A new approach for analyzing I/O in parallel scienti

The e cient use of high performance computing is usually focused on the use of computational resources. However, scienti c appli- cations currently produce a large volume of information. Therefore, the Input/Output (I/O) subsystem also should be used e ciently. In order to do so, it is necessary to...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Méndez, Sandra, Panadero, Javier, Wong, Alvaro, Rexachs del Rosario, Dolores, Luque Fadón, Emilio
Formato: Objeto de conferencia
Lenguaje:Inglés
Publicado: 2012
Materias:
Ciencias Informáticas
new approach
Analyzing I/O
parallel scienti
Parallel
Distributed
Acceso en línea:http://sedici.unlp.edu.ar/handle/10915/23632
Aporte de:
SEDICI (UNLP) de Universidad Nacional de La Plata
id I19-R120-10915-23632
record_format dspace
institution Universidad Nacional de La Plata
institution_str I-19
repository_str R-120
collection SEDICI (UNLP)
language Inglés
topic Ciencias Informáticas
new approach
Analyzing I/O
parallel scienti
Parallel
Distributed
spellingShingle Ciencias Informáticas
new approach
Analyzing I/O
parallel scienti
Parallel
Distributed
Méndez, Sandra
Panadero, Javier
Wong, Alvaro
Rexachs del Rosario, Dolores
Luque Fadón, Emilio
A new approach for analyzing I/O in parallel scienti
topic_facet Ciencias Informáticas
new approach
Analyzing I/O
parallel scienti
Parallel
Distributed
description The e cient use of high performance computing is usually focused on the use of computational resources. However, scienti c appli- cations currently produce a large volume of information. Therefore, the Input/Output (I/O) subsystem also should be used e ciently. In order to do so, it is necessary to know the application I/O patterns and establish a relationship between these patterns and the I/O susbsystem con g- uration. To analyze the I/O behavior of applications, we propose use a library of the PAS2P (Application Signature for Performance Prediction) tool. Parallel applications typically have repetitive behavior, and the I/O patterns of parallel applications also have that behavior. We propose to identify the portions (I/O phases) where the application does I/O. From these I/O phases, we extract an application model that can be used to evaluate it in di erent I/O subsystems considering the I/O phases and compute-communication phases. In this paper, we present the concepts used in the PAS2P methodology, which have been adapted for MPI-IO applications. We have extracted the I/O model of applications. This ap- proach was used to estimate the I/O time of an application in di erent subsystems. The results show a relative error of estimation lower than 10%.
format Objeto de conferencia
Objeto de conferencia
author Méndez, Sandra
Panadero, Javier
Wong, Alvaro
Rexachs del Rosario, Dolores
Luque Fadón, Emilio
author_facet Méndez, Sandra
Panadero, Javier
Wong, Alvaro
Rexachs del Rosario, Dolores
Luque Fadón, Emilio
author_sort Méndez, Sandra
title A new approach for analyzing I/O in parallel scienti
title_short A new approach for analyzing I/O in parallel scienti
title_full A new approach for analyzing I/O in parallel scienti
title_fullStr A new approach for analyzing I/O in parallel scienti
title_full_unstemmed A new approach for analyzing I/O in parallel scienti
title_sort new approach for analyzing i/o in parallel scienti
publishDate 2012
url http://sedici.unlp.edu.ar/handle/10915/23632
work_keys_str_mv AT mendezsandra anewapproachforanalyzingioinparallelscienti
AT panaderojavier anewapproachforanalyzingioinparallelscienti
AT wongalvaro anewapproachforanalyzingioinparallelscienti
AT rexachsdelrosariodolores anewapproachforanalyzingioinparallelscienti
AT luquefadonemilio anewapproachforanalyzingioinparallelscienti
AT mendezsandra newapproachforanalyzingioinparallelscienti
AT panaderojavier newapproachforanalyzingioinparallelscienti
AT wongalvaro newapproachforanalyzingioinparallelscienti
AT rexachsdelrosariodolores newapproachforanalyzingioinparallelscienti
AT luquefadonemilio newapproachforanalyzingioinparallelscienti
bdutipo_str Repositorios
_version_ 1764820466051055620

Ejemplares similares