Core group placement: allocation and provisioning of heterogeneous resources

We present a theoretical and empirical study on a recently introduced combinatorial optimization problem, namely core group placement problem. The problem arises from real-world business requirements as part of resource allocation in cloud management. In particular, it focuses on the allocation and provisioning of a set of heterogeneous resources serving multiple customers each with different service-level agreements. There exist certain business rules that govern the application stemming from privacy, performance, and capacity requirements. From a theoretical point of view, we prove that the problem is intrinsically hard, yet, from a practical point of view, we show how to formulate it as a constrained optimization program using constraint programming (CP), and alternatively, using mathematical programming (MP). Our experimental results demonstrate that the CP solution outperforms its MP counterpart. We then move toward a dynamic setting where the problem manifests itself in the real world. We show that CP model not only addresses the resource allocation problem but it also enables resource provisioning to take future considerations and system growth into account when making decisions. Overall, the CP solution stands out as a high-level, declarative solution that is efficient, easy to maintain and can address multiple scenarios.