RESEARCH PROPOSAL: Consolidation of VMs to improve energy efficiency and Reduce CO2 emission in cloud computing environments

RESEARCH PROPOSAL: Consolidation of VMs to improve energy efficiency and Reduce CO2 emission in cloud computing environments


Cloud computing is slowly becoming a vital technology among the small local
companies, global corporate, government agents and individual users. This is because it uses
remote servers linked to the Internet to manage, store as well as refine data without having to
use a personal computer or a local server. The cloud computing technology has been widely
accepted because of its potential to address the issue of energy reduction in the wider IT field.
This is achieved by operating a group of virtue servers using a single physical server.
However, the technology providers mainly concentrate on improving performance and
improving energy consumption only comes as a collateral benefit. As such, there should be
an additional technology subsidiary to the cloud computing, with whose help the reduction of
energy in IT will be achievable. This proposal work will provide a new technology which
solves this problem. The technology is VM (virtual machine) which comes with new energy
decreasing algorithms. On comparison with other energy reducing plans, these algorithms
proved to be more efficient and effective; 4.23 % reduced the power consumption within
reduced timelines.


Various providers of the cloud computing technology are introducing new data
systems to meet the ever-increasing demand for the services that comes with the technology.
The technology is now commonplace in many aspects of the world economies. This is mainly
because it has entirely revamped the modalities with which investments on computer services
are made on a daily basis. This technology has now been placed on state-of-the-art pedestal,
given the benefits that it avails to its users: provision of massive resources to be used in the
computing investments, the price per unit of the cloud billing reports is within reach of even

the least economically able users, and there is remarkable elimination of the appraisal, earnest
and inspections fee charged at the beginning of the contract between the providers and the
users. The services provided by the technology revolve in and around the PaaS (Platform as a
Service), SaaS (Software as a Service) and the IaaS (Infrastructure as Service services).
According to a server done by Farahnakian et al. (2009), the worldwide energy
consumption by the current information centres accounted for between 2 – 3 % of the total
energy used in the year 2003.This is because the cloud data use enormous watts of power
generating tremendous amounts of heat which are potential threats to the environment. This
could mainly be as a result of the fact that the clouding technology gravitates more on
improving performance than cutting on the joules used in the data centres across the world.
Even then, while the providers look into improving energy consumption and reducing the
collateral expenses, the best approach should ensure the same performance levels by the
technology but reduce the energy costs. In other words, whichever algorithm will be chosen,
it should only reduce the joules used leaving the benefits of the cloud computing intact
(Farahnakian et al. (2009). This is the reason the research has to be done to come up with the
best approaches which will address the energy concern.
There are many approaches to the power consumption problem, but the provider
should be open-minded when making a choice between the various strategies available. Some
of the most common methods include the use of the contemporary hardware technology
which has been designed to use less energy when running the cloud technology. An example
of such technologies is the Dynamic Voltage and Frequency Scaling (DVFS) which uses two
techniques to save energy: dynamic voltage scaling and frequency-voltage scaling. These two
techniques save power by reducing the voltage and frequency of the CPU as well as other
associated peripherals and post an energy efficiency margin of up to 56 % of the total watt
use depending on the frequency of the CPU. However, the approach reduces the overall

computer performance which is why the paper propose the integration of the VM (virtual
machine) in cloud computing to reduce CO2 emission in the environment and improve
energy efficiency on the computer. This approach is quite useful since the underlying strategy
is to reduce the inherent over-provision of the hardware of equipment. The hardware’s may
be over-provisioned when the providers have high demands for resources in quick notices
where they are forced to overwork the computers especially when they are signatories to
service level agreements (SLA).
There are various researches which show that the consolidation of the VM application
will lead to the enormous reduction of power consumption in attractive rates. This is the case
where more and more VMs have been consolidated using fewer resources. In other words, the
paper focuses on having new VMs rather than optimizing the existing ones. Additionally, the
article will introduce three strategies which come with the integration of the VMs. While one
approach is founded on the first-fit decreasing algorithm, the other two are based on the best
fit decreasing algorithms to examine the general effect of the VMs on the entire information
centre power consumption. These algorithms are implemented in the CloudSim toolkit and
compared with other algorithms. It was found that the algorithms performed better about
other algorithms.
Research Rationale
The research purpose is to identify the different ways through which consolidation of
VMs can improve energy efficiency and Reduce CO2 emission in cloud computing
environments. It will be achieved by carrying out an investigation whether effecting strategic
plans in cloud computing.

Objective of the Research

The objective of this research is:

 To offer suggestions through which Consolidation of VMs to improve energy
efficiency and Reduce CO2 emission in cloud computing environments.

Research Questions

1. What are the relevant modes for consolidation of VMs to improve energy efficiency
and Reduce CO2 emission in cloud computing environments?
2. What can we learn from analyzing consolidation of VMs in an attempt to improve
energy efficiency and Reduce CO2 emission in cloud computing environments?
3. How can consolidation of VMs cloud computing environments be enhanced?

Introduction to literature review

In this chapter, hypothetical data on consolidation of VMs to improve energy
efficiency and Reduce CO2 emission in cloud computing environments will be discussed. It
is a chapter that gives the comparison between what various authors believe regarding
Consolidation of VMs to improve energy efficiency and Reduce CO2 emission in cloud
computing environments.

Various Authors Similar Views

Different authors have embarked on the topic, and some have made exceptional
progress. The efficiency and cost of the various data centres determine the overall working of
a cloud network. There are many types of research which have been done to help the
providers cut on the production cost incurred when operating data centres for the
improvement of the overall performance of the cloud computing networks. To avail state-of-
the-art services to the customers, the cost and energy of the machine resources should
considerably reduce, and this is where the VM application comes in handy. Some of the
similar works are discussed below.

Borgetto et al. proposed a spectrum of VM relocation and allocation approaches
within the various data centres available in the cloud networking which still maintain the
standard services provided to the customers. To this effect, the number pf machines should be
minimal as well as there should be reduced VM migration to reduce the power consumption.
They propose a strategy to consolidate the VM technology which will monitor the process of
management which is in line with our research.
Barroso and Holzle (2013) discusses the various techniques that will be used to save
on power consumption in the sample servers that will have been garnered during the research.
The chosen techniques are related to the research question which is all about examining the
efficacy of the VMs in energy saving. Also, the techniques will apply one of the identified
simulation models which will be using the dynamic-voltage-frequency scaling (DVFS) and
sleeping modes to cut on power consumption. According to them, DVFS reduces the amount
of energy used by reducing the voltage based on the frequency of the current flowing through
the CPU which reduces the overall power consumption. In contrast, the sleeping or
hibernating modes saves power by disabling some of the features of a computer and only
waking them up when one needs to use them. This is different from our work which will be
using the VM placement strategy.
Shuo Fang et al. (2005) presents the implementation of the three algorithms which
will be utilized in the consolidation of the VM. To start with the, they introduce the Power
Aware Best Fit Decreasing (PABFD) algorithms which are a subsidiary of the wider Best Fit
Decreasing algorithm (BFD). However, this work is centred on optimizing the current VM
application on cloud computing and not the initial VM application problem as required by our
research. Also, the PABFD does not overlap with our algorithms which allow the physical
machines to suspend and impact on the entire energy consumption of the data centres.

Yue (2007) presented the practical aspect of the FFD for the VM application
problems where it will be seen that the current sources do not relate the output of the FFD
with other algorithms which will mainly feature in this research. Again, a distinction between
this work (DVFS) and ours is that the latter allows all the physical servers to suspend and the
former does not.
Batista and Vigliotti table two algorithmic variation related to energy efficiency and
the VM consolidation at large. The first one is the Iterated-EC-Net which is founded on the
Evolutionary Computer model by way of the iterated approach. This is a modification of the
Iterated-EC which reduces the network usage as well as the power use. The second one is the
iterated-KSPM-em which is again an iterated strategy founded on solving the issue once per
server. This algorithmic variation is just a modified version of the Iterated-KSP which
increases the size of the virtual storage by 2n times its actual size to accommodate the VM
consolidation which would require high virtual capacity storage. Although our research does
not include the Iterated-EC-Net, empirical results presented in a section will prove that out
proposed algorithms will be far much better than the one present by these authors. From the
energy efficiency related to our algorithms, there is no any better approach to the problem of
energy efficiency than ours. This is because the energy efficiency concern has adequately
been addressed in our research proposal.
Farahnakian et al. (2009) proposed a useful technique for energy saving in the VM
placement by applying the Minimum Correlation Coefficient (MCC). This strategy aims at
ensuring that energy is saved and the services provided are as efficient as they were before
the application of the VM in line with the Service Level Agreement which guides our work.
This strategy will be modelled by the Cloud Sim toolkit, discussed in the next section.

The Research Gap

The gaps discussed in this section are the missing pieces of information that would
have been regarded important in conducting research. The gaps show the areas that are under-
explored or not explored at all. This includes the sample type, size, location, research method,
data analysis, and the data collection methods applied. The following chapter introduces the
gaps citing existing typologies that are as well briefly described. On this basis, the desired
results are not easily acquired by the researcher. An overview of prior studies in the field of
consolidation of VMs, the scope and theoretical approaches of the studies are provided. The
majority of the existing literature only discuss of management terms in research-based
actions. This determines the focal point of many types of research that are done by various
people. The literature in this field calls for further analysis.

To identify how this research will be conducted the qualitative and quantitative
methods should be evaluated. Brandt, et al., (2007), has the view that 'action research is
planned within the subjective custom. This proposal, however, suggests the application of
both qualitative and quantitative approaches, either separately or combined in one. The
methodology for this research will be combining both, as both quantitative and qualitative
methods are concerned with studying observable facts in computing (Han, et al., 2016).
The interventions for this research are to be based on explicit activities that are
discipline-specific and are developed after considering the need to consolidation of VMs to
improve energy efficiency and Reduce CO2 emission in cloud computing environments
(Andrea, 2014). Brandt, et al., (2007), identify making decisions based on relevant research
and conducting a literature review before designing and implementing the intervention since
this has proven to be crucial for any successful study.

The data will be analysed by using the qualitative and quantitative methods as the data
gathered will necessarily be based on written materials, questionnaires, random selection and
random assignment sampling and use of other primary and secondary sources. This will be in
an attempt to gain in-depth knowledge in Consolidation of VMs to improve energy efficiency
and Reduce CO2 emission in cloud computing environments.
The questionnaires that will be formulated will be the source of quantitative
information and will be analysed by collecting the feedback from respondents, comparing the
results and suggesting the patterns and figures laid.
Research Design
This research will be based on information gathered from different sources
concerning Consolidation of VMs to improve energy efficiency and Reduce CO2 emission in
cloud computing environments. It will, therefore, require in-depth insights from a variety of
stakeholders and in this case, authors, respondents and individuals from different sectors.