SLA-Aware and Green Resource Management of IaaS Clouds

Abstract

Cloud computing has become a most popular computing pattern in both industry and academia. Infrastructure as a Service (IaaS) cloud platform provides an elastic way to require computing resources in a pay-as-you-go paradigm. While bringing great convenience to both customers and service providers, the cloud platform has some problems demanding prompt solution, one of which is how to achieve energy saving with a low Service Level Agreement (SLA) violation rate. Though existing works achieve optimization to a certain extent, most of them do not take the overall resource utilization balancing into account. In this paper, heuristic greedy algorithms are proposed based on a novel conception, Threshold. By regulating resource utilization of clouds towards Healthy Threshold, the overall balancing of resource utilization is achieved, so energy consumption can be reduced more effectively. Besides, we also propose a strict SLA enforcement strategy to further guarantee a low SLA violation rate. Results of real-world trace driven simulations show that our algorithms outperform other works when solving the problem.