Design And Implementation Of Web Servers Load Balancing For Cloud Service

With the rapid development of Internet technology,the proliferation of users and the diversification of demand have brought tremendous pressure on traditional network architecture.The emergence of cloud computing provides a new business computing model.Major companies have built their own public or private cloud platforms,relying on cloud computing technology to expand their business,and improve cluster computing,rapid response,high availability,flexible deployment and other capabilities.As a mature open source product with good scalability and compatibility,Open Stack is widely used in Iaa S cloud computing architecture deployed by enterprises.The cloud service provider builds a private cloud platform based on Open Stack,and provides private cloud services for industries,enterprises,and individuals.Users access cloud services through browsers to obtain corresponding resources.In the case of high concurrent access,it is difficult for a single web server to support normal cloud service business requirements,which is usually solved by web server cluster technology.In the cluster technology,the load balancing strategy directly affects the responsiveness of the cluster system.Therefore,based on the private cloud platform,this paper designs and implements an improved adaptive dynamic load balancing strategy to provide reliable cloud services and speed up the response of cloud platform in view of the shortcomings of the Weighted LeastConnection algorithm in Nginx.The main contribution is summarized as follows:1.Build the overall architecture of the private cloud platform system?Complete the cloud platform software architecture design,including web framework,development language,database design,etc.And complete the background programming of each module of the business layer.2.Build web server related network architecture,and use Nginx as reverse proxy server to realize web server load balancing cluster.At the same time,in order to prevent single point failure of Nginx,Nginx-keepalive master-slave hot standby is built.3.An improved adaptive dynamic load balancing strategy is designed to overcome the disadvantage of Weighted Least-Connection algorithm,which can not change the weights according to the performance of backend servers.Periodically obtain the performance of each back-end server,use the relevant algorithm designed in this paper to derive the weights of each node in the next cycle.The message with node IP address and weight is mounted in the service and registry,and the message is dynamically loaded into the Nginx upstream module by the third-party module to achieve adaptive dynamic load balancing.4.Setting up experimental network environment.Firstly,functional testing of Nginx master-slave hot standby is carried out.Secondly,the cycle T value was determined by JMeter stress test tool.Finally,the improved adaptive dynamic load balancing strategy and the Weighted Least-Connection algorithm are compared and analyzed through several experiments.According to the average response time and throughput evaluation parameters,the adaptive dynamic load balancing strategy can improve system performance.