EZTrace: a generic framework for performance analysis

Abstract : Modern supercomputers with multi-core nodes enhanced by accelerators, as well as hybrid programming models, introduce more complexity in modern applications. Exploiting efficiently all the resources requires a complex analysis of the performance of applications in order to detect time-consuming or idle sections. We present eztrace, a generic trace generation framework that aims at providing a simple way to analyze applications. eztrace is based on plugins that allow it to trace different programming models such as MPI, pthread or OpenMP as well as user-defined libraries or applications. This framework uses two steps: one to collect the basic information during execution and one post-mortem analysis. This permits tracing the execution of applications with low overhead while allowing to refine the analysis after the execution of the program. We also present a simple script language for \eztrace that gives the user the opportunity to easily define the functions to instrument without modifying the source code of the application.
Type de document :
Communication dans un congrès
IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid), May 2011, Newport Beach, CA, United States. 2011
Liste complète des métadonnées


https://hal.inria.fr/inria-00587216
Contributeur : François Trahay <>
Soumis le : mercredi 20 avril 2011 - 17:00:13
Dernière modification le : jeudi 10 septembre 2015 - 01:06:44
Document(s) archivé(s) le : jeudi 21 juillet 2011 - 02:39:23

Identifiants

  • HAL Id : inria-00587216, version 1

Collections

Citation

François Trahay, François Rue, Mathieu Faverge, Yutaka Ishikawa, Raymond Namyst, et al.. EZTrace: a generic framework for performance analysis. IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid), May 2011, Newport Beach, CA, United States. 2011. <inria-00587216>

Partager

Métriques

Consultations de
la notice

434

Téléchargements du document

281