DocumentCode
2020927
Title
A Lightweight Specialized Meta-scheduler for Business-Oriented Applications
Author
Chevalier, Jérémie ; Mouton, Stéphane
fYear
2009
fDate
25-29 May 2009
Firstpage
620
Lastpage
625
Abstract
The process of meta-scheduling is a key feature in the architecture of typical computational grids. Many meta-schedulers have reached maturity obstacles that slow down their use. On one hand, some meta-schedulers are tied to an environment requiring users to adopt the whole infrastructure in order to use them. On the other hand, remaining meta-schedulers are difficult to deploy. Generally speaking, most of the grid brokers are designed for an all-purpose use and may not fit the needs of specific domain or business constraints. The latter situation happened in the scope of the BEinGRID (business experiments in grid) research project. This paper describes how a lightweight, configurable and easy to deploy meta-scheduler was implemented to meet those constraints and to be able to easily change business and domain specific rules while performing job scheduling and resource allocation. The resulting generic grid broker has been applied to image rendering for architects.
Keywords
business data processing; grid computing; metacomputing; rendering (computer graphics); scheduling; business experiments in grid project; business-oriented applications; computational grids; image rendering; job scheduling; lightweight specialized meta-scheduler; resource allocation; specific domain; Asia; Business; Communications technology; Computational modeling; Computer architecture; Distributed computing; Grid computing; Processor scheduling; Rendering (computer graphics); Resource management; 3D image rendering; Globus Toolkit; Meta-scheduling; application-specific criteria;
fLanguage
English
Publisher
ieee
Conference_Titel
Modelling & Simulation, 2009. AMS '09. Third Asia International Conference on
Conference_Location
Bali
Print_ISBN
978-1-4244-4154-9
Electronic_ISBN
978-0-7695-3648-4
Type
conf
DOI
10.1109/AMS.2009.64
Filename
5072058
Link To Document