Energy Efficient Change Management in a Cloud Computing Environment

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 03 Issue: 01 | Jan-2016 www.irjet.net p-ISSN: 2395-0072
© 2016, IRJET | Impact Factor value: 4.45 | ISO 9001:2008 Certified Journal | Page 1216
Energy Efficient Change Management In
A Cloud Computing Environment
Ms. Anusuya G 1, Mr. Gopi R 2, Mr. Raja G 3, Ms. Nandhini S4
14Student, Dept. of Comp. Sci., Dhanalakshmi Srinivasan Engineering College,Tamilnadu,India
23Assistant Professor, Dept. of Comp.Sci.,Dhanalakshmi Srinivasan Engineering College,Tamilnadu,India
------------------------------------------------------------------------**********------------------------------------------------------------------------
Abstract - Achieving an energy-efficiency control and
simultaneously satisfying a performance guarantee have
become critical issues for cloud providers. The existing
work is power saving in Virtual Machine the potential
performances overheads of server consolidation were
evaluated. In this paper, three power-saving policies are
implemented in cloud systems to mitigate server idle power.
Three power-saving policies that (a) switching a server
alternately between idle and sleep modes, (b) allowing a
server repeat sleep periods and (c) letting a server Stays in
a sleep mode only once in an operation cycle are all
considered for comparison. The main objective is to
mitigate or eliminate unnecessary idle power consumption
without sacrificing performances. A server is allowed to
stay in an idle mode for a short time when there has no job
in the system, rather than switch abruptly into a sleep
mode right away when the system becomes empty. An idle
mode is the only operating mode that connects to a sleep
mode. Simulation results show that the benefits of reducing
operational costs and improving response times can be
verified by applying the power-saving policies combined
with the present algorithm as compared to a typical
system under a same performance guarantee. Our
proposed work is concentrated for the decrease the user
response time and cost using the request matching
algorithm, neg late the same user request for repeat.
Key Words: Energy-efficiency control, power-saving pol
icy, Cost optimization, response time.
1. INTRODUCTION
As cloud computing is predicted to grow, substantial
power consumption will result in not only huge
operational cost but also tremendous amount of carbon
dioxide (CO2) emissions [5], [6]. Therefore, an energy-
efﬁcient control, especially in mitigating server idle
power has become a critical concern in designing a
modern green cloud system. Ideally, shutting down
servers when they are left idle during low-load periods is
one of the most direct ways to reduce power
consumption. Unfortunately, some negative effects are
caused under improper system controls. Burst arrivals
may experience latency or be unable to access services.
There has a power consumption overhead caused by
awakening servers from a power-off state too frequently.
The worst case is violating a service level agreement
(SLA) due to the fact that shutting down servers may
sacriﬁce quality of service (QoS) [7], [8]. The SLA is
known as an agreement in which QoS is a critical part of
negotiation. A penalty is given when a cloud provider
violates performance guarantees in a SLA contract. An
efficient green control (EGC) algorithm is first proposed
for solving constrained optimization problems and
making costs/performances tradeoffs in systems with
different power-saving policies.
The challenges of controlling the service rate and
applying the N-policy to minimize power consumption
and simultaneously meet a response time guarantee are
first studied. To address the conflict issue between
performances and power-saving, a tradeoff between
power consumption cost and system congestion cost is
conducted.
An efficient green control (EGC) algorithm is
proposed to optimize the decision-making in service
rates and mode-switching within a response time
guarantee by solving constrained optimization problems.
As compared to a typical system without applying the
EGC algorithm, more cost-saving and response time
improvements can be achieved. The proposed algorithm
allows cloud providers to optimize the decision-making in
service rate and mode-switching restriction, so as to
minimize the operational cost without sacrificing a SLA
constraint.
2. SYSTEM MODEL
In this section we consider the existing system design an
d the proposed system.

2.1 Existing System
The existing design approaches are Power-Saving in
Virtual Machine, Power-Saving in Computing
Infrastructure.
Power-Saving in Virtual Machine:
Considered the problem of providing power budgeting
support while dealing with many problems that arose
when budgets virtualized systems. Their approach to VM-
aware power budgeting used multiple distributed
managers integrated into the virtual power management
(VPM) framework. By investigated the potential
performance overheads caused by server consolidation
and lived migration of virtual machine technology. The
potential performances overheads of server consolidation
were evaluated.
Power-Saving in Computing Infrastructure:
The Datacenter Energy Management project was focused
on modeling energy consumption in data centers, with a
goal to optimize electricity consumption. Their project
was focused on collecting data to define basic fuel
consumption curves. A Heterogeneity- Aware Resource
Monitoring and management system that was capable of
performing dynamic capacity provisioning (DCP) in
heterogeneous data centers.
Problem Identified:
∑ Power saving in virtual machine is not fully
achieved for power consumption in cloud based
storages.
∑ Multi User request in same time not properly
response.
∑ More emulsion of Carbon di oxide gas due to use
of server maintenance.
2.2 Proposed System
Distributed service system consists of lots of physical
servers, virtual machines and a job dispatcher. The job
dispatcher in designed system is used to identify an
arrival job request and forward it to a corresponding VM
manager that can meet its specific requirements. A server
is allowed to stay in an idle mode for a short time when
there has no job in the system, rather than switch
abruptly into a sleep mode right away when the system
becomes empty. An idle mode is the only operating mode
that connects to a sleep mode.
Benefits:
∑ Efficient way of saving the power consumption in
cloud based storages.
∑ Proper response to the entire requested user in
the cloud computing process.
∑ Avoid the over emulsion of Carbon di oxide gas
due to use of server maintenance.
∑ Practically implemented algorithm and
approaches.
3. DESIGN CONSTRUCTION
This Section consists of the following module design are
to be explained in this section.
3.1 Generation of Queuing Model
Generally Queue maintains several job request given by
the authorized users. The job request arrivals follow a
Poisson process with parameter and they are served in
order of their arrivals, that is, the queue discipline is the
first come first served (FCFS). There may have some job
requests that need to be performed serially at multiple
service stages. Then, applying phase-type distributions
allow us to consider a more general situation.
User Job
Request
Authenticate the
User Request
Job Queue
JR1JR2...JR n
Cloud Service Provider
(CSP)
Fig-1: Generation of Queuing Model
3.2 Implementation of ISN Policy
An energy-efficient control in a system with three
operating modes m = {Busy, Idle, Sleep}, where a sleep
mode would be responsible for saving power
consumption. A server is allowed to stay in an idle mode
for a short time when there has no job in the system,
rather than switch abruptly into a sleep mode right away
when the system becomes empty. An idle mode is the only
operating mode that connects to a sleep mode.
Two cases starting Busy Mode
Starting a busy mode when a job arrives in an idle mode;
Starting a busy mode if the number of jobs in a waiting
queue is more than the N value when a sleep period
expires.

ISN policy Idle Mode
Received Any Job
No
Sleep Mode
Cloud Service Provider (CSP)
Cloud Server
Jobs in queue
Busy Mode
Yes
Yes
No
Fig-2: Implementation of ISN Policy
3.3 Modeling the SN Policy
According to the switching process (directly to Sleep) and
the energy-efficient control (N policy), we have called
such an approach the “SN policy. A server switches into a
sleep mode immediately when no job is in the system. A
server stays in a sleep mode if the number of jobs in the
queue is less than the N value; otherwise, a server
switches into a busy mode and begins to work.
Cloud Server
SN policy Jobs in queue
Yes
No
Yes
No
Sleep Mode
Busy Mode
if jobs in queue > N
when ' t '
ends
Fig-3: Modeling the SN Policy
3.4 Enhance SI Policy
According to the switching process (from Sleep to Idle),
we have called such an approach “SI policy”. The step-by-
step decision processes and job flows of the SI policy. A
server switches into a sleep mode immediately instead of
an idle mode when there has no job in the system. A
server can stay in a sleep mode for a given time in an
operation period. If there has no job arrival when a
sleeping time expires, a server will enter into an idle
mode. Otherwise, it switches into a busy mode without
any restriction and begins to work.
Cloud Server
Busy ModeSleep Mode
No
Yes
Idle Mode
Sleep Time
ends
SI policy Has Job Arrives
Fig-4: Enhance SI Policy
3.5 Formulation of ECG Algorithm
An EGC algorithm is presented to solve the nonlinear
constrained optimization problem effectively. Meeting a
SLA constraint has the highest priority, followed by cost
minimization in deciding the optimal solution (M*; N*).
ISN Policy
SN Policy
SI Policy
Service Rate Energy Used
Minimum Co2
Estimation
Fig-5: Formulation of ECG Algorithm
3.6 Neg late Repeating Request
The multiple users request in add the queue for cloud
service provider and response for the based on any one
policy. But some time repeating same user and same
record in the add queue, slow down the response process
and significantly increase the cost. So we can overcome
this problem the cloud service provider check the repeat
request in the queue using Request Matching algorithm
neg late the same request for same process.

Queue
Neg late Same Request for
Same User
Request Matching algorithms
Check the Repeat User
Request in Queue
CloudServiceProvider
(CSP)
Add User Request
Fig-6: Neg late Repeating Request
4. PERFORMANCE EVALUATION
The system with the SI policy doesn’t reduce the sleep
probability as the arrival rate increases; hence, it results
in higher response times than other power-saving
policies with a higher arrival rate.
Chart-1: Response time comparison.
The Proposed power-saving policies can effectively
reduce cost, especially when an arrival rate is low. For a
cloud provider who focuses on reducing cost,
implementing the SN policy is a better choice to deal with
a wide range of arrival rates. Benefits when the startup
cost is high. As compared to a general policy, cost savings
and response time improvement can be verified.
Chart-2: Operational cost improvement rates
Although the general policy tries to keep the service rate
as low as possible, it still results in higher cost than other
policies, as shown in finally; we measure the cost
improvement rates, which calculate the relative value of
improvements to the original value instead of an absolute
value.
Chart-3: Operational cost comparisons.
Therefore, service rates are controlled at higher values w
ith power-saving policies and their idle times can be red
uced by switching into sleep modes. Conversely, the gene
ral policy focuses only on a performance guarantee and r
educes the service rate as low as possible for the purpos
e of saving operational cost. On the other hand, since a se
rver switches into a sleep mode only once in an operatio
n cycle for the SI policy, the variation among sleep proba
bilities is slight as the arrival rate increases.
5. CONCLUSION
To mitigate unnecessary idle power consumption, three
power saving policies with different decision processes
and mode switching controls are considered. A proposed
algorithm ECG allows cloud providers to optimize the
decision-making in service rate and mode-switching

restriction, so as to minimize the operational cost
without sacrificing a SLA constraint. To satisfy uncertain
workloads and to be highly available for users anywhere
at any time, resource over-provisioning is a common
situation in a cloud system. That can be resolved by ISN
policy. Even though switching between different modes of
server system is made easy but the authentication of CSP
is provided in future based on Service Rank.
ACKNOWLEDGEMENT
I would like to acknowledge my guide Mr. R. Gopi for gui
ding me and for his kind support.
REFERENCES
[1] R. Ranjan, L. Zhao, X. Wu, A. Liu, A. Quiroz, and M.
Parashar, “Peer-to-peer cloud provisioning: Service
discovery and load-balancing,” in Cloud Computing.
London, U.K.: Springer, 2010, pp. 195–217.
[2] R. N. Calheiros, R. Ranjan, and R. Buyya, “Virtual
machine provisioning based on analytical
performance and QoS in cloud computing
environments,” in Proc. Int. Conf. Parallel Process.,
2011, pp. 295–304.
[3] Y. C. Lee and A. Y. Zomaya, “Energy efficient
utilization of resources in cloud computing
systems,” J. Supercomput., vol. 60, no. 2, pp. 268–
280, 2012.
[4] A. Beloglazov, R. Buyya, Y. C. Lee, and A. Zomaya, “A
taxonomy and survey of energy-efficient data
centers and cloud computing systems,” Adv.
Comput., vol. 82, pp. 47–111, 2011.
[5] R. Buyya, C. S. Yeo, S. Venugopal, J. Broberg, and I.
Brandic, “Cloud computing and emerging IT
platforms: Vision, hype, and reality for delivering
computing as the 5th utility,” Future Generation
Comput. Syst., vol. 25, no. 6, pp. 599–616, 2009.
[6] L. Wang, G. Von Laszewski, A. Younge, X. He, M.
Kunze, J. Tao, and C. Fu, “Cloud computing: A
perspective study,” New Generation Comput., vol.
28, no. 2, pp. 137–146, 2010.
[7] R. Ranjan, R. Buyya, and M. Parashar, “Special
section on autonomic cloud computing:
Technologies, services, and applications,”
Concurrency Comput.: Practice Exp., vol. 24, no. 9,
pp. 935–937, 2012.
[8] M. Yadin and P. Naor, “Queueing systems with a
removable service station,” Operations Res., vol . 14,
pp. 393–405, 1963.
[9] W. Huang, X. Li, and Z. Qian, “An energy efficient
virtual machine placement algorithm with balanced
resource utilization,” in Proc. 7th Int. Conf.
Innovative Mobile Internet Serv. Ubiquitous
Comput., 2013, pp. 313–319.
[10] R. Nathuji, K. Schwan, A. Somani, and Y. Joshi, “VPM
tokens: Virtual machine-aware power budgeting in
datacenters,” Cluster Comput., vol. 12, no. 2, pp.
189–203, 2009.
[11] J. S. Yang, P. Liu, and J. J. Wu, “Workload
characteristics-aware virtual machine
consolidation algorithms,” in Proc. IEEE 4th Int.
Conf. Cloud Comput. Technol. Sci., 2012, pp. 42–49.
[12] G. P. Duggan and P. M. Young, “A resource allocation
model for energy management systems,” in Proc.
IEEE Int. Syst. Conf., 2012, pp. 1–3.
[13] M. Mazzucco, D. Dyachuky, and R. Detersy,
“Maximizing Cloud Providers Revenues via Energy
Aware Allocation Policies,” in Proc. IEEE 3rd Int.
Conf. Cloud Comput., 2010, pp. 131–138.
[14] Q. Zhang, M. Zhani, R. Boutaba, and J. Hellerstein,
“Dynamic heterogeneity-aware resource
provisioning in the cloud,” IEEE Trans. Cloud
Comput., vol. 2, no. 1, pp. 14–28, Jan.–Mar. 2014.
[15] A. Amokrane, M. Zhani, R. Langar, R. Boutaba, and G.
Pujolle, “Greenhead: Virtual data center embedding
across distributed infrastructures,” IEEE Trans.
Cloud Comput., vol. 1, no. 1, pp. 36– 49, Jan.–Jun.
2013.
[16] G. Wang and T. E. Ng, “The impact of virtualization
on network performance of amazon ec2 data
center,” in Proc. IEEE Proc. INFO- COM, 2010, pp.
1–9.
[17] D. A. Wu and H. Takagi, “M/G/1 queue with
multiple working vacations,” Perform. Eval., vol. 63,
no. 7, pp. 654–681, 2006.
[18] M. Zhang, and Z. Hou, “M/G/1 queue with single
working vacation,” J. Appl. Math. Comput., vol. 39,
no. 1–2, pp. 221–234, 2012.
BIOGRAPHY
Ms.G.Anusuya received the
B.Tech-IT degree from
Dhanalakshmi Srinivasan
Engineering College,
Perambalur in 2014. She is
currently doing her M.E-CSE
degree in Dhanalakshmi
Srinivasan Engineering
College,Perambalur,Tamilnadu,
India.

Energy Efficient Change Management in a Cloud Computing Environment

More Related Content

What's hot (20)

Similar to Energy Efficient Change Management in a Cloud Computing Environment (20)

More from IRJET Journal (20)

Recently uploaded (20)

Energy Efficient Change Management in a Cloud Computing Environment