e-ISSN: 2582-5208

International Research Journal of Modernization in Engineering Technology and

Science
Volume:04/Issue:01/ January-2022 Impact Factor- 6.752 www.irjmets.com

CHALLENGES OF CLOUD COMPUTE LOAD BALANCING ALGORITHMS

Mr. Gopala Krishna Sriram*1
*1 Software Architect, EdgeSoft Corp, McKinney, TX USA

ABSTRACT
Cloud computing reshaped the modern world by offering solutions to all problems faced by organizations. Cloud
computing provides computational services for users in a pay-per-use fashion. Hence, they do not need to
purchase these resources for their use. With the help of cloud computing services, users can use Software,
Hardware, infrastructure, and many other computational resources without taking the pain of their maintenance.
Because of their versatile services, cloud computing service providers face many challenges such as security,
privacy, quality of service and load balancing. In this research, we focus on load-balancing issues and investigate
significant challenges in the load-balancing domain of cloud computing. At the start, we introduce domain
knowledge related to cloud computing technologies and then briefly discuss load balancing techniques. At last, we
present some potential challenges in load balancing techniques.

Keywords: Cloud Computing, Load Balancing, Algorithms, Load Distribution, Heterogeneous Nodes, Single Point of
Failure, Etc.

I. Introduction

The excellent Cloud computing services reshape the world of Information Technology. Not only IT companies but
also other business organizations and individual's leverages cloud computing technology. The more core concept
of cloud computing is to provide users with computing services in a pay-per-use fashion [1]. Cloud allows users to
use a shared pool of computing resources. These resources are hardware-based; they include Software, network,
platform, and many other valuable resources. Several companies provide cloud services; the top of them are
Amazon Web Service (AWS), Google Cloud, Microsoft Azure, RackSpace, IBM Cloud; the primary goal of the cloud is
to provide distributed services to improve throughput and performance. Users task from all around the world are
distributed among widely spread data centers to improve speed and performance for those jobs. Cloud computing
distributed services must gain more popularity than other distributed technologies like grid computing peer-to-
peer computing. One reason for this popularity is the vast cloud services[2]. Cloud computing can be categorized
on two bases, one is location-based, and the other is offered services. The base of the location cloud can be
categorized as public, private and hybrid cloud. The cloud has three basic categories based on services provided,
but immense extended categories exist.

The essential services provide Software as a Service (SaaS), Platform as a Service (PaaS), Infrastructure as a
Service (IaaS). There can be many other services such as Hardware as a Service (HaaS), Communication as a
Service (CaaS), Database as a Service (DaaS), broadly we can tell that XaaS(anything as a service)

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[1]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and
Science
Volume:04/Issue:01/ January-2022 Impact Factor- 6.752 www.irjmets.com

Fig 1: cloud computing service model

Since cloud computing offers immense services, they need Quality of Services (QoS) monitoring to evaluate their
services for fulfilling user requirements. The cloud faces many challenges during this process, such as load
balancing, performance analysis, monitoring, throughput, response time, security, and privacy. Load balancing is a
significant challenge to avoid overloading or underloading virtual machines during service provision. Cloud
companies need to identify these potential challenges and find effective load balancing techniques to improve their
quality. This research has the following objectives
 State of the art related to different available load balancing techniques
 Taxonomies different load balancing techniques and investigate challenges of existing load balancing
techniques
 Point out different research areas for future researchers.

Rest of this paper, we will provide domain knowledge of our target study area, which is load balancing in cloud
computing. Afterwards, we will present currently available load balancing techniques and categorize them. At last,
we will propose a research area in load balancing in cloud computing.

II. Background
This section will provide preliminary knowledge about load balancing techniques, why we need load
balancing, the used technologies for load balancing and other domain knowledge.
Load balancing techniques are used to balance the load on virtual cloud machines so that each machine
can work equally according to its capacity. By load balancing, tasks can be equally distributed to each
virtual machine and hence get the best performance from each. With the help of load balancing
techniques, cloud services providers can manage workload on virtual machines. Virtual machines are
the core of cloud computing infrastructure. Cloud service providers provide hardware access to users
with the help of virtual machines. Virtual machines can act as one or more servers. One machine can act
on more than one server.
In contrast, more than one machine can also manage one server machine[3]. Furthermore, through load
balancing, cloud providers ensure that in case of failure of one machine to complete, operations will not
disturb. In addition to this, load balancing also provides scalability for those applications whose size can
grow or shrink with time. Since scalability is one of the significant characteristics of cloud computing, it
cannot be achieved without proper load balancing techniques[2]. The other responsibilities of load
balancing techniques are to provide energy efficiency, green cloud by reducing carbon emission, QoS
requirement fulfilment and other services[4, 5].
www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science
[2]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and
Science
Volume:04/Issue:01/ January-2022 Impact Factor- 6.752 www.irjmets.com

Fig 2: load-balancing model

III. Challenges Of Load Balancing

In this section, we will discuss the significant challenges cloud computing service providers face,
particularly for load balancing and other security-related aspects since cloud computing has gained the
attention of researchers due to its challenging nature. Cloud researchers work on virtual machine
migrations, fail tolerance, virtual machine security, QoS satisfaction, and other related issues[6]. Load
balancing is one of the significant challenges for cloud computing researchers. For the rest of this
section, we will provide a few challenges in load balancing.

 Geographically distributed nodes:

Data centers are primarily located in geographically isolated locations for large cloud computing
providers. These data centers act as a single server to solve user requirements. To make them act as a
single machine is a potential challenge, particularly for the environment where this distance is
considerable. There are many issues related to delay, such as communication speed, network
infrastructure and many more like this. If this separation is not at a long distance, network delay is
tolerated. Other challenges come even in geographically close data centers, such as choosing a suitable
algorithm that works well for all virtual machines.

 Single point of failure:

Currently, available load balancing algorithms mostly use a single machine to control the load
balancing of other virtual machines: Particularly, most virtual machines in the cloud environment work
in a distributed fashion. Hence there is a need for distributed load balancing algorithms so that the
failure of a single machine does not cause the failure of all computing resources.

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[3]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and
Science
Volume:04/Issue:01/ January-2022 Impact Factor- 6.752 www.irjmets.com

 Virtual machine migration:

More than one virtual machine can be configured on a single physical machine or server. These virtual
machines act as different systems. In the case underlying physical machines become overloaded, virtual
machines need to be migrated from one machine to other physical machines. Load-balancing for such
kinds of VM migrations is a challenge to be addressed.

 Heterogeneous nodes:
The early load-balancing researchers developed algorithms for homogeneous nodes. As user
requirements are complex, there comes a concept of heterogeneous nodes in a cloud environment. New
load-balancing techniques are much needed to meet the requirements of heterogeneous nodes.

 Load balancing scalability:

Scalability refers to the increasing or decreasing computational user resources according to their
demand. Scalability and on-demand availability are essential features of cloud computing. To achieve
scalability suitable load-balancing algorithm is mandatory.

 Algorithm complexity:
Load-balancing algorithms should make as efficient and straightforward as possible to overcome
overhead on cloud computing services

Fig 3: classification of load-balancing algorithm

Below we mentioned some load balancing techniques with their pros and cons.

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[4]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and
Science
Volume:04/Issue:01/ January-2022 Impact Factor- 6.752 www.irjmets.com

System
Technique Concept Pros Cons
State
Live VM
Workload
Migration Low task
balancing and
execution
Dynamic resource
Load balancing Reduced
management
VMs using end of response time, Homogeneous VM,
framework
service time reduced Independent task
processing time Actual instant
High resource processing power
Dynamic Predictive
Identifying use, low calculation is
load-balancing
overloaded and migration difficult, more
approach using
underloaded overhead, power consumption
machine
machine using reduced number Not tested on a real
learning
machine learning of migrations cloud
Dynamic
Minimum Allocates tasks
Modified active
Use of response time uniformly on single
VM load-
reservation table for tasks, load data centre
Dynamic balancing
for uniform balancing among Higher response
technique
allocation of VMs improves time and few
requests elasticity load-balancing
Novel
Improved Parameters.
load-balancing
Dynamic resource use,
Dynamic approach for
annexed method makespan, and
minimizing load
over static load QoS
on servers
balancing
Table 1: some general load-balancing techniques.

IV. Conclusion
Cloud computing is an ever-spading paradigm of the modern eras computing world. Cloud
computing researchers put their intensive care in various aspects; load-balancing is one of
them. Load-balancing refers to evenly distributing the computational load on virtual machines
of the cloud. In this research, we discussed some essential aspects of load-balancing
techniques. We focused mainly on the challenges of load-balancing techniques for future
researchers. This research can play a positive role for new researchers in cloud-computing
load-balancing to narrow their research problem.

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[5]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and
Science
Volume:04/Issue:01/ January-2022 Impact Factor- 6.752 www.irjmets.com

REFERENCES
[1] P. Kumar and R. Kumar, "Issues and challenges of load balancing techniques in cloud computing: A
survey," ACM Computing Surveys (CSUR), vol. 51, no. 6, pp. 1-35, 2019.
[2] D. A. Shafiq, N. Jhanjhi, and A. Abdullah, "Load balancing techniques in cloud computing
environment: A review," Journal of King Saud University-Computer and Information Sciences, 2021.
[3] S. Dhahbi, M. Berrima, and F. A. Al-Yarimi, "Load balancing in cloud computing using worst-fit bin-
stretching," Cluster Computing, pp. 1-15, 2021.
[4] A. Bisong and M. Rahman, "An overview of the security concerns in enterprise cloud computing,"
arXiv preprint arXiv:1101.5613, 2011.
[5] M. Armbrust et al., "A view of cloud computing," Communications of the ACM, vol. 53, no. 4, pp. 50-
58, 2010.
[6] J. S. Ward and A. Barker, "A Cloud Computing Survey: Developments and Future Trends in
Infrastructure as a Service Computing," arXiv preprint arXiv:1306.1394, 2013.

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[6]