Author:
Paolo Campegiani
Affiliation:
University of Rome Tor Vergata, Italy
Keyword(s):
Cloud, Green Computing, Virtual Machines Placement, Service Level Agreements, Resources Allocation, TCO, Genetic Algorithm, Bin Packing.
Related
Ontology
Subjects/Areas/Topics:
Economic Models of Energy Efficiency
;
Energy and Economy
;
Energy-Aware Systems and Technologies
;
Green Architectures for Grids, Clouds and Clusters
;
Green Data Centers
;
Optimization Techniques for Efficient Energy Consumption
;
Sustainable Computing and Communications
;
Virtualization Impact for Green Computing
Abstract:
We present an energy aware model for virtual machines placement in cloud computing systems. Our model manages resources of different kind (like CPU and memory) and energy costs that are depending on the kind and amount of deployed resources, incorporating capital expenses (costs of infrastructure and amortizations), operational expenses (electricity costs) and data center energy parameters as PUE, also with possibly different service levels for virtual machines. We show that the resulting model could be solved via a genetic algorithm, and we perform some sensitivity analysis on the model energy parameters.