Dynamic placement of virtualized resources for data centers in cloud

Abstract

Cloud Data Centers provides a range of solutions for systems deployment and operation. It is used to provide hosted applications for a third party to offer services to their customers by multiplexing server resources. Virtualization creates a virtual version of the resources by dividing the execution units into one or more execution units. In this paper, we present an approach that uses virtualization technology to allocate data center resources dynamically based on application demands and support green computing by optimizing the number of servers actively used. The resource allocation problem can be modeled as the bin packing problem where each server is a bin and each virtual machine is the item to be packed. Virtualization technology makes it easy to move running application across physical machines without any interruption. We abstract this as a variant of the relaxed classical on-line bin packing problem and develop a practical, efficient algorithm that works well in a real system according to Service Level Agreements. We adjust the resources available to each VM both within and across physical servers with memory de-duplication technologies. This can be used to adjust the VM layout for load balancing and energy saving purpose. Extensive simulation and experiment results demonstrate that our system achieves good performance compared to the existing work.