With the increasing popularity of cloud computing services, the more number of cloud data centers are constructed over the globe. This makes the power consumption of cloud data center elements as a big challenge. Hereby, several software and hardware approaches have been proposed to handle this issue. However, this problem has not been optimally solved yet. In this paper, we propose an online cloud resource management with live migration of virtual machines (VMs) to reduce power consumption. To do so, a prediction-based and power-aware virtual machine allocation algorithm is proposed. Also, we present a three-tier framework for energy-efficient resource management in cloud data centers. Experimental results indicate that the proposed solution reduces the power consumption; at the same time, service-level agreement violation (SLAV) is also improved.KEYWORDS cloud computing, dynamic consolidation, power efficiency, three-tier cloud data centers, virtual machine allocation, virtualization
| INTRODUCTIONWith the increasing speed of the Internet over the world, the cloud computing technology is becoming more and more attractive for both customers and service providers. 1,2 The rapid increase in the popularity of cloud services has led to increase in the number of cloud customers. It leads to a daily increase in cloud service demands. That is why main cloud providers such as Amazon, IBM, and Google etc have forced to increase their number of cloud data centers. This makes cloud computing area as one of the most challenging areas in computer science today. It needs a markedly increase in the amount of monetary costs. Hence, there should be a push to have more efficiency in maintenance and execution costs. There has been proposed a number of approaches to get close to this aim; however, it needs much more work to reach optimality. One of these approaches is to reduce the amount of power consumed by cloud resources and other network elements. The less the amount of power consumed, the lower is the amount of monetary costs which has to be paid pack by cloud providers. 3,4 Regarding power and cost efficiency, the amount of carbon footprint has increased drastically due to the increasing amount of power consumption. Hence, reducing power consumption not only reduces the amount of monetary costs for cloud operations but also reduces the threat of a natural disaster. 5,6 To do so, the dynamic virtual machine (VM) consolidation by live migration over runtime is known as a popular and welcoming approach. 7 By this means, we can migrate in or/and out VMs among hosts. More precisely, VMs located on a host with low load (underloaded host) can be migrated out to other active hosts, and then, the host can be switched to