Placement methods of Virtual Machines in servers

In a data center, the amount of data to be stored and processed is increasing more and more, given technological innovation and the diversity of services offered by companies in the cloud. Therefore, there is a need to know how to use server resources in a way that does not waste energy in data centers while optimizing SLAs (Service Level Agreement), and migration of virtual machines (VMs). This paper presents an algorithm named FSS-VM (Fuzzy Soft Set based-Virtual Machine) which consists of placing VMs in physical machines (PMs) taking into consideration the use of CPU, memory, RAM and correlation values. Another algorithm named DRL-VM is presented to make the placement of VMs in servers in an optimal way using heuristics, and taking into consideration software failures, the energy consumed by servers in the data center, and the collocation interference between VMs. As a result, FSS-VM has shown its performance on energy consumption and improvement of the QOS (Quality Of Service) comparing to other methods. The other method which is DRL-VM consists in making the placement of the VMs by choosing the optimal method among dot product, first fit, norm2, $\rho$-greedy, $\eta-$greedy and $\omega$-greedy. Virtual machine placement method is a process that aims to improve the treatment complexity of user’s requests, and optimize energy consumption, both methods need to be enhanced to get better results.