Please use this identifier to cite or link to this item:
Title: Fault Tolerant Heterogeneous Limited Duplication Scheduling algorithm for Decentralized Grid
Authors: Agarwal, Neha
Chauhan, Piyush
Keywords: Grid computing
Dependent task scheduling
Task duplication strategy
Decentralized fault tolerance
Issue Date: 2013
Publisher: Jaypee University of Information Technology, Solan, H.P.
Abstract: Fault tolerance is one of the most desirable property in decentralized grid computing systems, where computational resources are geographically distributed. These resources collaborate in order to execute workflow applications as fast as possible. In workflow applications, tasks are dependent on each other, so it becomes extremely vital that scheduling techniques should also have some decentralized fault tolerant mechanism. In this paper, we have proposed a decentralized fault tolerant mechanism which utilize the checkpoint concept; for Heterogeneous Limited Duplication (HLD) algorithm. HLD is based on task duplication scheduling in heterogeneous environment. There are two fold benefits firstly; if node failure occurs then rest of grid nodes sustain the execution of application. Secondly, less makespan of application is obtained using checkpoint concept. Therefore, application scheduled over decentralized grid systems (which are known for their unreliable behavior) will yield results fast utilizing algorithm proposed in this paper.
Appears in Collections:Journal Articles

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.