Abstract
Profiling-based performance visualization and analysis of program execution is widely used for tuning and improving the performance of parallel applications. There are several profiler-based tools for effective application performance analysis and visualization. However, a majority of these tools are not equally effective for performance tuning of applications consisting of 100’s to 10,000’s of tasks or applications generating several gigabytes to terabytes of trace information. This paper identifies architectural and usability limitations applicable to majority of existing performance analysis tools and proposes an alternative design to improve analysis of large amounts of trace-data. The new design addresses architectural as well as user-interface issues for terascale systems by providing scalable, flexible and automated mechanisms to analyze trace data. Using the proposed design, the authors have implemented an MPI application performance analysis tool, SeeWithin/Pro, as a proof-of-concept that the design can support flexible query-based analysis mechanisms to reveal complex performance statistics
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Miller, B.P., Callaghan, M.D., Cargille, J.M., Hollingsworth, J.K., Irvin, R.B., Karavanic, K.L., Kunchithapadam, K., Newhall, T.: The Paradyn Parallel Performance Measurement Tools. IEEE Computer 28(11) (November 1995)
Shirley, B., Breshears, C.P.: Usability Study of Portable Parallel Performance Tools, http://www.hpcmo.hpc.mil/Htdocs/UGC/UGC98/papers/2c
Yan, C.: Performance Tuning with AIMS—An Automated Instrumentation and Monitoring System for Multicomputers. In: Proceedings of the 27th Hawaii international Conference on Systems Sciences, ACM, New York (1994)
Java 2 Platform, Enterprise Edition. J2EE, http://java.sun.com/j2ee/
Herrarte, V., Lusk, E.: Studying parallel program behavior with Upshot. Argonne National Laboratory - Technical Report ANL-91/15 (1991)
DeRose, L., Zhang, Y., Reed, D.: SvPablo: A multi-language performance analysis system. In: Proceedings of 10th International Conference on Computer Performance Evaluation (September 1998)
Nagel, W.E., Arnold, A., Weber, M., Hoppe, H.-C., Solchenbach, K.: VAMPIR: Visualization and Analysis of MPI Resources. Supercomputer 63 12(1), 69–80 (1996)
OpenMP, http://www.openmp.org
Parallel Virtual Machine (PVM), http://www.epm.ornl.gov/pvm
van Maurice, R., Tourancheau, B., Vigouroux, X.-F.: Monitoring of Distributed Memory Multicomputer Programs. Technical Report
The Message Passing Interface (MPI) Forum, http://www.mpi-forum.org/
Pillet, V., Laboarta, J., Cortes, T., Girona, S.: PARAVER: A Tool to visualize and Analyze Parallel Code, CEPBA#UPC Report RR-95#03, University of Politencia, Catalonia (1995)
Automatic Performance Analysis: Real Tools (APART), http://www.fz-juelich.de/apart/
Michael, H.J., Viescas, J.L.: SQL Queries for Mere Mortals: A Hands-On Guide to Data Manipulation in SQL. Addison-Wesley Press, USA
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Chakravarthi, S., Kumar, C.R.K., Skjellum, A., Prahalad, H.A., Seshadri, B. (2003). A Model for Performance Analysis of MPI Applications on Terascale Systems. In: Dongarra, J., Laforenza, D., Orlando, S. (eds) Recent Advances in Parallel Virtual Machine and Message Passing Interface. EuroPVM/MPI 2003. Lecture Notes in Computer Science, vol 2840. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-39924-7_15
Download citation
DOI: https://doi.org/10.1007/978-3-540-39924-7_15
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-20149-6
Online ISBN: 978-3-540-39924-7
eBook Packages: Springer Book Archive