Energy Management of the System: An Empirical Investigation of Virtualization Approaches in Static and Dynamic Modes

Abstract

Background: Energy management has a vital impact on success of virtualization. Power consumption and generation of heat play critical roles in the core concept of energy management. Presently, energy consumption is continuously growing due to high usage of system resources and rising of demand in computer density. The high power usage and high heat generation lead to instabilities of their hardware systems and cause system reboot more often. Materials and Methods:To avoid above malfunctions and save from permanent damage to the system resources, this study are investigated and tested the power consumption and heat generation during the usage of low-level performance application, online education application and high-level performance application, respectively via the implementation of XEN, KVM and Docker virtualization technologies in between static and dynamic modes. Results:The result proved the power consumption and heat generation is higher for hypervisor-based virtualization and very lowfor container-based virtualization in all kind of applications. Even though the XEN and KVM are the same type of hypervisor is offering similar features but the output has deviated from each other.The limitation of VM/Docker on top of physical machine would control by mathematical formula, which created from our experiment. Conclusion:The overall analysis proved that XEN, KVM and Docker as the order of the power consumption rate and heat generation rate for all applications in high, medium and low, respectively.