ABSTRACT
Fei Teng Applied Mathematics and Systems Laboratory, Ecole Centrale Paris, Grande Voie des Vignes, 92295 Chaˆtenay-Malabry, France
Fre´de´ric Magoule`s Applied Mathematics and Systems Laboratory, Ecole Centrale Paris, Grande Voie des Vignes, 92295 Chaˆtenay-Malabry, France
As network speeds grow, it is possible to construct large-scale highperformance distributed computing environments to allow users to submit jobs from anywhere in the world. These jobs can then be run on any available computing resource. As a consequence these resources should be assigned effectively to provide reliable and fast distributed services and to reduce the turn-around time of user jobs and the scheduling scheme should be heavily considered. There are numbers of heterogeneous or homogeneous clusters which can provide the job queuing mechanism, scheduling policy and local resource management, including utility computing, cluster computing, grid computing and so on. Among them, grid system is the most popular one and highly discussed, developed by the researchers and information technologies (IT) developers during the past decade. Recently a new term, “cloud computing,” has emerged, which infers that computing is not operated on local computers, but on centralized facilities by third-party computing and storage utilities. Literally, clouds and other computing paradigm share the similar
to execute on many resources. However, clouds are more made available in a pay-as-you-go manner to the public or internal data center of business [Armbrust et al., 2009 ], [Buyya et al., 2000a ], [Buyya et al., 2000b ], [Buyya et al., 2001 ]. The cloud computing new characteristics differ from grid computing in the implementation details, which requires researchers and engineers to reconsider the resource scheduling strategies.
