MOM-VMP: multi-objective mayfly optimization algorithm for VM placement supported by principal component analysis (PCA) in cloud data center

Cloud computing provides consumers and organizations with shared pools of resources for data storage and processing and its optimization is essential as 98% of the allocated resources have been utilized only 86% of 98%. Hence, we carry out optimization to automatically allocate resources. In a cloud data center, Virtual machine placement is essential, and choosing the optimal physical machine to host the virtual machine is a critical step. The efficacy of the Virtual machine placement strategy has a considerable impact on cloud computing efficiency. Today, cloud computing optimization is needed for business goals and competition in the digital landscape for cost reduction (20-28%) and Energy consumption (16-22%), improving performance (30-42%) and scaling (12-14%) to meet changing business needs. Virtual machine placement optimization problems are a class of problems that arise in cloud computing when allocating resources to virtual machines across a set of physical machines or hosts. The goal is to optimize resource utilization (12-16%) while satisfying various constraints, such as performance requirements, availability, and energy efficiency than non-metaheuristic optimization techniques. Several virtual machine placement optimization problems include placement, consolidation, migration, and scheduling. Virtualization facilitated by virtual machine placement and migration meets the ever-increasing demands of a dynamic workload by transferring virtual machines inside cloud data center. Many resource management goals, including power efficiency, load balancing, fault tolerance, and system maintenance, are aided by virtual machine placement and migration. To propose a multi-objective Mayfly virtual machine placement algorithm with a massive cloud data center with different and multi-dimensional resources to handle these issues. A multi-objective, dynamic virtual machine placement strategy simultaneously reduces resource wastage, overcommitment ratio, migration time, service level agreement violation, and energy consumption. This paper presents a dynamic, multi-objective virtual machine placement strategy in cloud data centers based on overcommitment resource allocation to influence Virtual machine Physical machine mapping and achieved an increase in the range of 12.5-14.89% in allocation than the existing works. We validated our method by conducting a performance evaluation study using the CloudSim tool. The experimental results demonstrate that this article improves resource usage while reducing energy consumption, makespan, over-commitment, and physical machine overload.