Autonomic allocation of communicating virtual machines in hierarchical cloud data centers

Abstract

Cloud providers are typically hierarchically organized into interconnected data centers, each with a collection of racks of servers organized into clusters. The communication cost between two servers is a function of their relative location in the cloud infrastructure. Cloud consumers submit allocation requests for virtual machines, of different types and capacities, and provide an indication of the communication strength between all pairs of requested virtual machines. There is therefore a need for autonomic provisioning of virtual machines in a cloud environment. This paper formalizes the problem of finding an optimal allocation for the requested virtual machines that maximizes the cloud provider's revenue, which depend son how close the requested machines are allocated. This paper presents efficient heuristic algorithms for this NP-hard problem. Experiments show the heuristics to significantly outperform an allocation strategy that is oblivious to the communication strength between virtual machines. The proposed heuristics were also shown to generate between 80% and 90% of the optimal revenue.