Abstract:
The fast growth in demand for utility-based IT services has lead to the formation of large-scale Cloud data centers. The electrical energy consumption of these data centers results in high operational costs and carbon dioxide emissions. Cloud data centers benefit from the use of virtualization technology to reduce their energy consumption. This technology enables a Cloud data center to allocate its physical resources on demand and balance loads between its physical hosts by live migration of Virtual Machines (VMs). Consolidating VMs on fewer hosts reduces energy consumption, but the migration of VMs results in low Quality of Service (QoS) and potentially Service Level Agreement Violations (SLAVs).
Dynamic VM consolidation algorithms weigh up the merits of consolidation of VMs on hosts with the disadvantages of migration. Dynamic VM consolidation has three stages: a) Detecting over- and under-utilized hosts; b) Selecting one or more VMs for migration from those hosts; and c) Finding destination hosts for the se-lected VMs. This research consists of three contributions to dynamic consolidation knowledge that are tested using the CloudSim platform. The first contribution is a VM consolidation algorithm that contains different models for each stage. For the first stage, we developed different fine-tuned Machine Learning (ML) prediction models for individual VMs to predict the best time to trigger migrations from hosts. For the second stage, we lexicographically consider migration time and host CPU usage when selecting VMs to migrate. Finally, a new method based on the Best Fit Decreasing (BFD) algorithm was developed to select a destination host for the VMs being migrated. The second contribution is a novel VM selection algorithm that considers CPU utilization of the VMs on each host alongside any linear correlation between the CPU usage of the VMs. The third and final contri-bution is a new QoS metric for evaluating dynamic VM consolidation algorithms. This metric incorporates both performance loss during migration and performance degradation due to host over-utilization and overcomes a weakness in the current QoS metric within CloudSim.