Design And Implementation Of Cloud Computing Node State Management System

Cloud computing is a hot technology in the current information age.It uses the virtualization technology to achieve the integration of resources, and provides it to the users in the form of services, so that users can efficiently and reliably use the computing resources and storage resources.In the cloud computing system,the efficient management of server resource by node monitoring and virtual machine migration technology can improve the usage of server resource, and make the server load balancing and fault recovery.The paper designs a hierarchical node status monitoring system which is called CVM. The CVM consists of manager monitoring node (CVMan), cluster monitoring node (CVClu) and PM monitoring node (CVMon).There is only one CVMan in CVM.It is responsible for collecting metadata about the system.For robustness, there are multiple CVClu in a cluster. One is primary CVClu and others are backup CVClu. CVMon is implemented on each PM. CVMon is mainly responsible for collecting performance information about the PM and VMs which run on the PM. The paper also designs a Huffman manangement strategy about cluster. The results of the experiments demonstrate that the overhead of the monitoring system is relatively small, and also has a good accuracy.In order to achieve a better management of VM migration, the paper proposes a VM migration scheduling method based on simulated annealing algorithm that is SAMS. The method uses the CVMon of CVM to find abnormal VM and resource ordering policy SEQ analysis to obtain the potential purpose VM mapping scheme, and a path switching strategy can shorten the migration time as much as possible and make the entire cluster system maintain the state with load balancing after the completion of scheduling. The results of the experiments show that the algorithm has a good convergence and efficiency, but also can shorten the migration time well with the path switching strategy.Finally, the paper designs the system architecture of the node status management system. Combined with the monitoring system CVM and virtual machine migration scheduling method based on simulated annealing algorithm called SAMS, the paper implement a node status management system based on VM migration, which enables the node status monitoring and management realized by node migration scheduling based on the cloud computing environment.