Improving Dynamic Placement of Virtual Machines in Cloud Data Centers Based on Open-Source Development Model Algorithm

Although cloud computing can provide information technology services worldwide, data centers hosting cloud applications consume a lot of energy. At the same time, the ever-increasing growth of the number of providers in the market has led to an increase in greenhouse gas emissions and operational costs. Hence, the optimal configuration of the underlying cloud infrastructure as a green cloud computing solution is needed to manage energy consumption and costs in cloud data centers. An important issue in this field is Virtual Machine Placement (VMP) considering dynamic aspects and demand fluctuations to minimize energy consumption and costs without violating the Service Level Agreement (SLA). We believe that there is a scope of improvement in migrating Virtual Machines (VMs) and subsequently deciding to shutdown hosts with little-load. This study proposes an Open-Source Development Model Algorithm (ODMA) as a meta-heuristic approach to solve the VMP problem, which is named VMP-ODMA. VMP-ODMA seeks to dynamically consolidate VMs into a minimum number of active hosts by migrating VMs over cloud data centers. VMP-ODMA can perform the placement process dynamically and periodically by finding the best sequence of VMs for migration. In addition to minimizing the number of active hosts, a load balancing strategy is included in VMP-ODMA to improve the quality of service without violating the SLA. We demonstrate the effectiveness of the proposed scheme with extensive simulations. Experimental results show that VMP-ODMA can efficiently improve system performance and outperform the best results of existing methods ranging from 11 to 27%.

Automated summary

Cloud computing provides global IT services, but hosting cloud applications consume a significant amount of energy. The increasing number of providers has led to higher greenhouse gas emissions and operational costs. Therefore, optimizing the cloud infrastructure configuration is crucial to manage energy consumption and costs in data centers. Virtual Machine Placement (VMP) is a key issue in this field, aiming to minimize energy consumption and costs while adhering to the Service Level Agreement (SLA). This study proposes a meta-heuristic algorithm called VMP-ODMA, which dynamically consolidates VMs into a minimum number of active hosts through migration in cloud data centers. Extensive simulations demonstrate that VMP-ODMA efficiently improves system performance, surpassing existing methods by 11% to 27%.