Distributed wait state tracking for runtime MPI deadlock detection
From International Center for Computational Logic
Distributed wait state tracking for runtime MPI deadlock detection
Tobias HilbrichTobias Hilbrich, Bronis R. de SupinskiBronis R. de Supinski, Wolfgang E. NagelWolfgang E. Nagel, Joachim ProtzeJoachim Protze, Christel BaierChristel Baier, Matthias S. MüllerMatthias S. Müller
Tobias Hilbrich, Bronis R. de Supinski, Wolfgang E. Nagel, Joachim Protze, Christel Baier, Matthias S. Müller
Distributed wait state tracking for runtime MPI deadlock detection
Proc. of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC), 16:1--12, 2013. ACM
Distributed wait state tracking for runtime MPI deadlock detection
Proc. of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC), 16:1--12, 2013. ACM
- KurzfassungAbstract
The widely used Message Passing Interface (MPI) with its multitude of communication functions is prone to usage errors. Runtime error detection tools aid in the removal of these errors. We develop MUST as one such tool that provides a wide variety of automatic correctness checks. Its correctness checks can be run in a distributed mode, except for its deadlock detection. This limitation applies to a wide range of tools that either use centralized detection algorithms or a timeout approach. In order to provide scalable and distributed deadlock detection with detailed insight into deadlock situations, we propose a model for MPI blocking conditions that we use to formulate a distributed algorithm. This algorithm implements scalable MPI deadlock detection in MUST. Stress tests at up to 4,096 processes demonstrate the scalability of our approach. Finally, overhead results for a complex benchmark suite demonstrate an average runtime increase of 34% at 2,048 processes. - Forschungsgruppe:Research Group: Algebraische und logische Grundlagen der InformatikAlgebraic and Logical Foundations of Computer Science
@inproceedings{HSNPBM2013,
author = {Tobias Hilbrich and Bronis R. de Supinski and Wolfgang E. Nagel
and Joachim Protze and Christel Baier and Matthias S.
M{\"{u}}ller},
title = {Distributed wait state tracking for runtime {MPI} deadlock
detection},
booktitle = {Proc. of the International Conference for High Performance
Computing, Networking, Storage and Analysis (SC)},
publisher = {ACM},
year = {2013},
pages = {16:1--12},
doi = {10.1145/2503210.2503237}
}