Advanced search
Start date
Betweenand
Related content

Task scheduling with data locality in grids

Grant number: 14/02104-0
Support type:Scholarships in Brazil - Master
Effective date (Start): July 01, 2014
Effective date (End): February 28, 2015
Field of knowledge:Physical Sciences and Mathematics - Computer Science
Cooperation agreement: Coordination of Improvement of Higher Education Personnel (CAPES)
Principal Investigator:Eduardo Candido Xavier
Grantee:Marcelo Galvão Póvoa
Home Institution: Instituto de Computação (IC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil

Abstract

This study presents problems of task scheduling and data replication in distributed computing systems known as Data Grids. Both problems have already been extensively studied independently, but we focus on an integrated analysis which tries to optimize a single objective function. Significant theoretical results are not yet known for this approach, which is being subject of more recent research. We are interested in analyzing the problem in two approaches: theoretical (evaluate its hardness, compare variants and propose an approximation algorithm) and experimental (design and evaluation of efficient heuristics). Our results so far include a concise integer linear programming model and a constant factor approximation algorithm for a restricted problem formulation. The ultimate goal of this work is to present a robust algorithm with similar, if not better, performance compared to the results currently known. (AU)

Scientific publications
(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)
POVOA, MARCELO G.; XAVIER, EDUARDO C. Approximation algorithms and heuristics for task scheduling in data-intensive distributed systems. International Transactions in Operational Research, v. 25, n. 5, p. 1417-1441, SEP 2018. Web of Science Citations: 0.

Please report errors in scientific publications list by writing to: cdi@fapesp.br.