Abstract
Virtual machine placement (VMP) optimization is a crucial task in the field of cloud computing. VMP optimization has a substantial impact on the energy efficiency of data centers, as it reduces the number of active physical servers, thereby reducing the power consumption. In this paper, a computational intelligence technique is applied to address the problem of VMP optimization. The problem is formulated as a minimization problem in which the objective is to reduce the number of active hosts and the power consumption. Based on the promising performance of the grey wolf optimization (GWO) technique for combinatorial problems, GWO-VMP is proposed. We propose transforming the VMP optimization problem into binary and discrete problems via two algorithms. The proposed method effectively minimizes the number of active servers that are used to host the virtual machines (VMs). We evaluated the proposed method on various VM sizes in the CloudSIM environment of homogeneous and heterogeneous servers. The experimental results demonstrate the efficiency of the proposed method in reducing energy consumption and the more efficient use of CPU and memory resources.
Highlights
Cloud computing has transformed traditional IT into a promising paradigm in which the cloud is used as a utility [1,2]
We develop a grey wolf optimization (GWO)-based method for addressing the Virtual machine placement (VMP) optimization problem as a combinatorial problem
After the randomly distributed solution step has been completed for all wolves in the pack, GWO-VMP generates new solutions by updating the existing solutions for every wolf to search for an optimum distribution of virtual machines (VMs) on physical machine (PM)
Summary
Cloud computing has transformed traditional IT into a promising paradigm in which the cloud is used as a utility [1,2]. Virtual machine placement (VMP) optimization is a process of selecting the minimal number of PMs that can supply the required resources for hosting a specified number of VMs with the lowest possible power consumption. VMP optimization increases the energy efficiency and resource utilization of cloud data centers by introducing a solution in which VMs are hosted in the minimal number of active PMs. VMP optimization can prolong the stability of the datacenter before the reallocation of VMs becomes an urgent issue [12,13]. We develop a GWO-based method for addressing the VMP optimization problem as a combinatorial problem. The proposed method reduces the energy consumption of cloud computing by allocating VMs into the minimal number of active PMs. The proposed work formulated the VMP optimization problem as discrete and binary GWO problems. The proposed methods performed competitively compared to the state-of-the-art methods
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have