Providing IaaS resources automatically through prediction and monitoring approaches

Abstract

A cloud computing infrastructure management is proposed in this paper, which consists of two approaches that facilitate the provisioning of computing resources in a self-adaptive virtualized environment. Resource allocation is employed to predict the future of workload management and to employ a self-adaptive approach by using computational agents to monitor the Virtual Machines (VMs). The paper also includes the Return on Investment (ROI) formula that deals with the relationship between the prices for the Infrastructure-as-a-Service (IaaS) contracted by the customer and the effective use of this service. The experimental results show a significant improvement when self-configuration is used with agent-based computational modeling in contrast with the self-configuration based on prediction for future workload.