Parallel filesystems and parallel IO libraries @MdS

Apr 21, 2016, 9:00 AM → Apr 22, 2016, 5:30 PM CET

Maison de la simulation

The increase in computational power goes hand in hand with an increase in the size of the data to be managed, both on the input and on the output sides. IO can easily become a bottleneck for large scale architectures.

The understanding of parallel file system mechanisms and MPI-IO software layer enables users to efficiently use existing high level libraires like Netcdf or HDF5.In addition this course will provide an introduction to one of the emerging IO technology developed at CEA (France) and widely used il the climate community: XIOS.

Topics:

•     MPI-IO
•     Best practices for parallel IO, MPI-IO hints
•     Parallel file systems: Lustre, 
•     High level IO libraries (HDF5)
•     IO server concept (XIOS)

Instructors: M. Haefele (Maison de la Simulation, CNRS), Ph. Wautelet (Laboratoire d'Aerologie, CNRS), Ph. Deniel (TGCC, CEA), Manh-Ha Nguyen (LSCE, CEA)

Learning outcomes: After this course, participants should understand the tradeoffs implied by using a parallel file-system, and know how to efficiently use parallel IO libraries.

Prerequisites: Knowledge of C or Fortran programming language, parallel programing with MPI

Thu, April 21

9:00 AM → 10:30 AM MPI-IO

10:30 AM → 11:00 AM Coffee break

11:00 AM → 12:30 PM MPI-IO

12:30 PM → 2:00 PM Lunch

2:00 PM → 3:30 PM MPI-IO hands on

3:30 PM → 4:00 PM Coffee break

4:00 PM → 5:30 PM IO best practice & MPI-IO hints

Fri, April 22

9:00 AM → 10:30 AM HDF5 + Hands on

10:30 AM → 11:00 AM Coffee break

11:00 AM → 12:30 PM HDF5 + Hands on

12:30 PM → 2:00 PM Lunch

2:00 PM → 3:00 PM Lustre file system

3:00 PM → 3:30 PM Coffee break

3:30 PM → 5:30 PM XIOS