Performance Analysis of Cloud Hypervisor Using Network Package Workloads in Virtualization

Poovizhi, J. Mary Ramya and Devi, R. (2023) Performance Analysis of Cloud Hypervisor Using Network Package Workloads in Virtualization. In: 2023 12th International Conference on System Modeling & Advancement in Research Trends (SMART), Moradabad, India.

[thumbnail of Performance Analysis of Cloud Hypervisor Using Network Package Workloads in Virtualization _ IEEE Conference Publication _ IEEE Xplore.pdf] Archive
Performance Analysis of Cloud Hypervisor Using Network Package Workloads in Virtualization _ IEEE Conference Publication _ IEEE Xplore.pdf

Download (568kB)

Abstract

AWS (Amazon Web Services), Microsoft Azure, Cloud Zero, Kubernetes, and Google App Engine are cloud computing service providers that can manage client workloads and applications through virtualization and containerization. Computing resources are provided by large data centres that consume large amounts of Energy, contributing to global warming. Today's businesses and economies rely heavily on cloud data centres, the world's fastest-growing power users. The latest DCN has two main challenges: scalability and efficiency. A DCN's design directly impacts its scalability, but its power consumption is a significant factor in its cost. If this trend continues, many scientists predict that servers will consume more Energy over their lifetime than they cost. Large-scale infrastructure facilities such as clustering, grids, and clouds of thousands of heterogeneous computers present even more significant energy consumption problems. This paper addresses the energy consumption problem of the hypervisor shown in the cloud data centre. A single hypervisor does not exhibit the same performance on all platforms as other hypervisors in terms of power and energy consumption. Several vital conclusions presented in this paper will provide system designers and operators of data centres with valuable insights that will assist them in placing workloads and scheduling virtual machines in the most power-aware manner. Researchers in this paper provide insight into power-aware workload placement and VM scheduling for system designers and operators.

Item Type: Conference or Workshop Item (Paper)
Subjects: Computer Science > Computer Networks
Divisions: Computer Science
Depositing User: Mr IR Admin
Date Deposited: 19 Sep 2024 09:43
Last Modified: 19 Sep 2024 09:43
URI: https://ir.vistas.ac.in/id/eprint/6520

Actions (login required)

View Item
View Item