International Journal of Engineering Innovation & Research

Volume 3, Issue 4, ISSN: 2277 – 5668

Round Robin Approach for Better VM Load Balancing

in Cloud computing
Bhavana Bidarkar Shakheela Attikeri Ramya S. Pure
Assistant Professor, Department of CSE Assistant Professor, Department of CSE Assistant Professor, Department of CSE
GNDEC, Bidar VTU Belgaum, India GNDEC, Bidar VTU Belgaum, India GNDEC, Bidar VTU Belgaum, India
Email: [email protected] Email: [email protected] Email: [email protected]

Abstract – Cloud computing is known as digital service

delivery over the Internet by several applications which are
carried out by computer systems in distributed datacenters. .
It is a fast growing area in computing research and industry
today. With the advancement of the Cloud, there are new
possibilities opening up on how applications can be built and
how different services can be offered to the end user through
Virtualization, on the internet. There are the cloud service
providers who provide large scaled computing infrastructure
defined on usage, and provide the infrastructure services in a
very flexible manner which the users can scale up or down at
will. The establishment of an effective load balancing
algorithm and how to use Cloud computing resources
efficiently for effective and efficient cloud computing is one of Fig.1. shows a simple cloud environment.
the Cloud computing service providers' ultimate goals. It
supplies a high performance computing facilities which allow The virtualization forms the foundation of cloud
shared computation and storage over long distances. To technology where virtualization is an emerging technology
properly manage the resources of the service provider we that separate computing function and technology
require balancing the load of the jobs that are submitted to implementation from physical hardware. Cloud computing
the service provider. Load balancing is required as we don’t is the virtualization of computer program through the
want one centralized server’s performance to be degraded. In
internet connection rather than installing application on
available Virtual Machine Load balancing policies limitation
of cloud is that they don’t save the state of the previous
everywhere. Using virtualization user can access server or
allocation of virtual machine to a request from the user and storage without knowing specific storage detail.
the VM Load balancing algorithm requires execution each Virtualization can be applied to many types of computer
time a new request for VM allocation received from user. In resources: Infrastructure such as storage, network,
this paper, we present Round Robin Algorithm for better computer (as CPU, memory), platform (such as Linux,
load balancing in cloud Environment. windows OS), and software as a services.
Actually Cloud computing offers three layers of
Keywords – Cloud Computing, Load Balancing, Virtual abstraction
Machine, CloudSim, Data centre, Round Robin Algorithm.
Infrastructure as a Service (IaaS),
Platform as Services (PaaS),
I. INTRODUCTION Software as Services (SaaS)

Cloud computing is the most recent emerging epitome

to turn the vision of “computing utilities” into a real world.
Cloud computing is an emerging technology with advance
feature that focuses on the way in which we design
computing system, develop application and building
software with advancement. It is based on dynamic
provisioning concepts, which is applied on the services,
also to compute capability, storage, networking &
Information Technology (IT) infrastructure. In cloud
computing resources are made available through the
Internet and offered on a pay-per-use basis in anywhere
from Cloud computing service broker. Cloud computing
can help by providing massive computing power,
unlimited storage capacity, less maintenance cost,
availability of useful web-services etc. Fig 2: Virtualization of Cloud Computing

Copyright © 2014 IJEIR, All right reserved

499
 International Journal of Engineering Innovation & Research
Volume 3, Issue 4, ISSN: 2277 – 5668

II. RELATED STUDY

Cardellini et al. [1] In this paper he has Considered
different categories of web applications, and evaluate how
static, dynamic and secure web service requests affect
performance and quality of service at distributed web sites
Levy et al[2] Presents an architecture and prototype
implementation of performance management system for
cluster based web service. H.Aydin et.al[3] have proposed
minimizing the energy consumption and subsequently the
cost for the static system. Buiya et. al.[4] Have proposed
the architecture for market oriented allocation of resources
within clouds. Xinhui et. A[5]have proposed total cost of
ownership of cloud computing using a suite of metrics and
formulas. K.Mukherjee and G. Sahoo[10] Have proposed
a frame work for achieving better load balancing and job
scheduling in Cloud environment. K. Mukherjee and
Fig.3. The framework of tasks allocation process G.Sahoo [6][7][8][9] Have conducted research on different
issues of cloud computing regarding its security aspects,
SaaS provides different types of applications as a mathematical models and others.
Service for the end user. Middleware Services. Paas )
typically makes use of dedicated API’s to control the III. PROBLEM STATEMENT
behaviour of a server hosting engine which executes and
replicates the execution according to user request eg. With the ever increasing number of cloud service
force.com, Google App Engine. IaaS provides the providers, rising competition, and the growing popularity
hardware infrastructure of cloud consisting of physical of cloud, it is extremely important for cloud providers to
machines like clusters, datacenters etc to provide resources improve their services by improving the operational cost
in order to meet the consumer’s request on demand. To (response time and the processing time of requests).
provide massive computing and storage resources in cloud Hence, to satisfy the service level objectives (SLOs) [11],
environment here needs load balancing. Load balancing is it is very important to improve the low level requirements
the mechanism that decides which requesting nodes/client such as the CPU and memory requirements and the
will use the virtual machine and which requesting response time of the Userbase requests. One of the ways to
machines will be put on hold (figure 4).Load balancing is improve this is by providing efficient VM load balancing
also required to minimize the cost of machine and policies. However, there are only a limited number of VM
maximize the profit for the service being offered. Hence load balancing policies available at present. Also, in the
with these issues in mind we here in this paper propose an existing VM load balancing strategies, an algorithm has to
algorithm for rescheduling named round Robin algorithm. be run every time a new request for VM allocation is
received from the Userbase. This is because the existing
VM load balancing strategies do not save the previous
allocation state of a VM to a request from a given User
base. However, in case of the proposed strategy, a
hashmap is used to store the entries for the last VM
allocated to a request from a given Userbase. Thus, when a
request is received from a Userbase, if an entry for the
given Userbase exists in the hashmap and if the particular
VM is available, then there is no need to run the VM
allocation algorithm, which in turn saves a significant
amount of time.

IV. LOAD BALANCING

Fig.4. Load balancing in cloud environment
Load Balancing is a method to distribute workload on
In IaaS, the physical resources can be split into a the multiple computers or a computer cluster through
number of logical slices called Virtual Machine (VM’s). network links to achieve optimal resource utilization for
All VM Load Balancing methods are designed to maximizing throughput and minimizing overall response
determine which Virtual Machine is assigned to the next time. Load Balancing is used for avoiding too much
cloudlet task units. These VM are modelled using different overload on the resources and dividing the traffic between
tools Cloudsim- Simulation framework for its allocation to servers and data. Data can be sent and received without
the application. maximum delay. Load Balancing is used for minimizing
Copyright © 2014 IJEIR, All right reserved
500
 International Journal of Engineering Innovation & Research
Volume 3, Issue 4, ISSN: 2277 – 5668

the total waiting time of the resources. In cloud computing Initialize hash map with no entries;
load balancing are uses for balancing the load on virtual While (new request are recived by the Data Centre
machine and cloud resources. Controller)
Do
{
Data Center Controller queue the requests;
Data Centre Controller removes a request from
the beginning of the queue;
If(hash map contain any entry of a VM corresponding to
the current requesting user base && VM allocation status
== AVAILABLE)
{
The VM is reallocated to the user base request;
}
Else
{
Fig.5. Cloud load balancer
Allocate a VM to the user base request using Round Robin
Algorithm;
Existing Scheduling Algorithm in Cloud Computing
Update the entry of the user base and the VM in the hash
Virtual machine enables the abstraction of an OS and
map and the VM state list;
Application running on it from hardware there are three
}
types of VM Load Balancer that is Round Robin,
}
Throttled and active monitoring load balancing algorithms.
}
1. Round Robin Load Balancer – uses a simple Round
Robin algorithm to allocate VMs.
2. Active Monitoring Load Balancer – this version load V. RESULT ANALYSIS
balances the tasks between available VMs in a way that
evens out the number of active tasks on each VM at any The proposed algorithm (i.e. the Round Robin Load
given time. Balancer Algorithm) implemented for simulation. Java
3. Throttled Load Balancer – this ensures that only a pre- language is used for implementing VM load balancing
defined number of Internet Cloudlets are allocated to a algorithm. Table 3 shows the result based on Round Robin
single VM at any given time. If more request groups are VM Load Balancing algorithm for overall response time
present than the number of available VMs at a data center, of the cloud. In this min(ms) time, max (ms) time to
some of the requests will be queued until the next VM different number of virtual machines are analysed. Table 4
becomes available. shows the result based on Round Robin VM Load
One of the common problem with the above algorithms Balancing algorithm for Data Centre processing time of
is that they do not save the state of the previous allocation the cloud. In this min (ms) time, max(ms) time to different
of a VM to a request from given Userbase. As such, every number of virtual machines are analyzed.
time a request is received from the same Userbase the Table 1: For overall response time for Round Robin Load
algorithm needs to be run again, which increases the total Balancing
response of the requests. In this paper, we have addressed
this problem for the Round Robin VM load balancer.
The Proposed Algorithm- Round Robin VM Load
Balancing.
The proposed algorithm is an improvement over the
Round Robin VM Load Balancing algorithm. The Round
Robin algorithm does not save the state of previous
allocation of a VM to a request from a given user base
while the same state is saved in RR VM load balancer.
The Round Robin VM Load balancer maintains two data Table 2: For Data Centre processing time for Round
structure which is discussed below. Robin Load Balancing
o Hash Map- in which it stores the entry for the last VM
allocated to a request from a given user base.
o VM State List- this stores the allocation status (i.e. busy
available) of each VM.
ALGORITHM is-
Round_Robin_Load_Balancing ()
{
Initialize all the VM allocation status to
AVAILABLE in the VM state list;
Copyright © 2014 IJEIR, All right reserved
501
 International Journal of Engineering Innovation & Research
Volume 3, Issue 4, ISSN: 2277 – 5668

Analysed result shows that Round Robin Load REFERENCES

Balancing consumes less time for overall response time
and data centre processing time over Round Robin [1] V. Cardellini, E. Casalicchio, and M. Colajanni, “A performance
method. When number of virtual machine is increases then study of distributed architectures for the quality of Web
services”, in Proceding 34th Annual Hawaii International
it takes more time for over all response time and data
Conference on System Sciences,2001.
centre processing time. It decrease the problem of [2] R.Levy, J Nagarajaro, G. Pacifici, M. Spreitzer , A. Tantawi, and
deadlock and server overflow in cloud environment by the A. Youssef, “ Performance management for cluster based Web
new service broker policy in virtual machine that is Round services”, in Proceding IFIP/IEEE 8th International Symposium
on Integrated Network Management, 2003, pages 247-261,.
Robin VM load balancing algorithm.
[3] H. Aydin, R.G Melhem, D Mosse, and P. MejiaAlvarez.“Power-
Aware Scheduling for Periodic Ral-Time Task”, IEEE
Transactions on Computers, May 2004, , 53(5)
[4] Rajkumar Buyya, Chee Shin Yeo and Srikumar
Venugopal,“Market-Oriented Cloud Computing: Vision, Hype,
and Reality for Delivering IT Services as Computing
Utilities”,The 10th IEEE International Conference on High
Performance Computing and Communications, IEEE Computer
Society, 2008, pages 5- 13.
[5] Xinhui Li, Ying Li, Tiancheng Liu, Jie Qui, Fengchun
Wang“International Conference on Cloud Computing” IEEE
Computer, 2009 , pages 93-100
[6] K.Mukherjee and G.Sahoo “ Mathematical Model of Cloud
computing framework using Fuzzy Bee Colony optimization
Technique”, International Conference on Advances in
Computing, Control and Telecommunication Technologies,
IEEE Xplorer, 2009 International Journal of Computer
Applications (0975 – 8887)Volume 9– No.9, November 2010
[7] K.Mukherjee and G.Gahoo “ A Secure Cloud Computing”,
International Conference on recent trends in Information,
Telecommunication and computing, IEEE Xplorer, 2010
[8] K.Mukherjee and G. Sahoo “Performance Analysis of Cloud
Computing using Multistage Ant System” ,International Journal
of Computer Applications(0975- 8887), volume 1- No 20, 2010
[9] K.Mukherjee and G.Sahoo, “Cloud Computing: Future
Framework for e-Governance”, International Journal of
Computer Applications (0975- 8887), volume 7, No 7, October
2010.
[10] K.Mukherjee and G. Sahoo, “A framework for achieving better
load balancing and job scheduling in Grid environment”,
International Journal of Information Technology and knowledge
Management, volume 2, No - 1, pp- 199- 202, January-
June,2009.
[11] Qi Zhang, Lu Cheng, Raouf Boutaba, Cloud computing: state-of-
the-art and research challenges, Journal of Internet Services and
Applications, Springer, May 2010, Volume 1, Issue 1, pp 7-18.
[12] Amandeep Kaur sidhu1 and Supriya Kinger2, “Analysis of Load
Balancing Techniques in Cloud Computing”, International
Journal of Computers & Technology, volume 4, No. 2, March-
April 2013, pg 737- 741.
[13] Pooja Samal1 and Pranati Mishra2, “Analysis of Variants in
VI. SUMMARY AND CONCLUSION Round Robin Algorithms for Load Balancing in Cloud
Computing”, (IJCSIT) International Journals of Computer
The total work of this thesis work is done for five Science and Information Technologies, Volume 4 (3), 2013, pg.
servers. A virtual machine is a virtual form of computer no. 416- 419.
[14] Bhatiya Wickremansinghe1, Rodrigo N. Calheiros2and Dr.
hardware within software. Virtual machine is a software Rajkumar Buyya3, “CloudAnalyst: A CloudSim- based Visul
implementation that executes programs as if they were Modeller for Analysing Cloud Computing Environments and
actual physical machines. We also gives the detailed Applications”, IEEE Computer Society, 2010, pp. 446-452.
review on existing scheduling algorithm. The proposed [15] Jaspreet Kaur, “Comparison of load balancing algorithm in a
Cloud”, International Journal of Engineering Research and
Round Robin VM Load Balancing and existing Round Applications (IJERA), vol. 2, Issue 3, May- June 2012, pp. 1169-
Robin algorithm implemented Java language for 1173.
implementing VM scheduling algorithm in CloudSim [16] Syed Tauhid Zuheri1, Tamanna Shamrin2 and Rusia Tanbin3,
toolkit. Assuming the application is deployed in one data Firoj Mahmud4, “An Efficient Load Balancing Approach in
Cloud Environment by using Round Robin Algorithm”,
centres having virtual machine (with 2048 Mb of memory International Journal of Artificial and Mechatronics, volume 1,
in each VM running on physical processor capable of issue 5, 2013, pp 96-99.
speed of 1000 MIPS). These experimental results show
that Round Robin VM Load Balancing method improves
the performance by consuming less time for scheduling
virtual machine.
Copyright © 2014 IJEIR, All right reserved
502