Data sieving and collective I/O in ROMIO

Rajeev Thakur, William Gropp, Ewing Lusk

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The I/O access patterns of parallel programs often consist of accesses to a large number of small, noncontiguous pieces of data. If an application's I/O needs are met by making many small, distinct I/O requests, however, the I/O performance degrades drastically. To avoid this problem, MPI-IO allows users to access a noncontiguous data set with a single I/O function call. This feature provides MPI-IO implementations an opportunity to optimize data access. We describe how our MPI-IO implementation, ROMIO, delivers high performance in the presence of noncontiguous requests. We explain in detail the two key optimizations ROMIO performs: data sieving for noncontiguous requests from one process and collective I/O for noncontiguous requests from multiple processes. We describe how one can implement these optimizations portably on multiple machines and file systems, control their memory requirements, and also achieve high performance. We demonstrate the performance and portability with performance results for three applications-an astrophysics-application template (DIST3D) the NAS BTIO benchmark, and an unstructured code (UNSTRUC)-on five different parallel machines: HP Exemplar IBM SP, Intel Paragon, NEC SX-4, and SGI Origin2000.

Original languageEnglish (US)
Title of host publicationProceedings - Frontiers 1999, 7th Symposium on the Frontiers of Massively Parallel Computation
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages182-189
Number of pages8
ISBN (Electronic)0769500870, 9780769500874
DOIs
StatePublished - 1999
Event7th Symposium on the Frontiers of Massively Parallel Computation, Frontiers 1999 - Annapolis, United States
Duration: Feb 21 1999Feb 25 1999

Publication series

NameProceedings - Frontiers 1999, 7th Symposium on the Frontiers of Massively Parallel Computation

Other

Other7th Symposium on the Frontiers of Massively Parallel Computation, Frontiers 1999
CountryUnited States
CityAnnapolis
Period2/21/992/25/99

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Hardware and Architecture
  • Modeling and Simulation

Fingerprint Dive into the research topics of 'Data sieving and collective I/O in ROMIO'. Together they form a unique fingerprint.

Cite this