Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Computer network > General issues > Theory of computer networks

The Research and Implementation of Resource Adaptive Allocation in a Virtual Environment

Author ChenHuiPing
Tutor FengDan
School Huazhong University of Science and Technology
Course Computer System Architecture
Keywords Virtualization Resource Allocation Quality of Service Back Propagationneural network
CLC TP393.01
Type Master's thesis
Year 2013
Downloads 7
Quotes 0
Download Dissertation

Virtualization technology can isolate specific hardware architecture and softwaresystems, allows multiple applications to run on the same physical platform, thus provide areliable solution for the sharing of physical resources. As we all know, differentapplications demand for varied resources. The physical resources needed by applicationvaried by the workload and application quality for users. Existing virtualization platformscannot adjust resources allocation dynamically based on the workload of virtual machines,thus cannot meet the requirements of service quality. An adaptive resources allocationmethod is needed to adjust resource allocation among virtual machines in real time.We analyzed and summarized current research status of adaptive resource allocationin virtualization environment, designed and implemented an adaptive resource allocationmethod based on Xen platform. We predicted the application workload according to theresource consumption status obtained from the virtual machines, predicted the resourceallocation method meet the quality of service requirements under the application workload,and adjusted resource allocation according to virtual machine allocation strategy.The test result showed that our adaptive resource allocation prototype system canadjust resources allocation among virtual machines dynamically based on the changes ofvirtual machines’ application workload. The resource allocation method can meet thequality of service requirements and mostly optimal. And the system performance loss wasless than9%.

Related Dissertations
More Dissertations