Two-Phase Virtual Machine Placement Algorithms for Cloud Computing: An Experimental Evaluation under Uncertainty

Abstract

Cloud computing providers must support requests for resources in dynamic environments, considering service elasticity and overbooking of physical resources. Due to the randomness of requests, Virtual Machine Placement (VMP) problems should be formulated under uncertainty. In this context, a renewed formulation of the VMP problem is presented, considering the optimization of four objective functions: (i) power consumption, (ii) economical revenue, (iii) resource utilization and (iv) reconfiguration time. To solve the presented formulation, a two-phase optimization scheme is considered, composed by an online incremental VMP phase (iVMP) and an offline VMP reconfiguration (VMPr) phase. An experimental evaluation of five algorithms taking into account 400 different scenarios was performed, considering three VMPr Triggering and two VMPr Recovering methods as well as three VMPr resolution alternatives. Experimental results indicate which algorithm outperformed the other evaluated algorithms, improving the quality of solutions in a scenario-based uncertainty model considering the following evaluation criteria: (i) average, (ii) maximum and (iii) minimum objective function costs.