A Metascheduler architecture to provide QoS on the cloud computing

In this paper, a Metascheduler architecture is proposed to be used on Cloud Computing environments. The aim of this Metascheduler architecture is to handle the submission of a service and to manage the infrastructure of the Grid resources belonging to the Cloud. Futhermore, this Metascheduler also aims of building the best pool of available resources for the implementation of the requests issued by a user. For this, an economy model based in approach of supply/demand is adopted. Additionally, it is proposed in this paper a novel approach for the communications between pairs of entities of the collaborative environment. This communications is implemented through the hierarchical P2P scheme, a model that extend the hierarchical model. The central idea on this model is the use of the upper level Metascheduler (named Superscheduler) to control partial overload and to impose performance under service demand.