Advertisement

PERISCOPE: An Online-Based Distributed Performance Analysis Tool

  • Shajulin Benedict
  • Ventsislav Petkov
  • Michael Gerndt
Conference paper

Abstract

This paper presents PERISCOPE - an online distributed performance analysis tool that searches for a wide range of performance bottlenecks in parallel applications. It consists of a set of agents that capture and analyze application and hardware-related properties in an autonomous fashion. The paper focuses on the Periscope design, the different search methodologies, and the steps involved to do an online performance analysis. A new graphical user-friendly interface based on Eclipse is introduced. Through the use of this new easy-to-use graphical interface, remote execution, selection of the type of analysis, and the inspection of the found properties can be performed in an intuitive and easy way. In addition, a real-world application, namely, the GENE code, a grand challenge problem of plasma physics is analyzed using Periscope. The results are illustrated in terms of found properties and scalability issues.

Keywords

Performance Data Performance Property High Performance Computing Scalability Issue User Region 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Andreas Knüpfer, Holger Brunst, Jens Doleschal, Matthias Jurenz, Matthias Lieber, Holger Mickler, Matthias S. Müller and Wolfgang E. Nagel. The Vampir Performance Analysis Tool-Set. In Proc. of the 2nd Int. Work. on Parallel Tools for HPC, HLRS, Stuttgart, pages 139-155, Springer Publications, July 2008. Google Scholar
  2. 2.
    Chen, Y. and Parker, S. E. A δf particle method for gyrokinetic simulations with kinetic electrons and electromagnetic perturbations. In Comput. Phys. 189, 2 (Aug. 2003), DOI: http://dx.doi.org/10.1016/S0021-9991(03)00228-6. pages 463-475, 2003. zbMATHCrossRefMathSciNetGoogle Scholar
  3. 3.
    Eric Clayberg and Dan Rubel. Eclipse Plug-ins. In Addison-Wesley Professional, ISBN 978-0-321-55346-1 pages 107-135, 2008. Google Scholar
  4. 4.
    Markus Geimer, Felix Wolf, Brian J. N. Wylie, and Bernd Mohr. Scalable parallel trace-based performance analysis. In Proc. of the 13th Eur. PVM/MPI Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface (EuroPVM/MPI 2006), pages 303–312, Bonn, Germany, 2006. Google Scholar
  5. 5.
    M. Gerndt and K. Fürlinger. Specification and detection of performance problems with ASL. Conc. and Computation: Prac. & Exp., 19(11):1451–1464, Aug 2007. CrossRefGoogle Scholar
  6. 6.
    Michael Gerndt and Edmond Kereku. Search strategies for automatic performance analysis tools. In Anne-Marie Kermarrec, Luc Boug, and Thierry Priol, editors, Euro-Par 2007, volume 4641 of LNCS, pages 129–138. Springer, 2007. Google Scholar
  7. 7.
    Jeffrey Vetter and Chris Chambreau. mpiP: Lightweight, Scalable MPI Profiling. http://mpip.sourceforge.net, 2008.
  8. 8.
    F. Jenko. Massively parallel vlasov simulation of electromagnetic drift-wave turbulence. In Comp. Phys. Comm. 125 2000. Google Scholar
  9. 9.
    B.P. Miller, M.D. Callaghan, J.M. Cargille, J.K. Hollingsworth, R.B. Irvin, K.L. Karavanic, K. Kunchithapadam, and T. Newhall. The Paradyn parallel performance measurement tool. IEEE Computer, Vol. 28, No. 11, pp. 37-46, 1995. Google Scholar
  10. 10.
    Philip C. Roth and Barton P. Miller. The distributed performance consultant and the sub-graph folding algorithm: On-line automated performance diagnosis on thousands of processes. In Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (PPoPP’06), March 2006. Google Scholar
  11. 11.
    Shajulin Benedict, Matthias Brehm, Michael Gerndt, Carla Guillen, Wolfram Hesse and Ventsislav Petkov. Automatic Performance Analysis of Large Scale Simulations. In PROPER 2009, (in press), Springer Publishers 2009. Google Scholar
  12. 12.
    Sameer S. Shende and Allen D. Malony. The TAU parallel performance system. International Journal of High Performance Computing Applications, ACTS Collection Special Issue, 2005. Google Scholar
  13. 13.
    Felix Wolf and Bernd Mohr. Automatic performance analysis of hybrid MPI/OpenMP applications. In Proceedings of the 11th Euromicro Conference on Parallel, Distributed and Network-Based Processing (PDP 2003), pages 13–22. IEEE Computer Society, February 2003. Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Shajulin Benedict
    • 1
  • Ventsislav Petkov
  • Michael Gerndt
  1. 1.Fakultät für Informatik I10Technische Universität MünchenGarchingGermany

Personalised recommendations