Multi-Objective Energy-Efficient Virtual Machine Consolidation Using Dynamic Double Threshold-Enhanced Search and Rescue-Based Optimization

Multi-Objective Energy-Efficient Virtual Machine Consolidation Using Dynamic Double Threshold-Enhanced Search and Rescue-Based Optimization

Sweta Singh, Rakesh Kumar, Udai Pratap Rao
DOI: 10.4018/IJSSCI.315006
OnDemand:
(Individual Articles)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

The popularization of the cloud and its need to solve complex engineering application have alarmed energy and environmental concerns among the researchers. Achieving energy efficiency has become one of the most essential aims of the data center, offering more services with minimal energy consumption (EC). VM consolidation aims at adjusting the VMs to fewer PMs by live migration of VMs and then switching off the inactive servers, achieving energy efficiency. However, uncontrolled consolidation could violate the SLA. The paper contributes by considering the optimization problem targeting the EC and the number of VM migrations. Dynamic double threshold with enhanced search and rescue (DDT-ESAR) optimization has been introduced utilizing two thresholds; the first value defines the upper and lower bound for host classification, whereas the other is used to make migration decision. For migration, ESAR has been adopted for the most appropriate PM- VM mapping. The experimental analysis proves the efficiency where EC is computed to be 0.384kWh, SLA violations to be 6.33% and 64 number of migrations.
Article Preview
Top

Introduction

As defined by National Institute of Standards and Technology (NIST), cloud computing is termed a 5-4-3 model (Mishra et al., 2020; Abdelsamea et al., 2014), enabling convenient and on-demand network services by providing unlimited access to a pool of resources such as storage, network bandwidth, CPU, etc. These computing resources are sharable among multiple cloud users, which could be rapidly provided and released with minimal management and user-provider interaction. The cloud comprises several data centers geographically distributed at different locations providing on-demand faster services to the users over the Internet (Mohanapriya et al., 2018; Wu et al., 2016). Cloud computing services classified are as follows viz; Software as a Service (SaaS) incharge of promoting remote software services and applications; Platform as a Service (PaaS) incharge of providing a layout or platform for aplication development and deployment and last Infrastructure as a Service (IaaS) is responsible for providing user access to the computing resources namely storage, network services e.g. bandwidth, VMs and physical servers. IaaS has two main benefits over traditional distributed computing environment i.e on-demand provisioning and elasticity, making it most suitable for workflow application (Chakravarthi et al., 2020). Organizations use their own platforms and applications within a service provider’s infrastructure. These resources are hosted on the Cloud Service Provider (CSP) or third party infrastructure such as Amazon Web Services, Microsoft Azure or Google Cloud.

The main reason behind the popularization of the cloud is the scalability, flexibility, operational and front-end capital elimination with the delivery of infrastructure, software and platform to end-users, where they could lease and release the services requested and pay as per their usage (Mohanapriya et al., 2018). The service providers offer the services to the users by establishing data centers worldwide. These data centers, thus provide an illusion to the users of providing unlimited resources over the Internet by the big companies, including Amazon, Microsoft, Google, IBM, etc (Varasteh and Goudarzi, 2015). Each cloud data center constitutes a set of physical host machines or servers arranged as racks that run or host millions of VMs based on customers’ requests. This feature of cloud computing has driven considerable attention from academicians and industrial organizations to carry out their complex scientific and engineering computations over the cloud. This dependency on cloud for large scale scientific applications such as limnography, earthquake science, research, ocean science, etc., have raised environmental concerns (Mohanapriya et al., 2018). Big data analysis, voluminous data, growth of Internet, need of online data storage, backup, recovery, product testing and deployment, antivirus applications, e-commerce and e-governance applications, cloud application in education sector including e-learning, online distance-learning platforms, on-demand entertainment and cloud application in medical field are some application which has exponentially increased the dependency over the cloud and its data centers for computation, contributing to raised environmental concerns. The large-scale deployment, big data and dependency on the cloud, have also raised security and privacy concerns in the cloud environment (Gupta et al., 2018) As more will be deployment, more will be power/ energy demand and more it will leave carbon footprints on the environment.

Complete Article List

Search this Journal:
Reset
Volume 16: 1 Issue (2024)
Volume 15: 1 Issue (2023)
Volume 14: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 13: 4 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing