Research And Implementation Of A Load Balancing Technology Based On Data Correlation In Cloud Computing

Cloud computing is a new computing model that can access to the shared resource pool anytime and anywhere.In recent years,it has been widely concerned.But the diversity of user requirements in cloud computing environment,the heterogeneity and complexity of cloud computing resources often result in the load of servers is unbalanced and the resource utilization of system is down.The emergence of load balancing can effectively solve these problems.It can allocate a large number of concurrent task requests to the servers to ensure efficient system operation.When virtual machines in cloud computing environment process large amounts of computing tasks,if some of the virtual machines that handle the associated data are migrated to different servers while balancing the load,it will cause a lot of waste of system resources during the migration and computation.Based on the summary and analysis of existing load balancing Strategies,this paper takes into account the correlation between the virtual machines that handle the associated data and proposes a load balancing strategy based on data correlation.This strategy can improve the resource utilization of system by constructing the load-intensive data set to carry on the overall migration.The main work of this paper is as follows:(1)Proposing a load balancing strategy based on data correlation.Firstly,the strategy constructs the initialized data set by calculating the data correlation factor.Then,it constructs the load-intensive group by expanding the initialized data set according to the correlation of processing data between virtual machines.Finally,it takes the server with the shortest response time as the target migration server and achieves dynamic migration of load-intensive group.(2)Proposing an implementation framework of adaptive load balancing migration strategy based on mobile agent.Based on the introduction of mobile agent technology,this framework adopts agent-based three-tier architecture design method of server and performs load balancing migration through collaboration between different roles of agents.Finally,we use WSLA language to implement the decision of load balancing and the process of load migration.(3)Implementing a prototype system of cloud computing based on data correlation load balancing strategy and describing the specific implementation process by using cloud computing simulator CloudSim.The experimental results show that the load balancing strategy based on data correlation in cloud computing proposed in this paper can effectively realize the dynamic load migration between servers,reduce the communication overhead between servers and improve the resource utilization of system.