Damaris: Addressing Performance Variability in Data Management for Post-Petascale Simulations

Abstract : With exascale computing on the horizon, reducing performance variability in data management tasks (stor- age, visualization, analysis, etc.) is becoming a key challenge in sustaining high performance. This variabil- ity significantly impacts the overall application performance at scale and its predictability over time. In this paper, we present Damaris, a system that leverages dedicated cores in multicore nodes to offload data management tasks, including I/O, data compression, scheduling of data movements, in situ analysis and visualization. We evaluate Damaris with the CM1 atmospheric simulation and the Nek5000 computa- tional fluid dynamic simulation on four platforms, including NICS’s Kraken and NCSA’s Blue Waters. Our results show in particular that (1) Damaris fully hides the I/O variability as well as all I/O-related costs, which makes simulation performance predictable; (2) it increases the sustained write throughput by a factor of up to 15 compared with standard I/O approaches; (3) it allows almost perfect scalability of the simulation up to over 9,000 cores, as opposed to state-of-the-art approaches that fail to scale; (4) it enables a seamless connection to the VisIt visualization software to perform in situ analysis and visualization in a way that does not impact the performance of the simulation, nor its variability. In addition, we further extended our implementation of Damaris to also support the use of dedicated nodes and conducted a thorough comparison of the two approaches –dedicated cores and dedicated nodes– for I/O tasks with the aforementioned applications.
Type de document :
Article dans une revue
ACM Transactions on Parallel Computing, Association for Computing Machinery, 2016
Liste complète des métadonnées

Littérature citée [64 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01353890
Contributeur : Shadi Ibrahim <>
Soumis le : mercredi 31 août 2016 - 21:48:41
Dernière modification le : mercredi 16 mai 2018 - 11:23:28
Document(s) archivé(s) le : jeudi 1 décembre 2016 - 21:17:03

Fichier

Damaris - ACM Transactions on ...
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01353890, version 1

Citation

Matthieu Dorier, Gabriel Antoniu, Franck Cappello, Marc Snir, Robert Sisneros, et al.. Damaris: Addressing Performance Variability in Data Management for Post-Petascale Simulations. ACM Transactions on Parallel Computing, Association for Computing Machinery, 2016. 〈hal-01353890〉

Partager

Métriques

Consultations de la notice

957

Téléchargements de fichiers

157