November, 2009
by Dr. Tim R. Norton, Subject Area Chair for the CMG 2009 Modeling and Statistics Track
Capacity Planning is certainly getting more interesting with multi-tiered applications, virtualization of several types and cloud computing. Over the last few years, I've had numerous conversations where someone will say that the cost of adding more hardware is significantly less than the cost associated with measurements and modeling so why bother. Now that virtualization in becoming more prominent this attitude seems to be increasing. After all, virtual systems are free, right? Even if the hardware isn't really free, is it possible to use a virtualization pool to avoid the whole planning process? Advocates of this approach point out that it is hard, almost impossible, to get good business transaction measurements and forecasts. When the business becomes as dynamic as the virtual infrastructure, it sometimes seems like the only salvation is rapid reaction instead of methodical planning. On the other hand, how many of us capacity planners have watched management buy round after round of "cheap" hardware just to see performance stay the same, that is to say, lousy, and overall costs skyrocket. It seems like some form of planning is really needed but what is the balance? I've seen a company spend months on analysis involving dozens of employees, from engineers to administrators to managers, just to determine if a $2000 upgrade is really needed. Companies really can't afford to use the same methodology for a $5K server as they do for the $15M mainframe. That's pretty clear but where is the crossover? Is it always the same for every company? Is it always the same within a company?
As Subject Area Chair for the Modeling and Statistics Track this year I have organized a panel discussion to talk about the future of modeling. What I wanted to do was gather together a lot of different points-of-view about modeling and talk about some of these issues. How are real companies solving, or not solving, them? How can vendors help with new and better tools? What are university research centers doing that will help in the future? I felt that the best way to do this was to ask the top experts to come together to talk about the key topics. When you look at the list of panelists (see their biographies below) you can see that there is broad interest in this topic. Almost everyone invited agreed to participate and for those who could not it was because of scheduling or personal reasons, not for lack of interest. This panel will not be a series of prepared comments with a few minutes for questions at the end (I don't think I could do their biographies justice in the available time, let alone give each panelist enough time to present their ideas!) This will be an audience driven panel with the questions coming from you, the CMG attendees. The abstract below has my thoughts but what do you want to ask? Here is your chance to ask the tough questions and hear not only one answer but also likely, several different ones! The panelists are primed with the topics below, and I'll start with those if the audience is shy, but the objective is to hear from you, the CMGer, about what you need, to do your job better.
If you are interested and will be in Dallas for the CMG national conference, please bring your questions with you or email them to me at modeling@cmg.org and join us for:
PANEL: Hardware's Cheap so Why Do Modeling?
Session 356 (Abstract 9174) Tuesday 4:00-5:00 PM in Grapevine 4
Abstract:
The cost of hardware is rapidly trending down while other costs are rising even more rapidly. The result of this interplay is that cost saving opportunities are shrinking while the analysis takes increasingly more time, effort and money. This panel of world-renowned experts in application and systems modeling will candidly discuss this and other questions related to the future of modeling as a tool to achieve business objectives.
Discussion areas:
The panelists:
Dr. Jeff Buzen is best know as the "B" in BGS Systems and the originator of the BEST/1 family of performance modeling products. He is an A.A. Michelson Award winner, a past President of CMG, and an elected member of the National Academy of Engineering.
Adam Grummitt, Director of Metron and chairman of UKCMG, has been playing with computers since graduating from Cambridge way back. Doing research in mass spectrometry he used a PDP-8 and early IBM mainframes. He has since been an analyst, designer and programmer for end-users, software houses and as a consultant. He has been in performance management and capacity planning for many years, specializing in performance engineering. He is a founding Director of Metron and now does international strategic consultancy. He is on the executive of UKCMG and itSMF, gives workshops on capacity management worldwide and is a well-known speaker at CMG.
Ron Kaminski has been a performance analyst, capacity planner and modeler since the mid 1980s on every platform you can name except mainframes, and his most recent gig has those too! Ron has successfully modeled wars, stock exchange crashes, governments, social security systems, several gigantic retailers, airlines, grocery store chains and many internet startups and web systems. Ron also often goes "extra product" in pursuit of process pathology detection algorithms and other mayhem consuming resources that they shouldn't, and he shares his findings regularly at CMGs around the world. Ron thinks that every firm needs modeling tools, and will help you justify them!
Dr. Daniel A. Menasce is the Senior Associate Dean and a Professor of Computer Science at the Volgenau School of IT & Engineering, George Mason University, VA. He received his PhD in Computer Science from UCLA in 1978 and is the 2001 recipient of CMG's A.A. Michelson Award. He is a Fellow of the ACM and an author of over 195 papers and five books.
Dr. Michael Salsburg is a Distinguished Engineer for Unisys Systems and Technology, and is the Chief Architect for Real Time Infrastructures and Cloud Computing, which area strategic programs for Unisys. Previously, Dr. Salsburg was president of his own company, Performance & Modeling, Inc. Dr. Salsburg received his Bachelor of Science from the University of Pittsburgh in Mathematics in August 1972. He received his Masters of Science from the University of Delaware in Computer Science (Software Engineering and Artificial Intelligence) in 1982. He received his Doctor of Philosophy in Mathematics from Drexel University (Probability and Statistics) in 1992. Dr. Salsburg has been awarded two international patents in the area of infrastructure performance modeling algorithms and software. In addition, he has published dozens of papers and has lectured world-wide on the topic of computer performance evaluation. His current activities are focused on business-driven infrastructure management.
Debbie Sheetz is a Sr. Staff Consultant based in BMC Customer Support, at the Waltham, Massachusetts/USA location. She provides applied solutions for performance analysis and capacity planning challenges for customers, business partners, and BMC field consultants. She works with product R&D and marketing on refining existing solutions and designing new solutions. Prior to working with Distributed Systems performance management products, she had extensive involvement with AS/400 and mainframe product support and development. Originally hired to work on the first version of BEST/1 at BGS Systems, she has 33 years experience developing and supporting performance analysis software with BMC Software/BGS Systems.
Dr. Connie U. Smith, CTO of the Performance Engineering Services Division of L&S Computer Technology, Inc., is known for her pioneering work in defining the field of Software Performance Engineering (SPE) and integrating SPE into the development of new software systems. Dr. Smith received the Computer Measurement Group's prestigious A.A. Michelson lifetime achievement award for technical excellence and professional contributions for her SPE work. She is the author of the original SPE book, Performance Engineering of Software Systems, published in 1990 by Addison-Wesley, the more recent book: Performance Solutions: A Practical Guide to Building Responsive, Scalable Software also published by Addison-Wesley, and approximately 100 scientific papers. She is the creator of the SPE·EDTM performance engineering tool. She has over 25 years of experience helping clients design and implement software that meets performance requirements.
Bill Zahavi is a senior technologist at EMC Corp. where he has been developing tools and products for managing the performance of the Symmetrix Systems. Prior to EMC Bill was with Digital Equipment Corp. and with Sperry/Unisys always focusing on performance analysis and the development of tools in that arena. Bill is a designer and developer of both instrumentation and modeling tools and over the years has presented his work at CMG on several occasions. Currently Bill chairs the Boston Chapter of CMG.
Dr. Boris Zibitsker is a Chairman and CTO of BEZ Systems. He founded BEZ in 1983 with the objective to provide modeling and proactive performance modeling management solutions for complex computer systems based on relational DBMSs. From 1983 to 1990, he served as Adjunct Associate professor at DePaul University, where he developed graduate courses on "Computer and Information System Modeling", "Queuing Theory with Computer Systems Applications" and "Computer and Communication Networks Design and Analysis". Dr. Zibitsker has taught Performance and Capacity Management seminars for the Relational Institute, founded by pioneers of relational DBMS Dr. Tedd Codd and Chris Date. Since 1989, Boris focused BEZ Systems on developing commercial modeling tools for applications based on DBMS systems incorporating parallel processing. In 1992, BEZ Systems announced BEZPlus for Teradata, 1995 BEZPlus for DB2 , 1996 BEZPlus for Oracle and in 2008 BEZVision for multi-tier distributed systems. BEZ Systems named 2008 "Cool Vendor" by Gartner Group. Boris focuses his research on multi-tier distributed parallel processing environments. He has presented many papers and taught seminars in North and South America, Europe, Asia and Africa.