World Applied Sciences Journal 29 (Data Mining and Soft Computing Techniques): 159-165, 2014
ISSN 1818-4952
© IDOSI Publications, 2014
DOI: 10.5829/idosi.wasj.2014.29.dmsct.30
Corresponding Author: S. RamKumar, CSE, School of Computing, SASTRA University, India.
159
Towards Efficient Load Balancing and Green it Mechanisms in Cloud Environment
S. RamKumar, V. Vaithiyanathan and M. Lavanya
1 2 2
CSE, School of Computing, SASTRA University, India
1
School of Computing, SASTRA University, India
2
Abstract: Cloud Environment allows users to access on-demand IT resources over the cyberspace based on
pay per use pattern. Optimizing the resource utilizationand reducing the amount of active server machine usage
in clouds become a challenging task. The major intend of this work is to builda model which supports balancing
the workload in cloud datacenters and minimizing the amount of active servers to support Green-IT model.Based
on the conventional Earliest Deadline First(EDF) algorithm, this work presents an Adaptive Earliest Deadline
First (AEDF) algorithm for load balancing problem in order to achieve efficient consumption of cloud resources
and also to support Green Computing technology. Comparing with traditional load balancing algorithm,the
proposed method achieves high throughput, minimized cost and effective scheduling of tasks. Our proposed
model supports both Load balancing and Green-IT technology.
Key words: Cloud Environment Green Computing Load balancing Resourceutilization
INTRODUCTION software based virtualization techniques. Hybrid
Cloud Computing [1, 2] is the model for delivering based virtualization machinery and traditional Para-
IT resources, in which the resources are retrieved from the virtualization techniques. Para-virtualization is the
cyberspace using pay per use pattern. It allows the cloud virtualization technique used in xen server and it serves
consumer to extent the infrastructure dynamically in the interfaces to the virtual machine. Para-virtualization
within rapid period. In nature, the cloud contains uses the kernel for guest os is exactly loaded and
virtualization of resources (i.e. software and hardware) compiled before installing in to the virtual machine. This
these resources are maintained and managed in cloud hybrid virtualization technique can decrease the
environment itself. Cloud environment provides several complications and overhead in the conventional virtual
characteristics: machines.
On-demand utilization. Green IT it can maximize the utilization of cloud resources
Network of Shared Resources. in energy efficient manner. The Green IT also reduces the
Elastic Provisioning. energy expenditure. This proposed work intends to attain
Fine – grained metering. the both Load balancing technique and Green IT
Virtualization in cloud computing is the constructing
virtual system instead of real system, it may be an Green IT: Maximize the consumption of cloud
computational resources, hardware platforms, software virtual resources, also it decreases the cost of cloud
platforms, storage machine depends upon the user IT services and minimize the amount of active server
requirements. Now a day’s several investigations are machine.
focused on constructing efficient virtualization Load Balancing: Allocate the workload
technologies for IT purposes (i.e. Green computing and across several computers, servers and other
cloud computing). Hardware based virtualization resources to attain efficient resource
machinery introduced for reducing the constraints of the consumption.
virtualization [3] is the combination of both hardware
This work presents Green Computing technology or
technology: