Dynamic Resource Allocation Based on Priority in Various Data Centers Custom-Based Waiting Queue Technique in Cloud Computing

Abstract

For delivering the cloud services over the internet, cloud computing has become proficient infrastructural model for hosting cloud services. Server virtualization is key technology that enables cloud computing as a service, which authorizes dynamic sharing of physical resources. Virtualization introduces the problem of virtual machine placement that increases the overheads in load balancing. Existing infrastructure needs the strategy for the VM Placement as it may create poor allocation and load balancing issues. In most of the cases, due to lack of input parameters physical machines are partially loaded that creates issue of fragmentation which leads to in sufficient resources that causes more utilization of physical machines in any infrastructure. We did extensive survey in the said domain and found that, In load balancing approach, when VM place mentis done without measuring its lifetime, that creates fragments on PM. So we propose dynamic priority based spill over technique and add the concept of short life/long life container for solving the fragmentation issue.