Parallel File I/O with MPI-2

14.01.02


Start Lecture by clicking here


Content

Author:Rolf Rabenseifner

  1. Parallel File I/O with MPI-2
  2.     Acknowledgements
  3.     Motivation, I.
  4.     Motivation, II.
  5.     MPI-I/O Features
  6.     MPI-I/O, Principles
  7.     Logical view / Physical view
  8.     Definitions
  9.     Comments on Definitions
  10.     Opening an MPI File
  11.     Default View
  12.     Access Modes
  13.     File Info: Reserved Hints
  14.     Closing and Deleting a File
  15.     Writing with Explicit Offsets
  16.     MPI–IO Exercise 1: Four processes write a file in parallel
  17. File Views
  18.     Set/Get File View
  19.     Data Representation, I.
  20.     Data Representation, II.
  21.     Fileview examples
  22.     Example with Subarray, I.
  23.     Example with Subarray, II. — Distribution
  24.     Example with Subarray, III. — Reading the file
  25.     Example with Subarray, IV.
  26.     Example with Darray, I.
  27.     Example with Darray, II.
  28.     Example with Darray, III.
  29.     5 Aspects of Data Access
  30.     All Data Access Routines
  31.     Explicit Offsets
  32.     Individual File Pointer, I.
  33.     Individual File Pointer, II.
  34.     MPI–IO Exercise 2: Using fileviews and individual filepointers
  35.     MPI–IO Exercise 2: Using fileviews and individual filepointers, continued
  36. Shared File Pointer, I.
  37.     Shared File Pointer, II.
  38.     Collective Data Access
  39.     Application Scenery, I.
  40.     Application Scenery, II.
  41.     Application Scenery, III.
  42.     Non-blocking Data Access
  43.     Split Collective Data Access, I.
  44.     Split Collective Data Access, II.
  45.     Scenery – Split Collective
  46.     Other File Manipulation Routines
  47.     MPI I/O Error Handling
  48.     Implementation-Restrictions, I.
  49.     Implementation-Restrictions, II.
  50.     Implementation-Restrictions, III.
  51.     MPI-I/O: Summary
  52.     MPI–IO Exercise 3: Collective ordered I/O
  53.     MPI–IO Exercise 4: I/O Benchmark

Back to the Parallel Programming Workshop Overview