TU Darmstadt / ULB / TUbiblio

Trace-based Detection of Lock Contention in MPI One-Sided Communication

Hermanns, Marc-André ; Geimer, Markus ; Mohr, Bernd ; Wolf, Felix
eds.: Niethammer, Christoph ; Gracia, José ; Hilbrich, Tobias ; Knüpfer, Andreas ; Resch, Michael ; Nagel, Wolfgang E. (2017)
Trace-based Detection of Lock Contention in MPI One-Sided Communication.
Proceedings of the 10th International Workshop on Parallel Tools for High Performance Computing. Stuttgart, Germany (04.10. - 05.10.2016)
doi: 10.1007/978-3-319-56702-0_6
Conference or Workshop Item, Bibliographie

Abstract

Performance analysis is an essential part of the development process of HPC applications. Thus, developers need adequate tools to evaluate design and implementation decisions to effectively develop efficient parallel applications. Therefore, it is crucial that tools provide an as complete support as possible for the available language and library features to ensure that design decisions are not negatively influenced by the level of available tool support. The message passing interface (MPI) supports three basic communication paradigms: point-to-point, collective, and one-sided. Each of these targets and excels at a specific application scenario. While current performance tools support the first two quite well, one-sided communication is often neglected. In our earlier work, we were able to reduce this gap by showing how wait states in MPI one-sided communication using active-target synchronization can be detected at large scale using our trace-based message replay technique. Further extending our work on the detection of progress-related wait states in ARMCI, this paper presents an improved infrastructure that is capable of not only detecting progress-related wait states, but also wait states due to lock contention in MPI passive-target synchronization. We present an event-based definition of lock contention, the trace-based algorithm to detect it, as well as initial results with a micro-benchmark and an application kernel scaling up to 65,536 processes.

Item Type: Conference or Workshop Item
Erschienen: 2017
Editors: Niethammer, Christoph ; Gracia, José ; Hilbrich, Tobias ; Knüpfer, Andreas ; Resch, Michael ; Nagel, Wolfgang E.
Creators: Hermanns, Marc-André ; Geimer, Markus ; Mohr, Bernd ; Wolf, Felix
Type of entry: Bibliographie
Title: Trace-based Detection of Lock Contention in MPI One-Sided Communication
Language: English
Date: 9 May 2017
Publisher: Springer
Book Title: Tools for High Performance Computing 2016
Event Title: Proceedings of the 10th International Workshop on Parallel Tools for High Performance Computing
Event Location: Stuttgart, Germany
Event Dates: 04.10. - 05.10.2016
Edition: 1. Auflage
DOI: 10.1007/978-3-319-56702-0_6
Abstract:

Performance analysis is an essential part of the development process of HPC applications. Thus, developers need adequate tools to evaluate design and implementation decisions to effectively develop efficient parallel applications. Therefore, it is crucial that tools provide an as complete support as possible for the available language and library features to ensure that design decisions are not negatively influenced by the level of available tool support. The message passing interface (MPI) supports three basic communication paradigms: point-to-point, collective, and one-sided. Each of these targets and excels at a specific application scenario. While current performance tools support the first two quite well, one-sided communication is often neglected. In our earlier work, we were able to reduce this gap by showing how wait states in MPI one-sided communication using active-target synchronization can be detected at large scale using our trace-based message replay technique. Further extending our work on the detection of progress-related wait states in ARMCI, this paper presents an improved infrastructure that is capable of not only detecting progress-related wait states, but also wait states due to lock contention in MPI passive-target synchronization. We present an event-based definition of lock contention, the trace-based algorithm to detect it, as well as initial results with a micro-benchmark and an application kernel scaling up to 65,536 processes.

Divisions: 20 Department of Computer Science
20 Department of Computer Science > Parallel Programming
Date Deposited: 20 Apr 2018 12:22
Last Modified: 25 Jun 2024 06:14
PPN: 519356454
Export:
Suche nach Titel in: TUfind oder in Google
Send an inquiry Send an inquiry

Options (only for editors)
Show editorial Details Show editorial Details