A Hybrid Method for Reduction of Energy Consumption in Cloud Networks

Mehran Tarahomi, Mohammad Izadi
2019 Journal of Networking Technology  
Cloud computing is consisted of physical data centers each of which including hundreds or thousands of computers. A key technology which has enabled cloud computing feasible, is virtualization. Virtualization allows us to separate virtual machines in a way that each of these so-called virtualized machines can be configured on a number of hosts according to the type of application of the user. It is also possible to dynamically alter the allocated resources of a virtual machine. Different
more » ... of energy saving in data centers can be divided to three general categories: 1) energy saving methods based on balancing the loads of resources; 2) scheduling by the help of hardware facilities; 3) reducing the consumption of energy through being aware of thermal characteristics of the environment. One of the most important challenges in terms of cloud computing is to maintain an optimized load balance in cloud environment on servers because this issue is critical for energy consumption. Since load balance algorithms depend on the current behavior of the system, they act dynamically. Important factors which should be taken into account while developing such types of algorithms include load estimation, load ratio, adaptability of different systems, performance of the system, interaction between nods, scheduling, nature of the transferred service, selecting nodes and etc. By taking a detailed look on previous methods and challenges of energy saving in cloud networks, we provide a hybrid method which enables us to save energy through finding a suitable configuration for placement of virtual machines and through being aware of special features of virtual environments for scheduling and balancing their dynamic loads by the help of live migration method.
doi:10.6025/jnt/2019/10/3/92-103 fatcat:5stozg3y5ja4zhzdeasw3ulw4a