Computational Quality of Service

As computational science progresses toward ever more realistic multiphysics and multiscale applications, no single research group can effectively select or tune all components of a given application, and no solution strategy can seamlessly span the entire spectrum efficiently. Common component interfaces, along with programming language interoperability and dynamic composability, are key features of component technology that enable easy access to suites of independently developed algorithms and implementations; via the Common Component Architecture (CCA), such capabilities are now making inroads in scientific computing. The challenge then becomes how, during runtime, to make the best choices for reliability, accuracy, and performance.

Researchers in the CCA Working Group on Computational Quality of Service (CQoS) are addressing this challenge by developing tools for CQoS, or the automatic selection and configuration of components to suit a particular computational purpose and environment. The two main facets of CQoS tools are (1) measurement and analysis infrastructure and (2) control infrastructure for dynamic component replacement and domain-specific decision making.

Collaborations with four SciDAC applications teams motivate this work: parallel mesh partitioning in combustion (CFRFS project, BES), managing resources in quantum chemistry (BES), and efficiently solving linear systems in fusion and accelerator models (FACETS project, FES, and COMPASS project, HEP/NP/BES). For example, the quantum chemists recently developed a prototype component that automates the selection from among several integral evaluators in molecular computations; this approach achieved substantial performance improvement (greater than 25% in some preliminary tests) with respect to the worst-case configuration that a novice user might employ.

We are also collaborating with researchers in the Performance Engineering Research Institute (PERI) on the design and implementation of CQoS component infrastructure, in particular in the definition of common database formats and interfaces for representing performance data and associated metadata, both of which are required for implementing CQoS adaptive strategies.

• Quantum Chemistry
  • Regular recurring telecons: info and working notes
• Combustion
• Linear Solvers

Notes from session on 18 March, 2008:

• CQoS telecon: Tues, March 18, 2008 ... Most materials are accessible via this link.

Notes from session on 19 February, 2008:

• CQoS telecon: Tues, Feb 19, 2008 ... Most materials are accessible via this link.

Notes from CQoS BOF at the winter 2008 CCA Meeting

• CQoS BOF: Fri, 18 Jan, 2008 ... Most materials are accessible via this link.

Notes from session on 9 January, 2008:

• CQoS telecon: Wed, 9 Jan, 2008 ... Most materials are accessible via this link.

Notes from session on 6 December, 2007:

• CQoS telecon: Thu, 6 Dec, 2007 ... Most materials are accessible via this link.

Notes from CQoS BOF at the autumn 2007 CCA Meeting

• CQoS BOF: Fri, 12 Oct, 2007 ... Most materials are accessible via this link.

Notes from Session on 24 September, 2007:

• CQoS telecon: Mon, 24 Sept, 2007 ... Most materials are accessible via this link.

Notes from Session on 27 August, 2007:

• CQoS telecon: Mon, 27 Aug, 2007 ... Most materials are accessible via this link. Brief notes follow:
• Next telecon agenda: applications folks will provide specific examples of the material that they want to store in CQoS databases (runtime and/or historical) to help facilitate configuring/reconfiguring components.
• To do: Determine where to stash working documents so that these are accessible only to our working group, not the world at large. Options include:
  • Sameer volunteered to host at Univ of Oregon
  • David mentioned that eventually infrastructure to support this will be available via cca-forum.org ... Question: What is the timeframe for this?

Notes from Session on 18 July, 2007:

• CCA Pre-Meeting: 18 July, 2007, 1:30 - 5:30 pm ET, Silver Spring, MD ... Most materials are accessible via this link. Brief notes follow:
• Establish monthly group telecons to sync on overall progress. Next session: Aug 27, 9 am PT
• Applications teams' highest priority for CQoS infrastructure: tools for performance database query and management
• Next short-term goals (everyone report to the group on Aug 27): Quantum chemistry and combustion teams will develop detailed scenarios of how they initially want to use (some of) the CQoS tools. The CQoS infrastructure developers will focus on providing preliminary performance database query and management capabilities for use by the applications teams.

Action Items:

• Motivating Applications:
  • Jaideep et al.: Develop specific plan for using CQoS tools in combustion simulations
  • Joe, Meng-Shiou, Masha: Develop specific plan for using CQoS tools in quantum chemistry simulations

• CQoS Testbed:
  • Rob: Extend preliminary 2D data component according to needs of combustion application
  • Li, Boyana, Lois: Extend TOPS solver components as needed to support adaptive linear solvers in testbed
  • Other folks: Provide additional CQoS testbed content (simplified application kernels, etc.)

• Infrastructure:
  • Tammy: Integrate experimental semantics capabilities into main Babel repository
  • Allen, Sameer, Alan: Extend PerfDMF capabilities based on input from CQoS users
  • Li, Boyana (collaborating with Allen, Sameer, and Alan): Develop prototype performance database query and management components as part of CQoS infrastructure

CCA Pre-Meeting: 24 January 2007, 1:30 - 5:30 pm PT, Snoqualmie, WA [LCM: Need to add links to January presentations in pdf format]

[LCM: Please add additional references here.]

• L. McInnes, J. Ray, R. Armstrong, T. Dahlgren, A. Malony, B. Norris, S. Shende, J. Kenny, and J. Steensland, Computational Quality of Service for Scientific CCA Applications: Composition, Substitution, and Reconfiguration, Argonne National Laboratory preprint ANL/MCS-P1326-0206, February, 2006. [paper].
• A. Malony, S. Shende, N. Trebon, J. Ray, R. Armstrong, C. Rasmussen, and M. Sottile, Performance Technology for Parallel and Distributed Component Software, Concurrency and Computation: Practice and Experience, Wiley, 2004. [paper].
• B. Norris, J. Ray, R. Armstrong, L. McInnes, D. E. Bernholdt, W. R. Elwasif, A. D. Malony and S. Shende, Computational Quality of Service for Scientific Components, Proceedings of the International Symposium on Component-Based Software Engineering (CBSE7), Edinburgh, Scotland, May 2004, also available as Argonne National Laboratory preprint ANL/MCS-P1131-0204. [paper].
• J. Ray, N. Trebon, S. Shende, R. C. Armstrong, and A. Malony, Performance Measurement and Modeling of Component Applications in a High Performance Computing Environment : A Case Study, 18th International Parallel and Distributed Computing Symposium, April 2004, Santa Fe, NM. Also available as Technical Report SAND2003-8631, Sandia National Laboratories, Livermore, CA, Nov. 2003. [paper].
• S. Shende, A. D. Malony, C. Rasmussen, and M. Sottile, A Performance Interface for Component-Based Applications, Proc. International Workshop on Performance Modeling, Evaluation, and Optimization of Parallel and Distributed Systems, IPDPS'03, IEEE Computer Society, 278, 2003. [paper].
• S. Lefantzi, J. Ray, and S. Shende, Strong Scalability Analysis and Performance Evaluation of a SAMR CCA-based Reacting Flow Code Poster, SC2003 Conference, Nov. 2003. [poster].
• P. Hovland, K. Keahey, L. McInnes, B. Norris, L. F. Diachin, and P. Raghavan, A Quality of Service Approach for High-Performance? Numerical Components, Proceedings of Workshop on QoS in Component-Based Software Engineering, Software Technologies Conference, Toulouse, France, June 2003, also available as Argonne National Laboratory preprint ANL/MCS-P1028-0203. [paper].

[LCM: Add pointers to related work here.]


Mailing list: https://www.cca-forum.org/mailman/listinfo/cqos-wg

Working Group Coordinator: Lois Curfman McInnes, ANL