SaGM09iske

G.P. Santos, L. Garmendia, R.S. Montero, V. Lopez. Fuzzy System to Evaluate Resources Reliability in a Grid Environment. In 4th International ISKE Conference on Intelligent Systems and Knowledge Engineering, Pages 357-362, 2009.

Abstract

Job scheduling in Computational Grids is a complex task because of the lack of control over the resources. Under these conditions, it is impossible to know exactly which host will perform better for a given job. On the other hand, fuzzy systems have demonstrated in recent years to behave well in this kind of situations, with imprecision, incomplete information or a lack of an exact model of the optimization space. In this paper we present a fuzzy system, built on top of the GridWay meta-scheduler, to evaluate the reliability of each candidate resource. This information is used by GridWay to choose the resource where the job will be dispatched

Keyword

[ Grid ]

Contact

Ruben S. Montero

BibTex Reference

@InProceedings{SaGM09iske,
   Author = {Santos, G.P. and Garmendia, L. and Montero, R.S. and Lopez, V.},
   Title = {Fuzzy System to Evaluate Resources Reliability in a Grid Environment},
   BookTitle = {4th International ISKE Conference on Intelligent Systems and Knowledge Engineering},
   Pages = {357--362},
   Publisher = {World Scientific},
   Year = {2009}
}