Meeting agenda

Thursday, March 31, 2005

	Session
8:30 – 9:00	Breakfast, Registration
9:00 – 9:30	Welcome and Announcements
9:30 – 10:20	Millennium Performance Problems What will performance tools look like in the 21st century? What are the tough computer performance and capacity planning challenges that these tools will need to address and how should we set about developing those tools? It's difficult making predictions, especially predictions about the future, but in this talk I will identify five difficult problems: 1. Performance visualization 2. Self-measuring applications 3. The von Neumann bottleneck 4. The Internet simulator 5. Quantum computers that I would like to see addressed sooner rather than later this century. These problems have time horizons that are too long to be of immediate commercial interest; which raises a question about how best to stimulate enough interest to get them solved. The notion of a well-defined challenge or problem (sometimes with a monetary reward attached) has often provided a significant motivating influence in other sciences such as mathematics and physics. With that analogy in mind, I will propose and explain these five problems as the Millennium Performance Problems. Organizations such CMG might even consider holding a competition to extend this list. Dr. Neil Gunther,  Performance Dynamics Company Neil Gunther, M.Sc., Ph.D. is a leading industry computer performance consultant who founded Performance Dynamics Company (www.perfdynamics.com) in 1994. Prior to that, Dr. Gunther held research and management positions at San Jose State University, JPL/NASA (Voyager and Galileo missions), Xerox PARC research center and Pyramid/Siemens Technology. His performance and capacity planning classes have been given to such worldwide organizations as Boeing, FedEx, Motorola, Stanford University, and Sun Microsystems. Dr. Gunther is the author of numerous papers on computer performance topics, as well as two books: THE PRACTICAL PERFORMANCE ANALYST, McGraw-Hill, 1998 and most recently ANALYZING COMPUTER SYSTEM PERFORMANCE WITH PERL::PDQ, Springer-Verlag, 2005. He is well-known to CMG audiences for his presentations since 1993, and more recently for the extremely popular GUERRILLA CAPACITY PLANNING columns in the CMG MeasureIT online magazine. In 1996 Dr. Gunther was awarded Best Technical Paper at CMG, and in 1997 he was nominated for the A.A. Michelson Award. Dr. Gunther was born in Melbourne, Australia (which accounts for the odd accent) and is a member of the AMS, ACM, CMG, IEEE, SIGMETRICS and USENIX.
10:20– 10:40	Break
10:40 – 11:30	ITIL and Performance Management This presentation will briefly introduce the concepts and content of the Information Technology Infrastructure Library (ITIL), its relevance to IT, how it is evolving and the challenges of implementation. The focus of the presentation will then turn to ITIL's role in performance and capacity management. ITIL is a process driven framework with adds organization and discipline to how work is done in IT. From that perspective performance management benefits from more reliable information with which to assess performance and from the structure that enables those responsible for monitoring, assessing, recommending improvements for and implementing modifications to IT performance and capacity to be efficient and effective at what they do. Dialogue and feedback on ITIL and its relationship to performance management is encouraged. Loy Allen, Perot Systems Loy Allen is a leader of Perot Systems' Technology Management Consulting practice. He is responsible for guiding the development of and implementing IT related business processes, management practices and governance principles that bring business value for the Clients of Perot Systems. Loy has also worked to develop methodologies, such as the Roadmap to Business Innovation (RBI) and ITIL related best practices with in Perot Systems and to develop the Business Engineering Management track of the Engineer Development Program, an extensive training and leadership development program. Loy has an ITIL foundation certificate and over 5 years of experience developing ITIL processes, performing assessments, advising Clients and implementing ITIL processes and enabling organizational structures and technology.
11:30 – 12:30	Lunch - Sponsored by HyPerformix
12:30 – 1:20	Wide-Angle Capacity Planning at SSgA This presentation examines capacity planning at State Street Global Advisors (SSgA), which uses automated techniques to predict capacity issues across large numbers of servers. The methodology involves calculation of growth trends for each hour of the day, identification of peaks, and projections into the future, for several performance metrics. Factors such as exceptional conditions, new application roll-outs, business plans, and technology directions are also included in the analysis. Michael R. Stadelmann, State Street Global Advisors Michael Stadelmann is a technical analyst and project manager specializing in performance analysis and capacity management at State Street Global Advisors. He has experience at many Fortune 500-sized companies in projects involving UNIX, Windows, and OS/390(Z/OS) systems, including the implementation of monitoring, management, analysis, reporting systems, and business level reporting. His work has involved the measurement and analysis of enterprise performance for mainframe and distributed applications, such as web, client/server, and batch applications. He has written numerous papers and articles, and has presented at CMG, ICCM, Share, and other forums. At BGS Systems for 17 years, he led BEST/1 development efforts, support, and services on UNIX, Windows, OS/390, OpenVMS, and AS/400. At BMC Software, Mr. Stadelmann worked as an individual contributor as well as a senior manager for an international business-consulting unit. In this capacity he successfully consulted with clients at all levels of management to solve their technical and business needs.
1:20 – 2:10	Workload Manager - Our Experiences Implementing Goals By Online Transaction Response Time New application technology presented another opportunity to exploit Workload Manager capabilities. This session goes through setting up applications environments, analyzing the DB2 data, and points out some of the pit falls you can run into. Len Jejer, Raymond Smith, The Hartford Insurance Group Ray Smith has been in the field of Mainframe performance and capacity planning for the last twenty years. The last eight years he has been working for The Hartford Insurance Group as an Online Performance Analyst. Len Jejer has been in Data Processing for 25 years as an application programmer, systems programmer and performance analyst. He has been at The Hartford for the past 3 years.
2:10 – 2:30	Break
2:30 – 3:20	PASASM: A Method for the Performance Assessment of Software Architectures Architectural decisions are among the earliest made in a software development project. They are also the most costly to fix if, when the software is completed, the architecture is found to be inappropriate for meeting quality objectives. Thus, it is important to be able to assess the impact of architectural decisions on quality objectives such as performance and reliability at the time that they are made. This presentation describes PASA, a method for performance assessment of software architectures. It was developed from our experience in conducting performance assessments of software architectures in a variety of application domains including web-based systems, financial applications, and real-time systems. PASA uses the principles and techniques of software performance engineering (SPE) to determine whether an architecture is capable of supporting its performance objectives. The method may be applied to new development to uncover potential problems when they are easier and less expensive to fix. It may also be used when upgrading legacy systems to decide whether to continue to commit resources to the current architecture or migrate to a new one. The steps in the PASA method are described together with an overview of the techniques used to perform an architecture assessment. These techniques include the use of architectural styles, performance antipatterns and performance modeling as well as strategies for identification of alternatives for meeting performance objectives if problems are discovered. The method is illustrated with an example drawn from an actual assessment. Dr. Connie U. Smith and Dr. Lloyd G. Williams Dr. Lloyd G. Williams is President and Principal Consultant at PerfX (formerly Software Engineering Research), where he specializes in the development and evaluation of software architectures to meet quality objectives including performance, scalability reliability, modifiability, and reusability. His experience includes work on systems in fields such as process control, avionics, telecommunications, financial services, e-commerce and other Web-based systems, software development tools and environments, and medical instrumentation. He is an internationally-known leader in the theory and practice of Software Performance Engineering (SPE). Lloyd is the author of numerous technical papers and has presented professional development seminars and consulted on software development for more than 100 organizations worldwide. Dr. Connie U. Smith a principal consultant of the Performance Engineering Services Division of L&S Computer Technology, Inc., is known for her work in defining the field of SPE and integrating SPE into the development of new software systems. Dr. Smith received the Computer Measurement Group's prestigious AA Michelson Award for technical excellence and professional contributions for her SPE work. She also authored the original SPE book: Performance Engineering of Software Systems, published in 1990 by Addison-Wesley, and approximately 100 scientific papers. She is the creator of the SPE•ED™ performance engineering tool. She has over 25 years of experience in the practice, research and development of the SPE performance prediction techniques. Together, Drs. Williams and Smith have over 50 years of experience in software development. They have worked together for more than 15 years to help clients design and implement software that meets performance objectives. They have published numerous technical papers and articles, and are the authors of Performance Solutions: A Practical Guide to Creating Responsive, Scalable Software, published by Addison-Wesley.
3:20 – 4:10	Price performance optimization model In configuring a server, which is better, an additional CPU or a few more DIMMs? What’s the optimal balance between the I/O and memory configuration? IT staff and capacity planners face these sorts of questions all the time, but there appears to be no general solution or accepted methodology for making sub-system tradeoffs. This session presents such a methodology, based on both modeled performance and sub-system cost for an OLTP workload. Analysis of results indicates that memory configuration plays a crucial role in optimizing for price-performance. Jay Veazey , Hewlett-Packard