An asynchronous writing method for restart files in the gysela code in prevision of exascale systems

Abstract : The present work deals with an optimization procedure developed in the full-f global GYrokinetic SEmi-LAgrangian code (GYSELA). Optimizing the writing of the restart files is necessary to reduce the computing impact of crashes. These files require a very large memory space, and particularly so for very large mesh sizes. The limited bandwidth of the data pipe between the computing nodes and the storage system induces a non-scalable part in the GYSELA code, which increases with the mesh size. Indeed the transfer time of RAM to data depends linearly on the files size. The necessity of non synchronized writing-in-file procedure is therefore crucial. A new GYSELA module has been developed. This asynchronous procedure allows the frequent writing of the restart files, whilst preventing a severe slowing down due to the limited writing bandwidth. This method has been improved to generate a checksum control of the restart files, and automatically rerun the code in case of a crash for any cause.
Document type :
Conference papers
Complete list of metadatas

Cited literature [8 references]  Display  Hide  Download

https://hal.inria.fr/hal-01048745
Contributor : Julien Bigot <>
Submitted on : Friday, July 25, 2014 - 12:22:09 PM
Last modification on : Thursday, February 7, 2019 - 2:48:40 PM
Long-term archiving on : Tuesday, November 25, 2014 - 6:02:19 PM

File

proc134307.pdf
Publisher files allowed on an open archive

Identifiers

Collections

Citation

Olivier Thomine, Julien Bigot, Virginie Grandgirard, Guillaume Latu, Chantal Passeron, et al.. An asynchronous writing method for restart files in the gysela code in prevision of exascale systems. CEMRACS 2012, Jul 2012, Luminy, France. ⟨10.1051/proc/201343007⟩. ⟨hal-01048745⟩

Share

Metrics

Record views

458

Files downloads

188