Host Overloading Detection Based on EWMA Algorithm in Cloud Computing Environment

Abstract

Energy consumption and Service Level Agreement (SLA) in Cloud computing environment are important cloud management issues. Dynamic consolidation of the Virtual Machines (VMs) need effective and efficient distribution for VMs migration to hosts in data center. The process of VMs migration needs to evaluate host capability, VM placement and reallocation, which satisfy SLA criterions under a flexible service plan. Therefore, the plan is to select effective resource allocation to achieve cost minimization, reduce energy consumption and avoid SLA violation. We proposes Exponentially Weighted Moving Average (EWMA) algorithm to detect overloaded hosts, which deals with dynamic consolidation of VMs based on an analysis of historical data of the resource usage by VMs. It increases the accuracy in calculation of the upper threshold for host overloading and consequently increases accuracy in identification to deal with VMs migration issue.