A Self-Organizing Neural Network for Job Scheduling in Distributed Systems

The aim of this work is to describe a possible approach for the optimization of the job scheduling in large distributed systems, based on a self-organizing Neural Network. This dynamic scheduling system should be seen as adaptive middle ayer software, aware of current available resources and making...

Descripción completa

Detalles Bibliográficos
Autores principales: Newman, Harvey B, Legrand, Iosif
Lenguaje:eng
Publicado: 2001
Materias:
Detectors and Experimental Techniques
Acceso en línea:http://cds.cern.ch/record/687316
_version_ 1780901732314775552
author Newman, Harvey B
Legrand, Iosif
author_facet Newman, Harvey B
Legrand, Iosif
author_sort Newman, Harvey B
collection CERN
description The aim of this work is to describe a possible approach for the optimization of the job scheduling in large distributed systems, based on a self-organizing Neural Network. This dynamic scheduling system should be seen as adaptive middle ayer software, aware of current available resources and making the scheduling decisions using the past experience. It aims to optimize job specific parameters as well as the resource utilization. The scheduling system is able to dynamically learn and cluster information in a large dimensional parameter space and at the same time to explore new regions in the parameters space. This self-organizing scheduling system may offer a possible solution to provide an effective use of resources for the off-line data processing jobs for future HEP experiments.
id cern-687316
institution Organización Europea para la Investigación Nuclear
language eng
publishDate 2001
record_format invenio
spelling cern-6873162019-09-30T06:29:59Zhttp://cds.cern.ch/record/687316engNewman, Harvey BLegrand, IosifA Self-Organizing Neural Network for Job Scheduling in Distributed SystemsDetectors and Experimental TechniquesThe aim of this work is to describe a possible approach for the optimization of the job scheduling in large distributed systems, based on a self-organizing Neural Network. This dynamic scheduling system should be seen as adaptive middle ayer software, aware of current available resources and making the scheduling decisions using the past experience. It aims to optimize job specific parameters as well as the resource utilization. The scheduling system is able to dynamically learn and cluster information in a large dimensional parameter space and at the same time to explore new regions in the parameters space. This self-organizing scheduling system may offer a possible solution to provide an effective use of resources for the off-line data processing jobs for future HEP experiments.CMS-NOTE-2001-009oai:cds.cern.ch:6873162001-02-14
spellingShingle Detectors and Experimental Techniques
Newman, Harvey B
Legrand, Iosif
A Self-Organizing Neural Network for Job Scheduling in Distributed Systems
title A Self-Organizing Neural Network for Job Scheduling in Distributed Systems
title_full A Self-Organizing Neural Network for Job Scheduling in Distributed Systems
title_fullStr A Self-Organizing Neural Network for Job Scheduling in Distributed Systems
title_full_unstemmed A Self-Organizing Neural Network for Job Scheduling in Distributed Systems
title_short A Self-Organizing Neural Network for Job Scheduling in Distributed Systems
title_sort self-organizing neural network for job scheduling in distributed systems
topic Detectors and Experimental Techniques
url http://cds.cern.ch/record/687316
work_keys_str_mv AT newmanharveyb aselforganizingneuralnetworkforjobschedulingindistributedsystems
AT legrandiosif aselforganizingneuralnetworkforjobschedulingindistributedsystems
AT newmanharveyb selforganizingneuralnetworkforjobschedulingindistributedsystems
AT legrandiosif selforganizingneuralnetworkforjobschedulingindistributedsystems

Ejemplares similares