This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by Faerman, M. Articles by Casanova, H. Search for Related Content Social Bookmarking                         What's this?

Resource Allocation Strategies for Guided Parameter Space Searches

Adam Birnbaum

Francine Berman

SAN DIEGO SUPERCOMPUTER CENTER

Henri Casanova

DEPT. OF COMPUTER SCIENCE AND ENGINEERING, UNIVERSITY OF CALIFORNIA, USA

Computational Grids lend themselves well to parameter sweep applications, in which independent tasks calculate results for points in a parameter space. However, it is possible for a parameter space to become so large as to pose prohibitive system requirements. In these cases, user-guided searchespromise to reduce overall computation time. In this paper, we address an interesting challenge posed by these user-directed searches: how should computing resources be allocated to application tasks as the overall computation is being guided by the user? We present a model for user-guided searches, and then propose a number of resource allocation strategies and evaluate them in simulation. We find that prioritizing the assignments of tasks to computing resources throughout the search can lead to substantial performance improvements.

Key Words: resource allocation • scheduling • priorities • parallel computing • time sharing • parametric studies • guided search • steering • progressive refinement

International Journal of High Performance Computing Applications, Vol. 17, No. 4, 383-402 (2003)
DOI: 10.1177/10943420030174004


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?