Performance analysis of MPI-I/O primitives on a PC cluster

Authors:
Juan Touriño;Jacobo Barro;Ramón Doallo;Víctor M. Gulías
Affiliations:
University of A Coruña, 15071 A Coruña, Spain;University of A Coruña, 15071 A Coruña, Spain;University of A Coruña, 15071 A Coruña, Spain;University of A Coruña, 15071 A Coruña, Spain
Venue:
Proceedings of the 2002 ACM symposium on Applied computing
Year:
2002

Citing 5
Cited 0

Characterization of Message-Passing Overhead on the AP3000 Multicomputer

ICPP '02 Proceedings of the 2001 International Conference on Parallel Processing
A Parallel I/O Test Suite

Proceedings of the 5th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Data Sieving and Collective I/O in ROMIO

FRONTIERS '99 Proceedings of the The 7th Symposium on the Frontiers of Massively Parallel Computation
MPI-IO on a Parallel File System for Cluster of Workstations

IWCC '99 Proceedings of the 1st IEEE Computer Society International Workshop on Cluster Computing
PVFS: a parallel file system for linux clusters

ALS'00 Proceedings of the 4th annual Linux Showcase & Conference - Volume 4

Quantified Score

Hi-index	0.00

Visualization

Abstract

Cluster computing is an area of growing interest to support parallel and distributed applications. Many of these applications are I/O intensive and the limited bandwidth of the I/O subsystem of the cluster is an important bottleneck usually ignored. Thus, the performance of parallel I/O primitives is critical for the overall cluster performance. In this work, we try to characterize I/O routines supported by the standard MPI-2 on a PC cluster using the NFS and PVFS file systems in order to detect weak spots in the use of these routines and to predict their impact on the application's performance.