Scalable Event Tracing for High End Systems
We have developed a new performance measurement technique that is a hybrid between profiling and tracing, and a new, scalable presentation of event-based performance data that will facilitate quick understanding of the behavior of the parallel program. Our technique reduces the amount of data measured and stored by capitalizing on the repeated behavior of programs, and on the similarity of the behavior and performance of parallel processes in an application run. We retain enough information to recreate a complete (although approximate) trace of the parallel run. In addition, because the data has already been classified into behavior categories, we can present reduced, partially analyzed performance data to the user, highlighting the performance behaviors that comprised most of the execution time.
Acknowledgments
- This work was sponsored in part by a PSU Faculty Development grant.
Publications
-
Kathryn Mohror and Karen L. Karavanic, "Trace Profiling: Scalable Event Tracing on High-End Parallel Systems," Parallel Computing, Volume 38, Issues 4-5, April-May 2012, Pages 194-225, Elsevier.
-
Kathryn M. Mohror, "Scalable Event Tracing on High-End Parallel Systems," Ph.D. Dissertation, Portland State University Department of Computer Science, 2009.
-
Kathryn Mohror and Karen L. Karavanic, "Evaluating Similarity-Based Trace Reduction Techniques for Scalable Performance Analysis," SC2009, Portland, OR, November 2009.
-
Kathryn Mohror, Karen L. Karavanic and Allan Snavely, "Scalable Event Trace Visualization," Workshop on Productivity and Performance (PROPER 2009) at EuroPar 2009, Delft, The Netherlands, August 2009.
- Kathryn Mohror and Karen L. Karavanic, "Towards Scalable Event Tracing for High End Systems," High Performance Computation Conference (HPCC) 2007 Houston, TX, USA, September 2007.
- Kathryn Mohror and Karen L. Karavanic, "Scalable Event-based Performance Measurement in High-End Environments", Research Poster, SIGMETRICS Student Workshop, SIGMETRICS'07, San Diego, CA, June 13, 2007.
- Kathryn Mohror and Karen L. Karavanic, "A Study of Tracing Overhead on a High-Performance Linux Cluster," Research Poster, PPoPP'07, March 15, 2007.
- Kathryn Mohror and Karen L. Karavanic, "An Investigation of Tracing Overheads on High End Systems," PSU CS Department Technical Report TR-06-06, December, 2006.