Skip to Main content Skip to Navigation

Damaris: Leveraging Multicore Parallelism to Mask I/O Jitter

Matthieu Dorier 1 Gabriel Antoniu 2 Franck Cappello 3, 4, 5 Marc Snir 6 Leigh Orf 7
2 KerData - Scalable Storage for Clouds and Beyond
Inria Rennes – Bretagne Atlantique , IRISA-D1 - SYSTÈMES LARGE ÉCHELLE
3 GRAND-LARGE - Global parallel and distributed computing
CNRS - Centre National de la Recherche Scientifique : UMR8623, Inria Saclay - Ile de France, UP11 - Université Paris-Sud - Paris 11, LIFL - Laboratoire d'Informatique Fondamentale de Lille, LRI - Laboratoire de Recherche en Informatique
Abstract : With exascale computing on the horizon, the performance variability of I/O systems presents a key challenge in sustaining high performance. In many HPC applications, I/O is performed concurrently by all processes; this produces I/O bursts, which causes resource contention and substantial variability of I/O performance, significantly impacting the overall application performance. In this paper, we utilize the IOR benchmark to explore the influence of user-configurable parameters on I/O variability. We then propose a new approach, called Damaris, leveraging dedicated I/O cores on each multicore SMP node to efficiently perform asynchronous data processing and I/O. We evaluate our approach on two different platforms with the CM1 atmospheric model, one of the BlueWaters HPC applications. By gathering data into large files while avoiding synchronization between cores, our solution increases the I/O throughput by a factor of 6, hides all I/O-related costs, and enables a 600\% compression ratio without any additional overhead.
Complete list of metadatas

https://hal.inria.fr/inria-00614597
Contributor : Matthieu Dorier <>
Submitted on : Friday, August 12, 2011 - 3:43:58 PM
Last modification on : Monday, December 9, 2019 - 5:24:06 PM
Document(s) archivé(s) le : Sunday, November 13, 2011 - 2:22:09 AM

Files

RR-7706.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00614597, version 1

Collections

Citation

Matthieu Dorier, Gabriel Antoniu, Franck Cappello, Marc Snir, Leigh Orf. Damaris: Leveraging Multicore Parallelism to Mask I/O Jitter. [Research Report] RR-7706, 2011. ⟨inria-00614597v1⟩

Share

Metrics

Record views

81

Files downloads

99