A Predictive Approach for Selecting Suitable Computing Nodes in Grid Environment by Using Data Mining Technique - Advances in Computational Science and Engineering - page 190

A Predictive Approach for Selecting Suitable Computing
Nodes in Grid Environment by Using Data Mining
Technique
Asgarali Bouyer and Seyed Mojtaba Mousavi
Islamic Azad University- Miyandoab Branch,
Miyandoab Iran
{basgarali2,Mmseyed2}@Siswa.utm.my
Abstract. Grid Computing is a new technology for developing resources sharing
such as computational resources, data and so on between many organizations in
different geographical locations. Therefore resource management for finding, se-
lecting, and scheduling is more important in the grid. This paper has concen-
trated on nodes selection to assign proper jobs. There are many methods to do
this work but each of them has weakness. In this paper, we propose a new ap-
proach for fair Grid Resource Broker (GRB) to select optimal nodes, and then
we compare it to other existing methods. Our proposed approach applies Fuzzy
decision tree as evolution method for finding suitable resources in existing pool
of resources in the grid. Moreover, if our approach is not encountered with the
lack of resources, it will acts in best form and assigns to each task enough com-
putational power for it to finish within its deadline. The role of this resource bro-
ker service is executing of Fuzzy Decision Tree algorithm to find the best nodes
according to the requirements and conditions of the job on the Grid. The results
of experiments show a powerful effect in improving resource finding cycle.
Keywords: Grid Resource broker, Resource Selection, Data mining, Fuzzy
Decision Tree.
1 Introduction
Grid is a decentralized heterogeneous system that made up virtual organizations (VOs).
Each VO is composed of several different nodes. Each node can be server computers,
desktop PCs, clusters, and other kinds of hardware, which are sharing some resources
with other nodes. A main goal of grid computing is enabling applications to identify
resources dynamically to create distributed computing environments [1].
In order to perform job scheduling and resource management at Grid level, usually
it is used a Resource Broker or a meta-scheduler. A resource broker is fundamental in
any large-scale Grid environment. The task of a Grid resource broker and scheduler is
to dynamically identify and characterize the available resources, and to select and
allocate the most proper resources for a given job. In a broker-based management
system, brokers are responsible for finding and selecting best nods, ensuring the
trustworthiness of the service provider. Resource selection is an important issue in a