original weighted round-robin algorithm,fair algorithm and dynamic weighted roundrobin algorithm.By comparing and researching the test results,it is proved that the strategy proposed in this paper is feasible and effective.Compared with Nginx's original load balancing strategy,the dynamic load balancing strategy has more application value,mainly as follows:(1)The node performance evaluation is more complete,and the weight of the node is calculated according to the load information of the node,which is more effective than the personal experience to assign the value to the node.(2)The weight modification process is more scientific,and the weight modification threshold is set.When the variance of the resource usage rate of each server exceeds the set threshold,it indicates that the cluster load condition is unbalanced,and the weight modification process is started.(3)The AHP is used to calculate the weight vector in the evaluation function,and the consistency of the judgment matrix is tested to ensure that the obtained result is convincing.