UPM Institutional Repository

A new method for job scheduling in a non-dedicated heterogeneous system


Shokripour, Amin and Othman, Mohamed and Ibrahim, Hamidah and Subramaniam, Shamala (2011) A new method for job scheduling in a non-dedicated heterogeneous system. In: 1st World Conference on Information Technology, WCIT-2010, 6 October 2010 - 10 October 2010, Istanbul . (271 -275).


During the last decade, the use of parallel and distributed systems has become more popular. Dividing data is one of the challenges in this type of systems. Divisible Load Theory (DLT) is one of the most popular method for scheduling data distribution in parallel or distributed systems. Many researches were carried out in this field and most of them were assumed that all processors are dedicated for parallel or distributed systems. Unfortunately, it is not always true in real systems. Although Non-dedicated processors were investigated in some studies, unreliable or non-dedicated link connections have not been addressed extensively as well. In this research, assuming that we have four static parameters for processors in a non-dedicated system: (1) probability of disconnection in network links, (2) required time for repairing disconnection, (3) probability of loading a local task, and (4) required time for doing local tasks. Then, we present a closed-form formula for scheduling tasks in this type of systems. Experiment results show that the proposed method for non-dedicated system has better performance than applying the proposed method for dedicated system. © 2010 Published by Elsevier Ltd.

Download File

Full text not available from this repository.

Additional Metadata

Item Type: Conference or Workshop Item (Paper)
Subject: Heterogeneous computing
Divisions: Faculty of Computer Science and Information Technology
Keywords: Dedicated systems, Divisible Load Theory, Heterogeneous, Scheduling, Unreliable systems
Depositing User: Mohamad Jefri Mohamed Fauzi
Date Deposited: 21 Apr 2011 09:59
Last Modified: 21 Apr 2011 10:02
URI: http://psasir.upm.edu.my/id/eprint/12010
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item