This paper presents a power-aware scheduling policy algorithm of Virtual Machines into nodes called Green Cloud (GreenC) for Heterogeneous cloud systems. GreenC takes into account optimal assignments according to physical and virtual machine heterogeneity, the current host workload and communication between the different virtual machines. An initial test case has been performed by modelling the policies to be executed by a solver that demonstrates the applicability of our proposal for saving energy and also guaranteeing the QoS. The proposed policy has been implemented using the OpenStack software and the obtained results showed that energy consumption can be significantly lowered by applying GreenC to allocate virtual machines to physical hosts.
This work was supported by the MEyC under contracts TIN2011-28689-C02-02. The authors are members of the research groups 2009-SGR145 and 2014-SGR163, funded by the Generalitat de Catalunya.
English
Green cloud computing; SLA; Power-aware scheduling; Non-linear programming
Springer Verlag
MICINN/PN2008-2011/TIN2011-28689-C02-02
Reproducció del document publicat a https://doi.org/10.1007/s11227-014-1351-2
Journal of Supercomputing, 2015, vol. 71, núm 5, p. 1817-1832
(c) Springer Science+Business Media New York, 2015
Documents de recerca [17848]