Scalable implementations of MPI atomicity for concurrent overlapping I/O

W. K. Liao, Alok Choudhary, K. Coloma, G. K. Thiruvathukal, L. Ward, E. Russell, N. Pundit

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

5 Scopus citations

Abstract

For concurrent I/O operations, atomicity defines the results in the overlapping file regions simultaneously read/written by requesting processes. Atomicity has been well studied at the file system level, such as POSIX standard. We investigate the problems arising from the implementation of MPI atomicity for concurrent overlapping write access and provide two programming solutions. Since the MPI definition of atomicity differs from the POSIX one, an implementation that simply relies on the POSIX file systems does not guarantee correct MPI semantics. To have a correct implementation of atomic I/O in MPI, we examine the efficiency of three approaches: I) file locking, 2) graph-coloring, and 3) process-rank ordering. Performance complexity for these methods are analyzed and their experimental results are presented for file systems including NFS, SGI's XFS, and IBM's GPFS.

Original languageEnglish (US)
Title of host publicationProceedings - 2003 International Conference on Parallel Processing, ICPP 2003
EditorsP. Sadayappan, Chu-Sing Yang
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages239-246
Number of pages8
ISBN (Electronic)0769520170
DOIs
StatePublished - Jan 1 2003
Event2003 International Conference on Parallel Processing, ICPP 2003 - Kaohsiung, Taiwan, Province of China
Duration: Oct 6 2003Oct 9 2003

Publication series

NameProceedings of the International Conference on Parallel Processing
Volume2003-January
ISSN (Print)0190-3918

Other

Other2003 International Conference on Parallel Processing, ICPP 2003
CountryTaiwan, Province of China
CityKaohsiung
Period10/6/0310/9/03

Keywords

  • Application software
  • Bandwidth
  • Concurrent computing
  • File systems
  • Hardware
  • Laboratories
  • Large-scale systems
  • Multidimensional systems
  • Performance analysis
  • Programming

ASJC Scopus subject areas

  • Software
  • Mathematics(all)
  • Hardware and Architecture

Fingerprint Dive into the research topics of 'Scalable implementations of MPI atomicity for concurrent overlapping I/O'. Together they form a unique fingerprint.

Cite this