Figure 1 above illustrates the component of a typical data centre infrastructure (Andrae, 2019). Data Centres (DC) are significant consumers of energy, accounting for a large portion of the world's electricity consumption (Koronen et al., 2020). However, the continual expansion in data processing, storage facilities. and transmission raises concerns about the IT sector's projected energy need (Andrae and Edler 2015; Andrae 2019; Masanet 2020). This has led to increased interest in developing more energy-efficient data centres, known as "green data centres." (da Silva et al., 2022) Full virtualization is a strategy that involves installing a virtualization framework (hypervisor) on a physical or host machine to allow numerous virtual machines (VMs) to function simultaneously on the same physical server or computer hardware (Pereira et al., 2020; Nair and Santha, 2023) Every single virtual machine, or VM, has a distinctive Operating system (OS) as well as software and is totally separated from the others. This method is widely utilised in enterprise data centres as it is the most advanced type which allows different software and operating systems to operate on the same physical server (Alkhatib et al., 2021) Figure 3 Overview of the Effectiveness Data Centre Virtualization in a Green Data Centre Figure 5 Para-Virtualization actual hardware components of the primary system in para- virtualization (Tiburski et al., 2021). This method, while providing higher speed and _ scalability than full virtualization, necessitates more changes to the guest OS. Figure 5 below shows an overview of para-virtualization technique with the guest OS which runs with the virtualization framework known as hypervisor. Each virtualization approach has its own strengths and weaknesses, and the choice of approach depends on the specific use case and requirements of the system. Full virtualization is ideal for running multiple operating systems on a single physical server, para-virtualization provides better performance for specific workloads, and containerization is ideal for running distributed applications in a scalable and efficient manner. TABLE 1 HYPERVISOR TYPE FOR VIRTUALIZATION (SALAGRAMA AND BIBHU 2022) Figure 7 Forecast of servers and Data Centre Energy Usage A comparison of the many projections in the figure above illustrates that the likely future increase of data centre utilisation of energy has an extensive variation. The power consumption of data centres can be kept relatively constant in the "best-case" scenario. According to Andrae/Edler's 2015 "expected" figures as shown above, the yearly energy consumption could reach approximately 3,000 billion kWh/a if Moore's Law is no longer true and data centre performance dramatically i improves. Figure 9 Overview of Server Consolidation Live migration is regarded as the most effective method of lowering energy usage by minimising the number of active servers in the data centre (Ala’Anzy and Othman, 2020). Figure 9 below presents a server consolidation summary by using four servers as an example for migrating virtual machines and power off inactive servers. To minimise SLA breaches and maintain QoS, virtual machine live migration can migrate VMs between servers with significantly less system downtime leading to high availability (Ala’Anzy and Othman, 2020; Tran et al., 2022). In essence, server consolidation involves leveraging a more alluring feature for the server consolidation approach — Virtual Machine live migration while putting numerous virtual machines on a smaller number of physical servers to improve resource utilisation and lower energy consumption. (Helali and Omri, 2021) As a result, this functionality makes it possible to move a VM in operation from one physical hardware server to another without pausing the service.
