Abstract:Virtual machines (VMs) consolidation is one of the primary methods for improving the energy efficiency of cloud data centers. However, aggressive VMs consolidation methods may cause the hosts overload and produce massive inefficient VM migrations, so that host re-overload and even Quality of Service (QoS) degraded. Therefore, workloads are increasingly being deployed in virtualized cloud data centers to improve energy efficiency by using Virtual machines (VMs) consolidation. In this paper, we propose VMs selection algorithm and VMs placement algorithm for VMs consolidation to meet Quality of Service (QoS) for cloud applications while maximizing the energy efficiency and resource utilization of the data center. The VMs selection algorithm develops the relative CPU capacity gains to select the VMs which are selected to migrate from overloaded hosts. The VMs placement algorithm proposes an adaptive reservation CPU capacity for each physical host and allocates the selected VM to the hosts with idle CPU capacity. We experimented with real workloads, and results show that the algorithms can significantly reduce the energy consumption and improve QoS while decreasing the number of VM migrations.

Based on QoS and Energy Efficiency Virtual Machines Consolidation Techniques in Cloud