Migration of Idle Machines via Parking Server – a new & sophisticated Ant Colony Optimization algorithm to address Green Cloud Computing

Cloud computing is a modern concept that includes internet-based computing, as well as the abstraction of hardware devices. It is a convenient on-demand service that can be accessed ubiquitously from any corner of the world. Although this virtual concept has made our day-to-day life easy and efficient with its multitenancy, shareability and metered usage; the fact that its implementation involves wired connectivity of servers, hard-drives and other network devices has created high power and energy consumption problems. Regular carbon emission from cloud data centers has largely increased the carbon footprint. One of the reasons for high energy consumption and power emission from these data centers is the elongated runtime as well as slow and flawed working algorithms. If we can decrease the runtime with some fast and efficient heuristic algorithm, the power consumption problem will be solved. Migrating the idle virtual machines to a Parking Server where they can stay idle and consume minimum resources can address the problem quite effectively. In this paper, I have presented such an algorithm in the name of Migration of Idle Machines via Parking Server Algorithm. This algorithm focuses on parking the idle virtual machines on an isolated parking server until that machine is assigned a task. The algorithm has shown great results and satisfactory accuracy.