Consolidation and replication of VMs matching performance objectives

The users of actual computing infrastructures allowing the resource provision (such as clouds) are often asked to decide about the proper amount of equipment (virtual machines, VMs) required to execute their requests while satisfying a set of performance objectives. These types of decisions are particularly difficult since the direct correlation between the resources allocated and the performance offered is influenced by a number of factors such as the characteristic of the different class of requests, the capacity of the resources, the workload sharing the same physical hardware, the dynamic variation of the mix of requests of the different classes in concurrent execution. In this paper we derive the impact on several performance indexes by two popular techniques, namely, consolidation and replication, adopted in virtual computing infrastructures. In particular we present an analytical model to determine the best consolidation or replication options that matches given performance objectives specified through a set of constraints.