An Experimental Comparison of Algorithms for Virtual Machine Placement Considering Many Objectives

Two-phase virtual machine placement algorithms for cloud computing: An experimental evaluation under uncertainty

“…In experimental results previously obtained by the authors in [5], the First-Fit Decreasing (FFD) heuristic outperformed other evaluated heuristics in average; consequently, the mentioned heuristic was the only one considered in this work for the iVMP problem in all evaluated algorithms (A1 to A5), as summarized in Table 1. This way, this paper focuses in its main contribution: the VMPr phase.…”

“…In online algorithms for solving the presented VMP problem, placement decisions are performed at each discrete time t. The formulation of the presented iVMP (online) problem is based on [5] and could be enunciated as: Given a complex IaaS environment composed by a set of PMs (H), a set of active VMs already requested before time t (V (t)), and the current placement of VMs into PMs (i.e. x(t)), it is sought an incremental placement of V (t) into H for the discrete time t + 1 (i.e.x(t + 1)) without migrations, satisfying the problem constraints and optimizing the considered objective functions.…”

“…Taking into account the particularities of the presented complex IaaS environment, the FFD heuristic presents some modifications when comparing to the one presented in [5], mainly considering the cloud service request types previously described in Section 2.1. In fact, Algorithm 1 shows that cloud service destruction, scale-down of VM resources and cloud services scale-in are processed first, in order to release resources for immediate re-utilization (steps 1-3 of Algorithm 1).…”

“…It is important to consider that online decisions made along the operation of a cloud computing infrastructure may negatively affect the quality of obtained solutions of VMP problems when comparing to offline decisions [5].…”

“…To improve the quality of solutions obtained by online algorithms, VMP problems could be formulated as a twophase optimization problem, combining advantages of online and offline formulations for IaaS environments [5]. In this context, VMP problems could be decomposed into two different sub-problems: incremental VMP (iVMP) and VMP reconfiguration (VMPr), as presented in Figure 1.…”