Dynamic resource provisioning in cloud computing : a heuristic Markovian approach

Cloud computing provides more reliable and flexible access to IT resources, which differentiates it from other distributed computer paradigms. Managing the applications efficiently in cloud computing motivates the challenge of provisioning and allocating resource on demand in response to dynamic workloads. Most of investigations have been focused on managing this demand in physical layer and very few in application layer. This paper focuses on resource allocation method in application level that allocates appropriate number of virtual machines to an application which demonstrates a dynamic behavior in terms of resource requirements. By the knowledge of authors this is the first fully estimation based investigation in this field. Experimental results demonstrate that the proposed technique offers more cost effective resource provisioning approach considering cloud user demands.