Service Load Evaluation And Load Balacing Method In Virtual Machine

With the development of IT(Information Technology),the concept of network management is also undergoing profound changes,network management system based on Web Services mechanism and service oriented SOA(Service Oriented Architecture)architecture is becoming more and more popular.Self-organization,network centered,intelligentialize and service oriented has become the future development theme of network management technology.With the rapid growth of the number of services in the Web environment,it's a typical worth studying problem how to evaluate the Web service load.By using virtualization technology,virtual machines can be isolated from a physical server.Deploying various web services on the bottom of the virtual machine helps to enhance the level of service,to quickly meet the diverse need of management,which can ensure the security and the best use of physical resources.There is a high degree of collaboration between SOA and virtualization,with the increasing scale of network and management services,the virtualization technology to support and achieve these is increasingly important.The combination of SOA and virtualization helps to build a highly flexible network integration framework,to achieve the optimal allocation of resources management.In view of the phenomenon that the individual node load is too large due to the different service traffic in the virtual environment,this paper studies the calculation of the virtual machine load and the evaluation of the service load,selects the relevant influence factors to conduct experiments,proposes the calculation formula of the virtual machine load and the evaluation function of the service load,designs and implements the monitoring and dispatching system based on SOA in the virtual environment.The system mainly includes the monitoring of the virtual machine,the monitoring of the Web Services service on the virtual machine and the migration and scheduling and load balancing of the service.The system realizes the control of resources in virtual environment;triggering an alarm reporting information to the dispatch center while individual virtual machine overload;determining the destination node for service migration by scheduling center;a series of automated processes such as migration scheduling and load balancing.