Cloud computing applications for business is gradually becoming reality. Vendors such asMicrosoft, Amazon have launched their own commercial cloud computing platform solution. Cloudcomputing provides a dynamic service-oriented virtualized computing model by improving andintegrating grid computing, cluster technology and distributed systems technology. Virtualizationtechnology and the optimization of virtualized resource allocation play key roles in the large-scaledeployment of cloud computing platform. Hence, it is vital to investigate how to achieve efficientmanagement of virtual devices and cloud computing resources, improve resource utilization andsupport traditional services with cloud computing platform.Based on existing research, and cloud computing system implementation, this thesis hasanalyzed the management requirements of cloud computing platform as well as existing problemsin combination with virtualization technology. Meanwhile, by improving the current managementalgorithms and architectures, this thesis puts forward a complete schema of virtualized dynamicresource management system, including information gathering of physical hosts and virtual hosts,state monitoring of physical hosts, resourse warning of physical host and dynamic migrationstrategy of virtual hosts. Moreover, this thesis verifies the effectiveness of the system’s dynamicmanagement of CPU and memory resources through triggering memory resource alarms and CPUresource alarms. |