A Cost-Effective Criticality-Aware Virtual Machine Placement Approach in Clouds

Abstract

Virtual Machine (VM) placement is a key issue for addressing problems improving resources utilization, increasing customer satisfaction, reducing implementation cost, minimizing power consumption. Existing works on availability-aware VM placement deal with VM allocation task based on taking VM resource demand as constraints. In this paper, we present a novel Cost-Effective Criticality-Aware VM Placement (CECAVMP) approach based on criticality feature of VM, which has not been considered in previous works. A critical server free capacity controlling method is proposed to make a trade-off between availability and cost. We simulated VM placement using CECAVMP and other compared algorithms in different scale clouds. The experimental results show that CECAVMP keeps cloud application in higher availability range without loss of VM request satisfaction.