Cost-Efficient Algorithms for Critical Resource Allocation in Cloud Federations

Gomory-Hu based algorithms are proposed to select hosting resources and place services in a multiple providers federation context. The algorithms, use the Gomory-Hu tree, to detect critical nodes and edges in the end user requests to make selection and placement decisions. The algorithms take into account the providers' internal costs and network connectivity costs and hosting costs by the federation members. The algorithms are shown to be scalable and near-optimal through numerical simulations and evaluations.