Southern CMG
Home
SCMG Meetings
Officers
Interested in presenting at SCMG or sponsoring a meeting?
CMG National

SCMG Meeting Atlanta
October 5, 2005

Location:

SAS Institute Inc.
Atlanta Plaza
Executive Briefing Room (Rm 3206)
950 East Paces Ferry Road NE, Suite 3200
Atlanta, GA 30326

 


Agenda:

Time Session Presenter
8:00 – 8:45
Registration and Continental Breakfast
SAS Institute
8:45 - 9:35
A Theoretical Systems Management Application with a User-Centric Focus
Yinka Alaran
9:45 - 10:35 An Introduction to the IBM Processor Capacity Reference for IBM System z9 and eServer zSeries
(zPCR)
Walt Caprice
10:35 - 11:00 Break  
11:00 - 11:50 DDF Performance Analysis - Does it Really Have to be This Complicated Robert Chaney
12:00 - 1:15
Lunch
SAS Institute
12:15 - 1:00
Vendor Presentation SAS Institute
1:15 - 2:15
Capacity and Performance Free-For-All  
2:25 - 3:15
Case studies: What happens when SPE is implemented poorly in a software project? Kevin Mobley
3:25 - 4:15
Follow-up capacity discussion  
4:15 - 4:30 Concluding Remarks and Giveaways  

Speakers:

Walt Caprice
Walt Caprice has spent the last 33 years working in the computer industry. He spent his first 18 years working in Philadelphia for
Burroughs Corp. (Tech Rep.), Westinghouse in Lester (Systems Programmer), GTE in Mount Laurel (Systems Programmer) and IBM in Philadelphia (A branch office Large Systems SE and a Market Support Rep in the MVS Performance Group of the Area Systems Center). Walt has spent the last 15 years working at the IBM Washington Systems Center in Gaithersburg, MD. He has a worldwide reputation in the areas of z/OS and z/Series performance. Walt is a frequent presenter on performance topics at IBM as well as user conferences.

Kevin Mobley
Kevin Mobley created and managed several software performance engineering teams, groups of highly experienced engineering and software specialists who not only have deep knowledge of web and CRM applications, UNIX and Microsoft, but of networking and messaging technologies as well. These teams have focused on one goal: making sure software systems are free of performance and scalability bottlenecks, inefficiencies, memory leaks and other related problems that can manifest as usage increases.

Mobley has thirteen years of creating software systems for fortune 1000 companies in financial services, retail and travel sectors, and the Federal Government's defense department and the State Government of Georgia. He saved tens of millions of dollars for clients and employers by architecting and deploying high performance software systems. He has generated a minimum of one million dollars yearly in consulting revenue for software performance engagements.

Abstracts:

An Introduction to the IBM Processor Capacity Reference for IBM System z9 and eServer zSeries (zPCR)
Walt Caprice, Senior Consulting I/T Specialist, IBM Corporation
The zPCR tool has been used for years by IBM’ers and Business Partners to accurately estimate the capacity difference between different IBM mainframes. This tool is scheduled to be made available to customers as a no charge tool in 4Q2005. This session will provide an overview of the tool itself as well as where to get further information on the tool once it is generally available.

Case studies: What happens when SPE is implemented poorly in a software project?
Kevin Mobley, Fidelity Information Services
SPE at the end of the Development cycle
Issues: More than 45,000 hours of was spent on functional development of an enterprise application to replace an existing system with poor performance. The work on the new system was characterized as "just an upgrade," yet there was an assumption that the new application would performance than the application it was replacing. Only 200 hours was allocated to run a performance test at the end of the project.
Result: The actual abridged SPE effort at the end of the project uncovered "show stopper" issues with the new application, however there was not enough time to fix everything. The go-live date moved back five months in order to repair the most critical performance issues to bring performance up to the benchmark of the replacement application. Actually making the new application performance better than the old application was not feasible. The performance of the new enterprise application project was poorly received by the user community, resulting in a political nightmare for executives and the development staff.
Lack of an Enterprise SPE Plan
Issues: SPE methodology was followed for a single application and issues were uncovered before "go live" date. Other systems that interact with the application did not follow a SPE process
Result: The surrounding systems were slow and the application that utilized SPE was "perceived" to be slow. There were continuous war room meetings to optimize complete system. Executive management was at a lost to he "real" performance problem.


Sponsors:

SAS Institute
SAS IT Management Solutions let you go beyond traditional IT performance management and leverage the full potential of each IT resource across the enterprise. SAS IT Management Solutions provide integrated and intuitive products for IT management across the enterprise, sophisticated analytical reporting and data visualization and reliable information on IT usage, resources, services and costs.