Abstract— Cloud Computing mainly
deals with some company’s network of data center servers that’s accessible to
customers and various organizations. The demand in cloud has been increasing
enormously which lead to increase in the energy consumption of data centers,
which leads to different types of critical issues. This type of datacenters
consumes huge amounts of energy that leads to huge operational cost. Different
elements of cloud and the role of Cloud users are also discussed here in this
paper. A Survey is conducted in UAE and the results are provided. The Green
Cloud Simulator is used to generate results.

Keywords— Green cloud,
datacenter efficiency.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

                                                                                                                                    
I.     Introduction (Heading 1)

The emergence of
Cloud Computing is very rapidly growing by providing access to infrastructure
and services on – demand. Many companies are working with very huge datacenters
around the world. Cloud computing provides enormous amount of computer power
that may require processing of huge amount of data generated every day. Virtual
datacenters require high energy usage for its operation. So in order to reduce
this the applications of green cloud computing are used. This energy
consumption and carbon emission by cloud computing is a very great
environmental concern.

                                                                                                                                            
II.    Literature Survey

Talebi and Way
has described green computing focusing only on how to reduce energy by changing
the computer power saving modes 4. Another research done by Hemick and
Ritchard focused on how to save energy costs and increase the life span of
equipments but analysis was done from the services perspective 1. In addition
it is claimed that 80% of UAE’s ecological footprint is carbon electricity
consumptions have been number one reason in contributing to the carbon foot
print in the IT sectors 2.

As per Entesar
Al Hosani the IT department manager at EAD the green data center relies on
cooled racks, cooling subsystems and modular power which had important positive
consequences in overall capital expenditure savings and lower the operational
cost. According to the Green Grid data center power efficiency matrix the power
usage effectiveness (PUE) at EAD’s data center is 1.5 MVA which is an
acceptable range for this matrix 3.

RTA is
considered to be the first organization in UAE who signed for climate servers
computing organization to commit purchasing PCs and servers which are energy
efficient applying the green computing 6.

                                                                                                    
III.   UAE ORGANIZATIONS
USING GREEN CLOUD

There
was a simple survey to know how employees are aware about green cloud
computing. The survey had 52 responses from various cities in UAE. The
organizations belong to both public and private sectors.

RTA is now
planning to use the laptops instead of the PCsbecause it consumes energy 60%
less than the PCs and itaiming to make all desktop computers and laptops
100%qualified with the green computing in 2014 and they have a roadmap for
green computing 5

 

 The Environment Agency – Abu Dhabi guarantees all IT equipment will be
compatible with Energy Star standard. All desktop computers were replaced with
EPEAT Gold rated laptops which saved 50% in energy consumption 3

 

Fig1. Power usage for desktop and laptop

 

RTA is using
server blades to reduce carbon footprint and disk space 5

Cooling Rooms

Cooling rooms account for an upward of 65% of IT department power
usage. The power usage effectiveness at EAD’s data center is 1.5 MVA which is
in an acceptable range.

RTA datacenter
is also energy efficient. The cooling units switch off their fans while keeping
the temp optimized 5.

The practice of
Green Computing in UAE organizations

Fig2. Employee rate
on Green Cloud effectiveness

 

From the pie chart it is clear that the employees highlight that using
green cloud is more effective and it will take care of the environmental
sustainability.

 

Fig 3. Employees
willingness to practice

 

The above pie chart says that most of the employees are willing to
practice green computing in their organization.

                                                                                                                        
IV.   NEED FOR GREEN COMPUTING

Data centers are very expensive to maintain
and also affects the environment. High energy cost and carbon emission are very
high because if high electricity is needed to power and cool the servers hosted
in the data centers. Green cloud computing is used to address this problem to
manage the data center in an energy efficient manner.

 

                                                                                                  
V.    COMPONENTS OF GREEN CLOUD COMPUTING

 Load Balancing in Cloud

This distributes the excess load evenly across many different servers.
This helps in high performance improvement.

This type of load balancing helps to avoid overheating by balancing the
workload across different servers to reduce the energy consumed.

Since energy consumption is reduced we can reduce the carbon emission
to achieve green cloud computing.

Multi tenancy

This is more useful
than doing multiple copies of software installed on different servers.

Data center Efficiency

Cloud providers can improve the PUE of datacenters. Lowering level of
PUE will reduce 40% more energy efficient than traditional datacenters.

 

                                                                                                                       
VI.   GREEN CLOUD ARCHITECTURE

 

Fig 4.
Green Cloud Architecture

 

This Green cloud framework which considers the reduction of energy
consumption of clouds is described here. The different cloud users will place
their cloud service requests with the Green Broker. The Green Broker manages to
select the green cloud provider. The user requests can be software, platform or
infrastructure. The green offers have green services, pricing and time. The
carbon emission directory manages all data pertaining to energy efficiency.
Green Broker then calculates the carbon emission of all cloud providers.

 

                                                                                                            
VII.  GREEN CLOUD SIMULATION RESULTS

Green Cloud is a
simulator for energy-aware cloud computing datacenter. This gives a detailed
fine-grained modeling of the energy consumed by the data center like the
computing servers, network switches and communication connections.

 

GreenCloud is used to
develop new solutions in monitoring, resource allocation, and scheduling and
network infrastructures.

 

The following is the results generated using the
GreenCloud Simulator

 

*****************

BUILDING TOPOLOGY

*****************

 

Data center
architecture: three-tier small

Creating switches
CORE(1) AGGREGATION (2) ACCESS(3)…

Creating 36
servers…

Loading resource
specifications configuration files…

 

*********************

Creating cloud
users

*********************

Data center total
computing capacity: 144014400 MIPS

Creating 1 cloud
user(s)…

 

*********************

SIMULATION
PARAMETERS

*********************

 

Simulation time:
65.5 seconds

 

Power management
of computing servers: No

Power management
of network switches: No

 

Progress to      0 %

Progress to     10 %

Progress to     20 %

Progress to     30 %

Progress to     40 %

Progress to     50 %

Progress to     60 %

Progress to     70 %

Progress to     80 %

Progress to     90 %

******************

SIMULATION REPORTS

******************

Total tasks
submitted: 8264

DC load: 26.4%

Energy consumed by
servers: 83.2 W*h

Energy consumed by
switches: Core(51.4 W*h) Aggregation(102.8 W*h) Access(8.3 W*h) – 162.4 W*h

Total energy
consumed (servers + switches): 245.6 W*h

 

Average tasks per
server: 229

Average CPU load
per server: 0.3

Average Memory
load per server: 0.0

Average Storage
load per server: 0.0

Tasks failed on
servers: 0

Tasks failed
(rejected) on DC scheduler level: 0

 

 

Fig 5 GreenCloud
simulator results

 

                                                                                                                                                 
VIII. CONCLUSION

This paper presents a simple Simulation of
Green cloud which clearly demonstrates on how energy efficient it would be if
Green Cloud computing is used in practice. The related literatures are also
collected. The graphical representations for the survey made in UAE are also
presented. This will lead to a great environmental sustainability. Cloud
providers need to reduce the electricity demand of clouds and take major steps
by using renewable energy sources rather than just thinking of cost
minimization.

 

References

 

1 Herrick, Dan and Mark
Ritschard, Greening your Computing Technology November 2015.

2 Ecological footprint
scenario model development, April 2015.

3 EAD’s Green Data
center Wins Award 10 June 2015.

4 Talebi, Mujtaba and
Thomas Way, Methods, Metrics and motivation for Green Computer Science Program.

5 Kumar, Jaiwanth
Senior Project Manager in IT infrastructure and Services Section RTA. 28 April
2014.

6
RTA launches Green IT initiative energy efficient practices Urged, 22 Arpil
2011.