Program for Performance and Capacity 2014 by CMG

Time Salon 4 Salon 5 Salon 6 Conference Room E Conference Room C

Monday, November 3

07:00 am-08:00 am Breakfast
08:00 am-09:15 am 171: Business Service Profiling and Advanced Analytics 172: WLM - Performing a Cursory WLM Review 173: Hands On Lab: Jmeter for Performance Testing 174: Performance Optimization on Co-processors 175: Taking Your ITSM Strategy Beyond the Service Desk Workshop
09:15 am-09:30 am Break
09:30 am-10:30 am 171: Business Service Profiling and Advanced Analytics 172: WLM - Performing a Cursory WLM Review 173: Hands On Lab: Jmeter for Performance Testing 174: Performance Optimization on Co-processors 175: Taking Your ITSM Strategy Beyond the Service Desk Workshop
10:30 am-10:45 am Break
10:45 am-11:45 am 171: Business Service Profiling and Advanced Analytics 172: WLM - Performing a Cursory WLM Review 173: Hands On Lab: Jmeter for Performance Testing 174: Performance Optimization on Co-processors 175: Taking Your ITSM Strategy Beyond the Service Desk Workshop
12:00 pm-01:00 pm Lunch
01:15 pm-02:30 pm Annual Business Meeting
02:45 pm-03:30 pm       Session Chair/Monitor Training 1st Timers Orientation
03:30 pm-03:45 pm Break
03:45 pm-04:45 pm 251: Stack Metrics: A Taxonomy of Metrics Supporting the Capacity Planning Stack 252: Parameter Estimation of Asymptotically Improved Super-Serial Scalability Law 253: CMG-T: Networking 254: Entropy-Based Anomoly Detection for SAP z/OS Systems [Best Paper] 255: Performance Measurements and Post Data Reduction: Using Radview WebLoad and Analytics to Load Test Georgia Tech Research Institute Portal Web Site in a Linux Environment
04:45 pm-05:00 pm Break
05:00 pm-06:00 pm 261: Vaulting Over Career Speedbumps - Transforming Loss Into Opportunity 262: The Real Story on Cloud and Analytics Performance 263: CMG-T: Networking 264: Challenges in Implementing CodeVita - Coding Competition for Students 265: A Guide to EMC VNX Performance Metrics
06:30 pm-07:45 pm Dinner
08:00 pm-10:30 pm Networking Session

Tuesday, November 4

07:00 am-08:00 am Breakfast
08:00 am-09:15 am 301: The High Performance Computing Story at Georgia Tech - Salon 4
09:15 am-09:30 am Break
09:30 am-10:30 am 311: Technical Debt - The Secret Drain on Capacity Dollars and Performance 312: PANEL: Virtualization: Something Old, Something New, Something Borrowed, Something Blue 313: Rationale for an IPv6 PDM Extension Header 314: No Session Scheduled 315: Adjusting Performance Testing to the World of Agile
10:30 am-10:45 am Break
10:45 am-11:45 am 321: Performance Implications of NUMA - What You Don't Know Could Hurt You! 322: PANEL: Measuring New Workloads: Cloud Analytics, Mobile, Social 323: Heisenberg and Utilization 324: Fear and Loathing in Automation-Land 325: Data Delivery Network: Next-Generation Application Performance
12:00 pm-01:00 pm Lunch
01:15 pm-02:15 pm 331: Ensuring Performance in a Fast-Paced Environment 332: Graph Analysis Trends and Opportunities 333: CMG-T: Virtualization - Part 1 334: Avoiding Analysis Paralysis: The Search for the Ultimate Question 335: Performance Measurements and Tools in a VDI Environment - Our Experiences
02:15 pm-02:30 pm Break
02:30 pm-03:30 pm 341: How Fast Can You Go - Optimizing Memory Cache 342: PANEL: Modern Industry Trends and Performance Assurance 343: CMG-T: Virtualization - Part 2 344: It's Time to Retire Our 1970's User Demand Model for Transaction Computing and Adopt One that Reflects Modern Web User Traffic Flow 345: A Specification of Performance Test Types
03:30 pm-03:45 pm Break
03:45 pm-04:45 pm 351: Ensuring the Performance of Mobile Apps on Every Device and Network 352: Building Performance in a Retail Order Management System 353: CMG-T: Virtualization - Part 3 354: Slow Time Versus Downtime 355: How to Save $450K by Capacity Planning an Oracle Server
04:45 pm-05:00 pm Break
05:00 pm-06:00 pm BOFs BOF: Pros, Cons & Experiences with Array Based Replication vs. Server & Application Based Replications Regional Officers Meeting BOFs BOFs
06:30 pm-07:45 pm Dinner
08:00 pm-10:30 pm Networking Session

Wednesday, November 5

07:00 am-08:00 am Breakfast
08:00 am-09:15 am 401: Mobile Web Stress: Understanding the Neurological Impact of Poor Performance - Salon 4
09:15 am-09:30 am Break
09:30 am-10:30 am 411: Sources of Traffic Demand Variability and Use of Monte Carlo for Network Capacity Planning 412: PANEL: How Fast is Fast Enough? 413: Evaluating Performance and Capacity with Calculated Metrics 414: CMG-T: Modeling & Forecasting - Part 1 415: webMethods Performance - Best Practices and Case Study from a Large Retail
10:30 am-10:45 am Break
10:45 am-11:45 am 421: Managing What You Don't Own - How to Test and Monitor Cloud-based Services 422: Meter Anything from Component to a Cloud: A General Purpose Utilization Monitor 423: Load Testing: Which Tool to Choose? 424: CMG-T: Modeling & Forecasting - Part 2 425: How to Present Your IT Business Case to Your Finance Partner
12:00 pm-01:00 pm Lunch
01:15 pm-02:15 pm 431: Capacity Planning for Location-Independent Computing (LIC) 432: Balancing Cost and Performance Tradeoffs: Benchmarking VMware Storage Performance and Best Practices 433: CMG-T:Java - Part 1 434: CMG-T: Modeling & Forecasting - Part 3 435: Accounting for Application Performance in an Agile Environment
02:15 pm-02:30 pm Break
02:30 pm-03:30 pm 441: ITIL Based Capacity Management for the Internet of Things 442: PANEL: APM Tools and Technologies: What Do You Need? 443: CMG-T:Java - Part 2 444: As if by Magic...the Dashboard Appeared - Building Useful Business Service Capacity Reports 445: Performance Engineering Mobile Applications: Tips and Tricks
03:30 pm-03:45 pm Break
03:45 pm-04:45 pm 451: The Case of the Broken Development Process 452: Demystifying Oracle Databases Capacity Management with Workload Characterization 453: CMG-T:Java - Part 3 454: How to Handle Noisy Neighbors? 455: System z Processor Consumption, or What is Consuming the CPU?
04:45 pm-05:00 pm Break
05:00 pm-06:00 pm     463: Inside Products: IP Problem Finder 464: Syspertec - Improving the Performance of Web-Initiated CICS Transaction Workloads 465: Metron-Athene - 360° Capacity Management with athene® and athene® ES/1
06:00 pm-06:30 pm   2015 Kick-Off & Planning Meeting      
06:30 pm-07:45 pm Dinner
08:00 pm-10:30 pm Networking Session

Thursday, November 6

07:00 am-08:00 am Breakfast
08:00 am-09:15 am 501: Disruptive Technology: Let's Upset Your Quiet IT Life - Salon 4
09:15 am-09:30 am Break
09:30 am-10:30 am 511: Newer Computing Professionals, and a Newer Paradigm - What Do Young Mainframe CPE (Computer Performance Evaluation) and Capacity Planners Need to Know? 512: Using Big Data to Identify Micro-Outages in Real Time 513: CMG-T: Analytics - Part 1 514: The Art and Science of Technical Presentations 515: Workload Analytics for Engineering Performance of Information Systems
10:30 am-10:45 am Break
10:45 am-11:45 am 521: Capping Capacity vs. Capping Demand 522: Defining and Improving the Client Side Performance of Web Applications 523: CMG-T: Analytics - Part 2 524: Calibrate Workload Model for Accurate Performance Testing 525: Software Testability for Performance Testing
12:00 pm-01:00 pm Lunch
01:15 pm-02:15 pm 531: CMG-T: Mainframe - Part 1 - What's Happening to the Mainframe? Cloud? Mobile? Social? Big Data? 532: PANEL: What Workloads Work Best with Flash 533: User Perspective: Detecting Memory Leaks Versus Memory Hogs on UNIX 534: Practical Applications of Locality Sensitive Hashing for Unstructured Data 535: Performance Engineering of Investment Bank's Risk Compute Grid Application
02:15 pm-02:30 pm Break
02:30 pm-03:30 pm 541: CMG-T: Mainframe - Part 2 - Top New z/OS Performance Functions Every Sysprog Should Understand 542: Key Performance Metrics for Your Continuous Delivery Pipeline 543: Common Metrics Don't Count 544: In-Memory tables 545: Using Regression to Estimate Query Resource Consumption
03:30 pm-03:45 pm Break
03:45 pm-04:45 pm 551: CMG-T: Mainframe - Part 3 - Understanding the RMF CPU Activity Report 552: Dynamic Derivation of Analytical Performance Models in Autonomic Computing Environments [Best Paper] 553: A New Approach to Capacity Forecasting 554: wuqiSpank.org Brings Transparency to the Subterranean World of SQL Performance 555: Insider’s View of a Real-World Project: Innovative Strategies and Solutions for Performance Testing Composite Applications
04:45 pm-05:00 pm Break
05:00 pm-06:00 pm BOFs BOFs BOFs   BOFs
06:30 pm-07:45 pm Dinner
08:00 pm-10:30 pm Networking Session

Monday, November 3

Monday, November 3, 07:00 - 08:00

CONF: Breakfastgo to top

Dining Room - Lobby Level
Workshop Attendees Only

Monday, November 3, 08:00 - 09:15

171 (Workshops): Business Service Profiling and Advanced Analyticsgo to top

Workshop
Richard Gimarc, CA Technologies; Adrian Johnson
Room: Salon 4

A Business Service Profile is a set of related data that characterizes the usage and capacity of a business service. The Profile uses metadata to relate individual data elements describing workload, resource usage, and infrastructure. The result is a logical network of data describing a Business Service that can be viewed, analyzed, and summarized along any of the dimensions defined by the supporting metadata. This workshop will introduce, illustrate and leverage the concept of a Business Service Profile to perform the following tasks: • Create the fundamental set of data and relationships required to analyze the usage and capacity of a business service. It will be demonstrated that the Profile enables the capacity planner and performance analyst to focus his/her efforts.

• Develop a data center capacity plan. The high-level data center plan is developed from the set of lower-level business service capacity plans that describe the usage and capacity of the business services supported by the data center. The key ingredient used to develop a business service capacity plan is the Business Service Profile.

• Monitor the day-to-day performance of business services. Today's data centers support hundreds of business services across thousands of servers; this creates a problem because of the large volume of data that must be examined. Advanced analytics can be applied to a Business Service Profile to develop intelligent dashboards that focus the analyst's attention on what warrants his/her attention. This workshop will explore the concept, development, presentation and utility of a Business Service Profile. The following topics will be addressed: • Business Service Profile terminology • Techniques for creating and presenting a Profile • Sample Business Service Profiles • Creating a business service capacity plan using a set of Business Service Profiles • Developing a reference template (based on the Profile) for monitoring the usage and capacity of a business service • Examples of using advanced analytics to create dashboards for monitoring business service usage and expected capacity demands.

172 (Workshops): WLM - Performing a Cursory WLM Reviewgo to top

Workshop
Peter Enrico, Enterprise Performance Strategies, Inc.
Room: Salon 5

During this workshop Peter Enrico will walk the attendees through a cursory review of their WLM Goals and Importance Levels. In particular, during this workshop Peter instruct the attendees to do the following reviews: Importance Level Review Velocity Goal Review Response Time Goal Review Multiple Period Review CMG attendees planning on attending this workshop are strongly encouraged to contact Peter Enrico at least a week prior to the workshop if they want to submit to him raw SMF data. For attendees that submit raw SMF data, Peter will generate a set of performance reports needed to do the WLM analysis reviews discussed in the workshop. If you want to submit Peter raw SMF data for this workshop, please send him an email at Peter.Enrico@EPStrategies.com

173 (Workshops): Hands On Lab: Jmeter for Performance Testinggo to top

Workshop
Robert Jahn, Collaborative Consulting
Room: Salon 6

This workshop offers hands on exercise for the process of designing, running, and compiling load test result using Jmeter. Jmeter is an open source desktop application under the Apache project. Jmeter is designed to load test functional behavior and measure performance for Web Applications as well as other protocols. Pre-requisites: Bring a laptop. Workshop is hands-on. Jmeter will require Java JDK 6 and we will install Jmeter. The instructor will be running Windows 7 laptop, but Mac Netbooks work fine too. Feel free to install and setup Jmeter ahead of time, but we will take time in class to ensure everyone is setup. Objectives:

By the end the student should know how to: • Install Jmeter • Use core Jmeter features for constructing a script • Control the load levels and duration of a test script • Collect and analysis test results • Make use of Jmeter plugins • Use command line execution

Agenda Section 1: Jmeter Introduction Section 2: Jmeter installation Section 3: Labs • Lab 1 Basic use of Jmeter with simple script • Lab 2 Adding in functionality to scripts • Lab 3 Using Jmeter Proxy Server to record script • Lab 4 Running tests and data analysis Section 4: Jmeter Testing Topologies Section 5: More Labs • Lab 5 Jmeter from the Command Line • Lab 6 Jmeter plugins Section 6: Where to get more information

174 (Workshops): Performance Optimization on Co-processorsgo to top

Workshop
Manoj Nambiar, TCS
Conference Room E

CPU technologies have been changing thanks to various physical and power limitations. New co-processors (also known as accelerators) – namely many integrated core (MiC from Intel) and GPU (NVidia, AMD etc) architectures have been evolving – they consume less power than traditional CPUs but are capable delivering equally good performance. With power considerations being high in scaling data-centers it expected to see more proliferation of these coprocessors. Simply porting does not guarantee best performance - applications have to be re-engineered keeping underlying hardware in perspective - and it’s not just parallelization. There are a lot more tricks to enable applications achieve optimal performance on the new platforms. Of course it also helps to understand the application well – nevertheless there are ways to optimize application performance even if you are not a domain expert. The participants in this workshop will learn methods/tricks to optimize application performance on Intel MiC coprocessors and NVIDIA GPUs.

Workshop Contents • Introduction to Coprocessors

• NVIDIA GPU platform o Hardware Architecture o Programming with CUDA C o Performance Monitoring o Performance Tuning  including code based examples and case studies

• Intel MiC platform o Hardware Architecture o Parallel programming with MiC  including introduction to openmp o Performance Monitoring o Performance Tuning  including code based examples and case studies

Participant Pre-requisites: Familiarity with basic computer architectures, Programming in C/Fortran

175 (Workshops): Taking Your ITSM Strategy Beyond the Service Desk Workshopgo to top

Workshop
Charles Williams
Conference Room C

How does an IT organization get away from reactive, unscientific service costing guess work? One way is to implement the Information Technology Infrastructure Library (ITIL) framework's IT Financial Management practices to predict current and future keep-the-lights-on (KTLO) activities and service costs, and the impact of change? Effectively implementing the ITIL framework's IT Financial Management disciplines requires a consumption-based planning model approach that (1) aligns IT workload activities to business needs, (2) defines IT workload drivers and related activity cost, and (3) provides the capability to quickly and accurately define current costs, and to estimate the impact of change. This paper describes a new approach for Information Technology (IT) service cost modeling. Workshop Outline: • Introduction • IT Service Cost Modeling Primer • A New Approach to IT Service Cost Modeling • Building A Simple Consumption-based Model (CBM) Example • Benefits of Consumption-based Modeling (CBM) • Conclusion

Monday, November 3, 09:15 - 09:30

CONF: Breakgo to top

Monday, November 3, 09:30 - 10:30

171 (Workshops): Business Service Profiling and Advanced Analyticsgo to top

Workshop
Richard Gimarc, CA Technologies; Adrian Johnson
Room: Salon 4

A Business Service Profile is a set of related data that characterizes the usage and capacity of a business service. The Profile uses metadata to relate individual data elements describing workload, resource usage, and infrastructure. The result is a logical network of data describing a Business Service that can be viewed, analyzed, and summarized along any of the dimensions defined by the supporting metadata.

172 (Workshops): WLM - Performing a Cursory WLM Reviewgo to top

Workshop
Peter Enrico, Enterprise Performance Strategies, Inc.
Room: Salon 5

During this workshop Peter Enrico will walk the attendees through a cursory review of their WLM Goals and Importance Levels. In particular, during this workshop Peter instruct the attendees to do the following reviews: Importance Level Review Velocity Goal Review Response Time Goal Review Multiple Period Review

173 (Workshops): Hands On Lab: Jmeter for Performance Testinggo to top

Workshop
Robert Jahn, Collaborative Consulting
Room: Salon 6

This workshop offers hands on exercise for the process of designing, running, and compiling load test result using Jmeter. Jmeter is an open source desktop application under the Apache project. Jmeter is designed to load test functional behavior and measure performance for Web Applications as well as other protocols.

174 (Workshops): Performance Optimization on Co-processorsgo to top

Workshop
Manoj Nambiar, TCS
Conference Room E

CPU technologies have been changing thanks to various physical and power limitations. New co-processors (also known as accelerators) – namely many integrated core (MiC from Intel) and GPU (NVidia, AMD etc) architectures have been evolving – they consume less power than traditional CPUs but are capable delivering equally good performance. With power considerations being high in scaling data-centers it expected to see more proliferation of these coprocessors. Simply porting does not guarantee best performance - applications have to be re-engineered keeping underlying hardware in perspective - and it’s not just parallelization. There are a lot more tricks to enable applications achieve optimal performance on the new platforms. Of course it also helps to understand the application well – nevertheless there are ways to optimize application performance even if you are not a domain expert. The participants in this workshop will learn methods/tricks to optimize application performance on Intel MiC coprocessors and NVIDIA GPUs.

Workshop Contents • Introduction to Coprocessors

• NVIDIA GPU platform o Hardware Architecture o Programming with CUDA C o Performance Monitoring o Performance Tuning  including code based examples and case studies

• Intel MiC platform o Hardware Architecture o Parallel programming with MiC  including introduction to openmp o Performance Monitoring o Performance Tuning  including code based examples and case studies

Participant Pre-requisites: Familiarity with basic computer architectures, Programming in C/Fortran

175 (Workshops): Taking Your ITSM Strategy Beyond the Service Desk Workshopgo to top

Workshop
Charles Williams
Conference Room C

How does an IT organization get away from reactive, unscientific service costing guess work? One way is to implement the Information Technology Infrastructure Library (ITIL) framework's IT Financial Management practices to predict current and future keep-the-lights-on (KTLO) activities and service costs, and the impact of change? Effectively implementing the ITIL framework's IT Financial Management disciplines requires a consumption-based planning model approach that (1) aligns IT workload activities to business needs, (2) defines IT workload drivers and related activity cost, and (3) provides the capability to quickly and accurately define current costs, and to estimate the impact of change. This paper describes a new approach for Information Technology (IT) service cost modeling.

Monday, November 3, 10:30 - 10:45

CONF: Breakgo to top

Monday, November 3, 10:45 - 11:45

171 (Workshops): Business Service Profiling and Advanced Analyticsgo to top

Workshop
Richard Gimarc, CA Technologies; Adrian Johnson
Room: Salon 4

A Business Service Profile is a set of related data that characterizes the usage and capacity of a business service. The Profile uses metadata to relate individual data elements describing workload, resource usage, and infrastructure. The result is a logical network of data describing a Business Service that can be viewed, analyzed, and summarized along any of the dimensions defined by the supporting metadata.

172 (Workshops): WLM - Performing a Cursory WLM Reviewgo to top

Workshop
Peter Enrico, Enterprise Performance Strategies, Inc.
Room: Salon 5

During this workshop Peter Enrico will walk the attendees through a cursory review of their WLM Goals and Importance Levels. In particular, during this workshop Peter instruct the attendees to do the following reviews: Importance Level Review Velocity Goal Review Response Time Goal Review Multiple Period Review

173 (Workshops): Hands On Lab: Jmeter for Performance Testinggo to top

Workshop
Robert Jahn, Collaborative Consulting
Room: Salon 6

This workshop offers hands on exercise for the process of designing, running, and compiling load test result using Jmeter. Jmeter is an open source desktop application under the Apache project. Jmeter is designed to load test functional behavior and measure performance for Web Applications as well as other protocols.

174 (Workshops): Performance Optimization on Co-processorsgo to top

Workshop
Manoj Nambiar, TCS
Conference Room E

CPU technologies have been changing thanks to various physical and power limitations. New co-processors (also known as accelerators) – namely many integrated core (MiC from Intel) and GPU (NVidia, AMD etc) architectures have been evolving – they consume less power than traditional CPUs but are capable delivering equally good performance. With power considerations being high in scaling data-centers it expected to see more proliferation of these coprocessors. Simply porting does not guarantee best performance - applications have to be re-engineered keeping underlying hardware in perspective - and it’s not just parallelization. There are a lot more tricks to enable applications achieve optimal performance on the new platforms. Of course it also helps to understand the application well – nevertheless there are ways to optimize application performance even if you are not a domain expert. The participants in this workshop will learn methods/tricks to optimize application performance on Intel MiC coprocessors and NVIDIA GPUs.

Workshop Contents • Introduction to Coprocessors

• NVIDIA GPU platform o Hardware Architecture o Programming with CUDA C o Performance Monitoring o Performance Tuning  including code based examples and case studies

• Intel MiC platform o Hardware Architecture o Parallel programming with MiC  including introduction to openmp o Performance Monitoring o Performance Tuning  including code based examples and case studies

Participant Pre-requisites: Familiarity with basic computer architectures, Programming in C/Fortran

175 (Workshops): Taking Your ITSM Strategy Beyond the Service Desk Workshopgo to top

Workshop
Charles Williams
Conference Room C

How does an IT organization get away from reactive, unscientific service costing guess work? One way is to implement the Information Technology Infrastructure Library (ITIL) framework's IT Financial Management practices to predict current and future keep-the-lights-on (KTLO) activities and service costs, and the impact of change? Effectively implementing the ITIL framework's IT Financial Management disciplines requires a consumption-based planning model approach that (1) aligns IT workload activities to business needs, (2) defines IT workload drivers and related activity cost, and (3) provides the capability to quickly and accurately define current costs, and to estimate the impact of change. This paper describes a new approach for Information Technology (IT) service cost modeling.

Monday, November 3, 12:00 - 13:00

CONF: Lunchgo to top

Dining Room - Lobby LEvel
Workshop Attendees Only

Monday, November 3, 13:15 - 14:30

CONF: Annual Business Meetinggo to top

Salon 4
CMG Board of Directors

Monday, November 3, 14:45 - 15:30

CONF: Session Chair/Monitor Traininggo to top

CMG Volunteer Coordinator
Conference Room E

CONF: 1st Timers Orientationgo to top

CMG Volunteer
Conference Room C

Monday, November 3, 15:30 - 15:45

CONF: Breakgo to top

Monday, November 3, 15:45 - 16:45

251 (Featured Speaker): Stack Metrics: A Taxonomy of Metrics Supporting the Capacity Planning Stackgo to top

Richard Gimarc, CA Technologies; Amy Spellmann, The 451 Group; Adrian Johnson, CA Technologies
Room: Salon 4

The Capacity Planning Stack was introduced as a way to organize, discuss and implement a capacity plan that spans the entire Digital Infrastructure. This paper will examine the fundamental metric classes that support the Capacity Planning Stack. Stack metrics are partitioned into three main classes; Business, IT and Facilities. Each metric class supports different layers within the Capacity Planning Stack. Collectively the metric classes bridge the gap between the high-level business requirements and the lower-level data center that hosts the IT infrastructure.

Viewing the universe of metrics in terms of the taxonomy supporting the Stack enables the practitioner to more easily organize and delineate the data requirements for Digital Infrastructure Capacity Planning. For successful capacity planning, these metrics must be correlated and communicated. This paper will illustrate the use of the technique with a number of practical examples.

252 (CP): Parameter Estimation of Asymptotically Improved Super-Serial Scalability Lawgo to top

Jayanta Choudhury, Teamquest Corp.
Room: Salon 5

253 (CMG-T): CMG-T: Networkinggo to top

CMG-T
Manoj Nambiar, TCS
Room: Salon 6

254 (ITSM): Entropy-Based Anomoly Detection for SAP z/OS Systems [Best Paper]go to top

Tim Browning
Conference Room E

255 (APM): Performance Measurements and Post Data Reduction: Using Radview WebLoad and Analytics to Load Test Georgia Tech Research Institute Portal Web Site in a Linux Environmentgo to top

James Logan, Georgia Tech
Conference Room C

Monday, November 3, 16:45 - 17:00

CONF: Breakgo to top

Monday, November 3, 17:00 - 18:00

261 (IPT): Vaulting Over Career Speedbumps - Transforming Loss Into Opportunitygo to top

Denise Kalm, Kalm Kreative
Room: Salon 4

262 (CONF): The Real Story on Cloud and Analytics Performancego to top

Elisabeth Stahl, IBM
Room: Salon 5

Don't miss this session on the real story of the latest in Cloud and Analytics performance. Workload optimization and Cloud models will be highlighted from a performance and non-functional requirements focus. Challenges and complexities with Cloud and Analytics performance metrics will be discussed, as well as benchmark best practices and trends.

263 (CMG-T): CMG-T: Networkinggo to top

CMG-T
Manoj Nambiar, TCS
Room: Salon 6

264 (PET): Challenges in Implementing CodeVita - Coding Competition for Studentsgo to top

Amol Khanapukar, TCS
Conference Room E

265 (CP): A Guide to EMC VNX Performance Metricsgo to top

Jim Quigley, Consolidated Edison of New York
Conference Room C

Monday, November 3, 18:30 - 19:45

CONF: Dinnergo to top

Dining Room - Lobby Level

Monday, November 3, 20:00 - 22:30

CONF: Networking Sessiongo to top

Lounge/Lobby Bar - Lobby Level

Tuesday, November 4

Tuesday, November 4, 07:00 - 08:00

CONF: Breakfastgo to top

Dining Room - Lobby Level

Tuesday, November 4, 08:00 - 09:15

301 (Featured Speaker): The High Performance Computing Story at Georgia Tech - Salon 4go to top

Keynote
Ron Hutchins, PhD, Chief Technology Officer, Georgia Tech

The keynote will highlight the High Performance Computing (HPC) business and Information Technology at Georgia Tech. IT infrastructure will be discussed as well as a unique virtualization model that maximizes Georgia Tech's computing.

Tuesday, November 4, 09:15 - 09:30

CONF: Breakgo to top

Tuesday, November 4, 09:30 - 10:30

311 (CP): Technical Debt - The Secret Drain on Capacity Dollars and Performancego to top

Denise Kalm, Kalm Kreative
Room: Salon 4

312 (CONF): PANEL: Virtualization: Something Old, Something New, Something Borrowed, Something Bluego to top

Moderator: Kevin Mobley
Room: Salon 5

As a follow-on to the Keynote presentation, this panel will examine retro, current, and future IT virtualization technologies. Panelists will compare and contrast virtualization models, highlight successful implementations, and take questions from the audience on their experiences.

PANELISTS: Rick Lebsack, Ron Hutchins, Elisabeth Stahl, Johnathan Paul, and Michael Salsburg

313 (Featured Speaker): Rationale for an IPv6 PDM Extension Headergo to top

Nalini Elkins, Inside Products, Inc.
Room: Salon 6

To diagnose performance and connectivity problems, metrics on passive traffic are critical for timely end-to-end problem resolution. Such diagnostics may be real-time or after the fact, but must not impact an operational production network. These metrics are defined in the IPv6 Performance and Diagnostic Metrics Destination Option (PDM). The base metrics are: packet sequence number and packet timestamp. Other metrics may be derived from these for use in diagnostics. This document specifies such metrics, their calculation, and usage.

314 (CONF): No Session Scheduledgo to top

Conference Room E

315 (PET): Adjusting Performance Testing to the World of Agilego to top

Alex Podelko, Oracle
Conference Room C

Tuesday, November 4, 10:30 - 10:45

CONF: Breakgo to top

Tuesday, November 4, 10:45 - 11:45

321 (APM): Performance Implications of NUMA - What You Don't Know Could Hurt You!go to top

Claire Cates, SAS
Room: Salon 4

322 (CONF): PANEL: Measuring New Workloads: Cloud Analytics, Mobile, Socialgo to top

PANEL
Elisabeth Stahl, Moderator
Room: Salon 5

Panelists include: Elisabeth Stahl, Martin Spier, Steve Weisfeldt, Pete Mastin

323 (PET): Heisenberg and Utilizationgo to top

Joseph Temple, Low Country North Shore Consulting, LLC
Room: Salon 6

324 (CONF): Fear and Loathing in Automation-Landgo to top

Denise Kalm, Kalm Kreative
Conference Room E

Remember when you only had a handful of servers or a single mainframe to manage? Back in those days, we also had teams of experts working on capacity and performance management. Now, one person is managing 1000′s of servers, at least one private cloud, virtualization and more with changes being applied to the system every day, if not every hour. It isn’t possible to manage as you have in the past – you need to change the game, eliminating manual processes wherever possible. Learn how to work smarter and have more fun on the job by letting go of the past and venturing boldly into the future of capacity planning. Build a vision and the management skills to create it.

325 (APM): Data Delivery Network: Next-Generation Application Performancego to top

Vic Cekvenich, Apakau, Inc.
Conference Room C

Tuesday, November 4, 12:00 - 13:00

CONF: Lunchgo to top

Dining Room - Lobby Level

Tuesday, November 4, 13:15 - 14:15

331 (Featured Speaker): Ensuring Performance in a Fast-Paced Environmentgo to top

Martin Spier, Netflix
Room: Salon 4

Netflix accounts for more than a third of all traffic heading into American homes at peak hours. Making sure users are getting the best possible experience at all times is no simple feat and performance is at the core of this experience. In order to ensure performance and maintain development agility in a highly decentralized environment/(organization?), Netflix employs a multitude of strategies, such as production canary analysis, fully automated performance tests, simple zero-downtime deployments and rollbacks, auto-scaling clusters and a fault-tolerant stateless service architecture. We will present a set of use cases that demonstrate how and why different groups employ different strategies to achieve a common goal, great performance and stability, and detail how these strategies are incorporated into development, test and devops with minimal overhead.

332 (Featured Speaker): Graph Analysis Trends and Opportunitiesgo to top

James Riedy, Georgia Tech
Room: Salon 5

High-performance graph analysis is unlocking knowledge in problems like anomaly detection in computer security, community structure in social networks, and many other data integration areas. While graphs provide a convenient abstraction, real-world problems' sparsity and lack of locality challenge current systems. This talk will cover current trends ranging from massive scales to low-power, low-latency systems and summarize opportunities and directions for graphs and computing systems.

333 (CMG-T): CMG-T: Virtualization - Part 1go to top

Johnathan Paul, Siemens
Room: Salon 6

334 (CONF): Avoiding Analysis Paralysis: The Search for the Ultimate Questiongo to top

Adrian Johnson, CA Technologies
Conference Room E

We all know the Ultimate Answer is 42 but when faced with the performance and capacity analysis of large and complex IT systems, what is the Ultimate Question? Playing with numbers can be fun, but we don’t all have a ‘brain the size of a planet’ so it can also be time-consuming and to what value…? This presentation aids the search for the Ultimate Question by addressing data and analysis requirements in terms of a scale ranging from interesting through useful to necessary.

335 (PET): Performance Measurements and Tools in a VDI Environment - Our Experiencesgo to top

Bandula Abeysundara, Teradici Corp.
Conference Room C

Tuesday, November 4, 14:15 - 14:30

CONF: Breakgo to top

Tuesday, November 4, 14:30 - 15:30

341 (APM): How Fast Can You Go - Optimizing Memory Cachego to top

Claire Cates, SAS
Room: Salon 4

342 (PET): PANEL: Modern Industry Trends and Performance Assurancego to top

Panel
Alex Podelko, Moderator
Room: Salon 5

343 (CMG-T): CMG-T: Virtualization - Part 2go to top

Johnathan Paul, Siemens
Room: Salon 6

344 (PET): It's Time to Retire Our 1970's User Demand Model for Transaction Computing and Adopt One that Reflects Modern Web User Traffic Flowgo to top

James Brady, State of Nevada
Conference Room E

345 (PET): A Specification of Performance Test Typesgo to top

Alan Korncoff, Boeing
Conference Room C

Tuesday, November 4, 15:30 - 15:45

CONF: Breakgo to top

Tuesday, November 4, 15:45 - 16:45

351 (Featured Speaker): Ensuring the Performance of Mobile Apps on Every Device and Networkgo to top

Steve Weisfeldt, Neotys
Room: Salon 4

Today's remarkable mix of cloud computing, ever-smarter mobile devices, and prolific application development has changed the way applications are developed and tested. These days deployed applications deliver different content and functionality depending on whether the user is accessing it via a browser, a cell phone, a tablet, etc. Moreover, applications are accessed over a myriad of network configurations, including wireless and mobile networks. These new approaches have a significant impact on the performance of applications; creating a previously unseen set of testing challenges. Every client connecting to the application has a different kind of network constraint (mobile network, Wi-Fi, Ethernet…etc.) and these networks constraints have serious implications on the behavior and performance of applications. A crucial part of the load testing process is to be able to emulate all network constraints and introduce not only connection speeds but also parameters such as packet loss and network latency are required to test for the most realistic scenarios. This presentation will discuss the ramifications of these new technologies and constraints on testing for mobile application performance. It will suggest approaches and considerations to ensure that high application performance is delivered to all end-users, at all times, regardless of device.

352 (PET): Building Performance in a Retail Order Management Systemgo to top

Ramapantula Udaya Shankar, TCS
Room: Salon 5

353 (CMG-T): CMG-T: Virtualization - Part 3go to top

Johnathan Paul, Siemens
Room: Salon 6

354 (Featured Speaker): Slow Time Versus Downtimego to top

Tammy Everts
Conference Room E

Eventually, every website fails. If it's a household-name site, like Amazon, then news of that failure gets around faster than a rocket full of monkeys. That's because downtime hurts. As a for-instance, in 2013 Amazon suffered a 40-minute outage that allegedly cost the company $5 million in lost sales. That's a big number, and everybody loves big numbers. But when it comes to performance-related losses, is it the biggest number?

In this session, we'll review real-world examples that compare the cost of site slowdowns versus outages. We'll also talk about how to overcome the challenges of creating as much urgency around the topic of slow time as there is around the topic of downtime.

355 (CP): How to Save $450K by Capacity Planning an Oracle Servergo to top

Jamie Baker, Metron Technologies
Conference Room C

Tuesday, November 4, 16:45 - 17:00

CONF: Breakgo to top

Tuesday, November 4, 17:00 - 18:00

CONF: BOFsgo to top

Birds of a Feather Sessions
Room: Salon 4

CONF: BOF: Pros, Cons & Experiences with Array Based Replication vs. Server & Application Based Replicationsgo to top

Birds of a Feather Sessions
Bryan Drake
Room: Salon 5

CONF: Regional Officers Meetinggo to top

Invitiation Only
Denise Kalm, CMG VP
Room: Salon 6

CONF: BOFsgo to top

Birds of a Feather Sessions
Conference Room E

CONF: BOFsgo to top

Birds of a Feather Sessions
Conference Room C

Tuesday, November 4, 18:30 - 19:45

CONF: Dinnergo to top

Dining Room - Lobby Level

Tuesday, November 4, 20:00 - 22:30

CONF: Networking Sessiongo to top

Lounge/Lobby Bar - Lobby Level

Wednesday, November 5

Wednesday, November 5, 07:00 - 08:00

CONF: Breakfastgo to top

Dining Room - Lobby Level

Wednesday, November 5, 08:00 - 09:15

401 (Featured Speaker): Mobile Web Stress: Understanding the Neurological Impact of Poor Performance - Salon 4go to top

Plenary Session
Tammy Everts

Slow pages hurt mobile user metrics, from bounce rate to revenues. At Radware, we wanted to understand the science behind this, so we engaged in the first documented study of the neurological impact of poor performance on mobile users. Your takeaway from this session is hard data that you can use to make a case for investing in mobile performance in your organization.

Wednesday, November 5, 09:15 - 09:30

CONF: Breakgo to top

Wednesday, November 5, 09:30 - 10:30

411 (CP): Sources of Traffic Demand Variability and Use of Monte Carlo for Network Capacity Planninggo to top

Alexander Gilgur, Google, Inc.
Room: Salon 4

412 (PET): PANEL: How Fast is Fast Enough?go to top

Panel
Alex Podelko, Moderator
Room: Salon 5

413 (APM): Evaluating Performance and Capacity with Calculated Metricsgo to top

Dale Feiste, Metron-Athene
Room: Salon 6

414 (CMG-T): CMG-T: Modeling & Forecasting - Part 1go to top

Michael Salsburg, Unisys
Conference Room E

Although most computing environments are heterogeneous, computer system modeling is, in most ways, platform neutral. The same techniques and tools can be used to model zSeries, Unix / Linux, and Windows. At the heart of these models is the essential queueing network. This course provides the details of the essential queueing network, including the necessary statistics that need to be collected from the system, as well as various modeling techniques that yield insights that cannot be gleaned from observing the actual computer system. Once the model is validated, it can be used to explore "what-if" scenarios where either the workload or the underlying configuration can be changed in the model so that the resulting service levels can be observed. If time permits, an additional section on the subject of time series estimation and forecasting will be presented. This course will not teach you everything you need, but it will give you a full survey of the various approaches with a full bibliography for future reference.

Part 1 - Modeling and Forecasting Basics This is the first of three sessions. Computer performance modeling is mainly focused on understanding how business activity and the infrastructure can be analyzed to understand the impact on IT services. The key to this activity is to understand how requests for service queue for usage of resources. This first session provides basic definitions, the history of modeling queueing systems and some basic analytical queueing models.

415 (PET): webMethods Performance - Best Practices and Case Study from a Large Retailgo to top

Sasidar Nemani, TCS
Conference Room C

Wednesday, November 5, 10:30 - 10:45

CONF: Breakgo to top

Wednesday, November 5, 10:45 - 11:45

421 (Featured Speaker): Managing What You Don't Own - How to Test and Monitor Cloud-based Servicesgo to top

Craig Hyde, Rigor
Room: Salon 4

The recent proliferation of SaaS and IaaS systems has reduced the amount of time, money, and expertise needed to deploy certain applications by orders of magnitude. However, once deployed, these services can be challenging to manage without the visibility and control allowed by self-hosted solutions.

This session presents processes and tools used to test and tune complex, cloud-based services by decoupling the performance of the underlying systems and service delivery. Several available techniques, technologies, and best practices for managing cloud-based services will be highlighted.

422 (Featured Speaker): Meter Anything from Component to a Cloud: A General Purpose Utilization Monitorgo to top

Bruse McNutt, IBM
Room: Salon 5

A persistent and recurring problem in the field of system performance management is to compare the current use of the system with the maximum use permitted by the workload, hardware and software. Measurements based on processor cycle counts provide a well understood and effective solution, but have insufficient scope to cover many problems of interest. This paper takes a fresh look at system utilization monitoring.

423 (PET): Load Testing: Which Tool to Choose?go to top

Alex Podelko, Oracle
Room: Salon 6

424 (CMG-T): CMG-T: Modeling & Forecasting - Part 2go to top

Michael Salsburg, Unisys
Conference Room E

Although most computing environments are heterogeneous, computer system modeling is, in most ways, platform neutral. The same techniques and tools can be used to model zSeries, Unix / Linux, and Windows. At the heart of these models is the essential queueing network. This course provides the details of the essential queueing network, including the necessary statistics that need to be collected from the system, as well as various modeling techniques that yield insights that cannot be gleaned from observing the actual computer system. Once the model is validated, it can be used to explore "what-if" scenarios where either the workload or the underlying configuration can be changed in the model so that the resulting service levels can be observed. If time permits, an additional section on the subject of time series estimation and forecasting will be presented. This course will not teach you everything you need, but it will give you a full survey of the various approaches with a full bibliography for future reference.

Session 2 - Queueing Network Models This is the second of three sessions. Building on session 1, the discussion turns to understanding the distribution of requests for service as well as the distribution for the service times for each request. It is shown how the understanding of these distributions contributes to developing accurate models that predict IT service end-to-end times. During this session, another approach to computer performance modeling, simulation modeling, is introduced. The basics of simulation modeling to predict computer performance are presented.

425 (IPT): How to Present Your IT Business Case to Your Finance Partnergo to top

Randy McCoy, DataKinetics
Conference Room C

Wednesday, November 5, 12:00 - 13:00

CONF: Lunchgo to top

Dining Room - Lobby Level

Wednesday, November 5, 13:15 - 14:15

431 (CP): Capacity Planning for Location-Independent Computing (LIC)go to top

Yiping Ding, Riverbed Technologies
Room: Salon 4

432 (CP): Balancing Cost and Performance Tradeoffs: Benchmarking VMware Storage Performance and Best Practicesgo to top

Priyanka Arora, Birla Institute of Technology & Science
Room: Salon 5

433 (CMG-T): CMG-T:Java - Part 1go to top

Peter Johnson, Unisys
Room: Salon 6

Attendees at these CMG-T sessions will benefit from my many years of doing Java performance tuning, including in our lab where we ran industry standard benchmarks, in our application excellence centers where I have helped tuning our customer's real-world applications, and in day-to-day operations of applications we have running in our data center. I will cover the following topics: 1) Analyzing the garbage collector (GC) a) Understanding how the GC works b) Gathering GC data (there are three different formats for this data) c) Graphing the GC data and understanding what the graphs mean d) Examining some real-world examples, what the GC graphs showed, what tuning was done, what the results were (hint: significantly improved performance) 2) Survey of various GC algorithms a) Description of the default collector b) Description of the parallel collector c) Description of the mostly-concurrent mark/sweep collector i) Using a parallel collector in conjunction with the mark/sweep collector d) Description of the new garbage-first collector available in JDK 7 e) Description of the pros and cons of each collector 3) Miscellaneous JVM tuning and tips for solving real-world Java issues.

434 (CMG-T): CMG-T: Modeling & Forecasting - Part 3go to top

Michael Salsburg, Unisys
Conference Room E

Although most computing environments are heterogeneous, computer system modeling is, in most ways, platform neutral. The same techniques and tools can be used to model zSeries, Unix / Linux, and Windows. At the heart of these models is the essential queueing network. This course provides the details of the essential queueing network, including the necessary statistics that need to be collected from the system, as well as various modeling techniques that yield insights that cannot be gleaned from observing the actual computer system. Once the model is validated, it can be used to explore "what-if" scenarios where either the workload or the underlying configuration can be changed in the model so that the resulting service levels can be observed. If time permits, an additional section on the subject of time series estimation and forecasting will be presented. This course will not teach you everything you need, but it will give you a full survey of the various approaches with a full bibliography for future reference.

Session 3 - Simulation and Forecasting This is the third of three sessions. It builds on the previous two sessions. Using the simulation concepts from the second session, the attendees are introduced to a simulation model that simulates a hypervisor that is used for server virtualization. The last portion of the session is focused on analytical methods to forecast trends. This includes the basics of linear regression as well as the basics of time series forecasting.

435 (Featured Speaker): Accounting for Application Performance in an Agile Environmentgo to top

Steve Weisfeldt, Neotys
Conference Room C

Many organizations today are struggling with questions about to account for the performance of an application in an Agile environment. The lack of a concrete set of best practices for incorporating performance into Agile testing has left many wondering, "Am I doing this right?" Many teams doing iterative development have a number of tests that are running with their Continuous Integration builds, but haven't been able to effectively incorporate load and performance testing into that automation. Furthermore, few teams write user stories and tasks in a way that capture performance goals along with functional specifications. In this session, I will discuss: • Benefits gained from being more Agile with performance testing • Challenges faced when testing performance in an iterative environment • Best practices for ensuring an app will perform after ever iteration • Tips for automating the load and performance testing of CI builds • Use cases from successful Agile performance testing implementations

Wednesday, November 5, 14:15 - 14:30

CONF: Breakgo to top

Wednesday, November 5, 14:30 - 15:30

441 (CP): ITIL Based Capacity Management for the Internet of Thingsgo to top

Chris Molloy, IBM
Room: Salon 4

442 (PET): PANEL: APM Tools and Technologies: What Do You Need?go to top

Panel
Alex Podelko, Moderator
Room: Salon 5

A diverse panel of APM experts will discuss different APM technologies and areas of their applicability: RUM (real-user monitoring), synthetic monitoring, deep diagnostics, network APM, IT Operations Analytics, and others. Pros and Cons of different approaches. Do we need an integrated tool or a set of best-of-breed products? How to evaluate tools? Where to start implementation?

Panelists: David Halbig (First Data), Craig Hyde (Rigor), Martin Spier (Netflix), Charles Johnson (Metron).

443 (CMG-T): CMG-T:Java - Part 2go to top

Peter Johnson, Unisys
Room: Salon 6

Attendees at these CMG-T sessions will benefit from my many years of doing Java performance tuning, including in our lab where we ran industry standard benchmarks, in our application excellence centers where I have helped tuning our customer's real-world applications, and in day-to-day operations of applications we have running in our data center. I will cover the following topics: 1) Analyzing the garbage collector (GC) a) Understanding how the GC works b) Gathering GC data (there are three different formats for this data) c) Graphing the GC data and understanding what the graphs mean d) Examining some real-world examples, what the GC graphs showed, what tuning was done, what the results were (hint: significantly improved performance) 2) Survey of various GC algorithms a) Description of the default collector b) Description of the parallel collector c) Description of the mostly-concurrent mark/sweep collector i) Using a parallel collector in conjunction with the mark/sweep collector d) Description of the new garbage-first collector available in JDK 7 e) Description of the pros and cons of each collector 3) Miscellaneous JVM tuning and tips for solving real-world Java issues.

444 (CP): As if by Magic...the Dashboard Appeared - Building Useful Business Service Capacity Reportsgo to top

Adrian Johnson, CA Technologies
Conference Room E

445 (PET): Performance Engineering Mobile Applications: Tips and Tricksgo to top

Mohit Verma, Tufts Health Plan
Conference Room C

Wednesday, November 5, 15:30 - 15:45

CONF: Breakgo to top

Wednesday, November 5, 15:45 - 16:45

451 (APM): The Case of the Broken Development Processgo to top

Michael Salsburg, Unisys
Room: Salon 4

452 (CP): Demystifying Oracle Databases Capacity Management with Workload Characterizationgo to top

Stefano Doni
Room: Salon 5

453 (CMG-T): CMG-T:Java - Part 3go to top

Peter Johnson, Unisys
Room: Salon 6

Attendees at these CMG-T sessions will benefit from my many years of doing Java performance tuning, including in our lab where we ran industry standard benchmarks, in our application excellence centers where I have helped tuning our customer's real-world applications, and in day-to-day operations of applications we have running in our data center. I will cover the following topics: 1) Analyzing the garbage collector (GC) a) Understanding how the GC works b) Gathering GC data (there are three different formats for this data) c) Graphing the GC data and understanding what the graphs mean d) Examining some real-world examples, what the GC graphs showed, what tuning was done, what the results were (hint: significantly improved performance) 2) Survey of various GC algorithms a) Description of the default collector b) Description of the parallel collector c) Description of the mostly-concurrent mark/sweep collector i) Using a parallel collector in conjunction with the mark/sweep collector d) Description of the new garbage-first collector available in JDK 7 e) Description of the pros and cons of each collector 3) Miscellaneous JVM tuning and tips for solving real-world Java issues.

454 (CP): How to Handle Noisy Neighbors?go to top

Rema Hariharan, Ebay
Conference Room E

455 (CONF): System z Processor Consumption, or What is Consuming the CPU?go to top

Peter Enrico, Enterprise Performance Strategies, Inc.
Conference Room C

The first step to any processor analysis is to understand your processor configuration and settings. The second step is to understand what workloads, address spaces, and transactions are consuming the fixed processor resource. It is only after understanding what and how the processor is being consumed can you conduct any sort of processor tuning or optimization exercise.

During this presentation Peter Enrico will show you how to conduct a processor resource consumption analysis. You will be provided with a top down approach to better understand processor measurements available to help you gain a drilldown insight into how the CPU resource is being consumed, and by what LPARs, Workloads, and transactions. Shown is what is known as a drill down approach for a processor performance analysis.

Wednesday, November 5, 16:45 - 17:00

CONF: Breakgo to top

Wednesday, November 5, 17:00 - 18:00

463 (CONF): Inside Products: IP Problem Findergo to top

Nalini Elkins
Room: Salon 6

464 (CONF): Syspertec - Improving the Performance of Web-Initiated CICS Transaction Workloadsgo to top

Exhibitor Presentation
Patrick Fournier
Conference Room E

An ever-growing number of consumers conduct their personal business and shop online every day. It sends avalanches of web-initiated transactions hitting the supporting - oftentimes CICS - applications. Unfortunately CICS and associated system components add over 90% overhead to this web-initiated TP workload which they are not designed for. This presentation exposes the causes for overhead with web-initiated CICS TP, and the technology behind a new minimalist transaction server that runs CICS transactions with up to 50% improvement in footprint, response times, and throughput. Discussion items include: 1) changing nature of CICS TP workloads, 2) understanding CICS TP overhead, 3) design brief for lighter and faster transaction server, 4) benchmarking against standard CICS/SOAP, 5) impact on CICS programs, 6) alternative architectures, 7) proposed implementation strategy, and 8) next steps.

465 (CONF): Metron-Athene - 360° Capacity Management with athene® and athene® ES/1go to top

Exhibitor Presentation
TBD
Conference Room C

Wednesday, November 5, 18:00 - 18:30

CONF: 2015 Kick-Off & Planning Meetinggo to top

Bryan Drake
Room: Salon 5

Wednesday, November 5, 18:30 - 19:45

CONF: Dinnergo to top

Dining Room - Lobby Level

Wednesday, November 5, 20:00 - 22:30

CONF: Networking Sessiongo to top

Lounge/Lobby Bar - Lobby Level

Thursday, November 6

Thursday, November 6, 07:00 - 08:00

CONF: Breakfastgo to top

Thursday, November 6, 08:00 - 09:15

501 (Featured Speaker): Disruptive Technology: Let's Upset Your Quiet IT Life - Salon 4go to top

Glenn Anderson, IBM

Ready for a break from the detailed bits and bytes of IT? Come join long time conference speaker and instructor Glenn Anderson as he explores the world of disruptive technologies - new ways of doing things that overturn traditional business methods and practices. Some you have heard of; others probably not. Augmented reality, ubiquitous computing, likeonomics, cognitive computing, the internet of things, and more. Exciting, challenging, entertaining and career enhancing. Come let Glenn through a wrench in your idea of the future.

Thursday, November 6, 09:15 - 09:30

CONF: Breakgo to top

Thursday, November 6, 09:30 - 10:30

511 (CP): Newer Computing Professionals, and a Newer Paradigm - What Do Young Mainframe CPE (Computer Performance Evaluation) and Capacity Planners Need to Know?go to top

Greg Caliri, BMC Software
Room: Salon 4

512 (Featured Speaker): Using Big Data to Identify Micro-Outages in Real Timego to top

Pete Mastin, Cedexis
Room: Salon 5

Major outages from Cloud and CDN providers are well documented. What flows beneath the radar are the more prosaic Micro-outages that permeate the Internet. Using Billions of Real End User Measurements a day and Heat-maps this talk discusses in detail the 150-300 impactful outages a month on over 100+ Platforms. These measurements must be captured and analyzed in real time to make them useful however. We also discuss how the data can be signal processed to make it actionable in real time.

513 (CMG-T): CMG-T: Analytics - Part 1go to top

Ray Wicks
Room: Salon 6

This is a two part session which reviews some of the statistical techniques which can be useful in performance analysis and capacity planning

Part 1 reviews some of the analytic (statistical) concepts and their psychology used in all of statistics. The process of seeing and describing reality in terms of numbers and graphs is foremost. This analysis is essential to grasping statistical concepts that follow. Emphasis in this part will provide the under pinning of more complex statistical ideas: average, distribution, standard deviation, coefficient of variation.

514 (CONF): The Art and Science of Technical Presentationsgo to top

Glenn Anderson, IBM
Conference Room E

Ever seen a great presenter and wondered – how does he do it? What makes it different from the way I present? Drawing on experiences speaking around the world, from classes to conferences to comedy club stand-up routines, Glenn will help you make your next presentation useful, memorable, and enjoyable for both you and your audience! Topics include preparation, speaking techniques, and the good, bad and ugly of PowerPoint. Glenn Anderson is a long-time instructor and presenter with IBM Technical Training, with a wealth of experience developing and delivering very technical material.

515 (PET): Workload Analytics for Engineering Performance of Information Systemsgo to top

Sundar Narayanan Subramanian
Conference Room C

Thursday, November 6, 10:30 - 10:45

CONF: Breakgo to top

Thursday, November 6, 10:45 - 11:45

521 (Featured Speaker): Capping Capacity vs. Capping Demandgo to top

John Baker, MVS Solutions
Room: Salon 4

While many organizations are enjoying the benefits of IBM’s Sub-Capacity pricing model in order to lower Monthly License Charges (MLC), fewer have embraced the financial opportunity offered via the optional ‘soft capping’ features Defined Capacity and Group Capacity. This paper will explore the rationale and propose solutions to allow full exploitation of these features.

522 (CONF): Defining and Improving the Client Side Performance of Web Applicationsgo to top

Craig Hyde, Rigor
Room: Salon 5

People have been trying to make websites faster ever since the browser was invented, and in that time, the web performance industry has made a vast number of advancements, each addressing specific performance bottlenecks. Thus, as the web has evolved, so have the methods for measuring and improving its performance.

This session will cover: ● A brief overview of the evolution of performance on the internet ● How to select and collect the proper metrics to define client-side performance ● Best practices for performance tuning modern web applications

523 (CMG-T): CMG-T: Analytics - Part 2go to top

Ray Wicks
Room: Salon 6

This is a two part session which reviews some of the statistical techniques which can be useful in performance analysis and capacity planning.

Part 2 reviews the basic techniques will be expanded to talk about the comparison of measurement results (T-test) and the use of techniques which can be useful in performance analysis and capacity planning, namely regression analysis and time series analysis (a.k.a. trending).

524 (PET): Calibrate Workload Model for Accurate Performance Testinggo to top

Benjamin Mao, Ticketmaster
Conference Room E

Web performance testing, in short, is to simulate real user access workload and web system behaviour in test environment. Performance engineers often spent significant time to build and validate performance test workload model with their best effort and knowledge. The reality is that the workload models in performance testing often missed their real-world scenario targets due to missing “significant non critical” web usages or lack of accurate workload model validation. The accuracy of workload model is directly reflected back into the understanding of web app usage pattern in production. Without an accurate workload model, not only performance test results could be incorrect,but also many following tasks, performance bottleneck recreating, server performance tuning, capacity planning,and application framework validation could be at wrong targets. I will walk you through a practical way of building accurate workload model that will be matching to the targeted prod workload much closer comparing to the workload model created in current performance testing practice. The high accurate workload model and followed performance test results will bring high confidence on web app performance of new releases to project team.

525 (PET): Software Testability for Performance Testinggo to top

Saikathiresan Subramanian
Conference Room C

Thursday, November 6, 12:00 - 13:00

CONF: Lunchgo to top

Dining Room - Lobby Level

Thursday, November 6, 13:15 - 14:15

531 (CMG-T): CMG-T: Mainframe - Part 1 - What's Happening to the Mainframe? Cloud? Mobile? Social? Big Data?go to top

CMG-T
Glenn Anderson
Room: Salon 4

If you have been too busy applying PTFs and updating PARMLIB lately to notice where our beloved mainframe is headed, then this session is for you! Join long-time instructor and conference speaker Glenn Anderson and get a quick fix update on how z/OS now plays in the world of mobile computing, social business, cloud computing, big data, and any other new buzzwords that have popped up in the last year. Leave this session smart enough to carry on an intelligent conversation with your peers and management on keeping System z connected and relevant in the modern world.

532 (CONF): PANEL: What Workloads Work Best with Flashgo to top

Ray Wicks, Moderator
Room: Salon 5

Panelists Include: Bruce McNutt, Michael Salsburg, Ellen Friedman, Elisabeth Stahl

533 (APM): User Perspective: Detecting Memory Leaks Versus Memory Hogs on UNIXgo to top

Dima Seliverstov, BMC Software
Room: Salon 6

534 (APM): Practical Applications of Locality Sensitive Hashing for Unstructured Datago to top

Jake Drew, SMU
Conference Room E

535 (PET): Performance Engineering of Investment Bank's Risk Compute Grid Applicationgo to top

Nilesh Gajjar, Cognizant Technology Solutions
Conference Room C

Thursday, November 6, 14:15 - 14:30

CONF: Breakgo to top

Thursday, November 6, 14:30 - 15:30

541 (CMG-T): CMG-T: Mainframe - Part 2 - Top New z/OS Performance Functions Every Sysprog Should Understandgo to top

CMG-T
Glenn Anderson
Room: Salon 4

Have you lost track of the most recent z/OS announcements that affect performance? Or have you heard about them, but never had time to dig in and understand them? Never fear! Join long-time z/OS instructor Glenn Anderson for this session and come away a little smarter and better prepared. Topics covered will include SMF logstream mode, Enterprise Data Compression (zEDC), pageable large pages, Flash Express, Warning Track, work-dependent enclaves, WLM and the DB2 Analytics Accelerator, and RMF XP.

542 (Featured Speaker): Key Performance Metrics for Your Continuous Delivery Pipelinego to top

Andreas Grabner, Compuware
Room: Salon 5

Do you consider the performance impact when adding a new single JavaScript file, a single AJAX call or a new database query to your app? Neglect able you say? I disagree - and so shall you. Andreas shows you the severe impact small changes can have on performance and scalability. Many small changes will have an even bigger impact and its therefore important to catch them early. If you are working on a delivery pipeline make sure to look into performance, scalability and architectural metrics such as # of Resources on your page, Size of Resources, Number of Requests hitting your web servers, database statements executed, log messages created. Monitoring these metrics allows you to add a new quality gate to your delivery pipeline and prevents major problems. In this talk you learn a handful of metrics to teach to your devs, test and ops folks. We learn why they are important by looking at specific examples that caused problems in real life apps and also how to monitor them.

543 (CONF): Common Metrics Don't Countgo to top

Joseph Temple, Low Country North Shore Consulting, LLC
Room: Salon 6

There has been a long and ongoing search for the a "Common Metric" for comparing the capacity of Servers. Analysts and Consultants make a great deal of money selling their version. None of them work. With luck some knowledge and experience we can modify their ratios. But if that is the case why not simply create metrics directly from specifications and modify them? Do we really need to pay someone for a metric?

544 (PET): In-Memory tablesgo to top

Larry Strickland, DataKinetics
Conference Room E

545 (APM): Using Regression to Estimate Query Resource Consumptiongo to top

Baron Schwartz, VividCortex
Conference Room C

Thursday, November 6, 15:30 - 15:45

CONF: Breakgo to top

Thursday, November 6, 15:45 - 16:45

551 (CMG-T): CMG-T: Mainframe - Part 3 - Understanding the RMF CPU Activity Reportgo to top

CMG-T
Glenn Anderson
Room: Salon 4

The z/OS RMF CPU Activity Report is essential in understanding z/OS performance. This session will discuss the layout and components of the RMF CPU Activity Report, and how to interpret the key metrics.

552 (CP): Dynamic Derivation of Analytical Performance Models in Autonomic Computing Environments [Best Paper]go to top

Mahmoud Awad, George Mason University
Room: Salon 5

553 (CP): A New Approach to Capacity Forecastinggo to top

Linda Carroll, IBM
Room: Salon 6

554 (PET): wuqiSpank.org Brings Transparency to the Subterranean World of SQL Performancego to top

Erik Ostermueller, FIS
Conference Room E

555 (CONF): Insider’s View of a Real-World Project: Innovative Strategies and Solutions for Performance Testing Composite Applicationsgo to top

Terri Calderone, Orasi Software, Inc.
Conference Room C

Composite applications have become one of the most complicated―and unavoidable―project types that testers face. With these applications needing to interface successfully with myriad third-party components and services―sometimes simultaneously―performance testing them adequately and in a sufficient number of scenarios can be a challenge. In this presentation, Terri Calderone, Principal Consultant for Orasi Software, Inc.’s Performance Testing Services, will discuss a complex composite application testing project where she was actually a member of the testing team. Her on-the-ground experience in this effort, which spanned several years in all its phases, provides valuable insight that CMG attendees can apply to their own testing projects. The presentation will cover: • Process of performance testing a complex, composite application with multiple input points to the system, and the innovative solutions and approaches that the team devised and introduced. • Challenges of actively testing 15 different applications simultaneously and how the team resolved them. • Benefits and difficulties of performing “day in the life” testing with large teams (in this case, up to 35 testers on the team at once). • Lessons learned from the effort―and from working in a continuous delivery (one build per week) environment over a long period of time.

Thursday, November 6, 16:45 - 17:00

CONF: Breakgo to top

Thursday, November 6, 17:00 - 18:00

CONF: BOFsgo to top

Room: Salon 4

CONF: BOFsgo to top

Room: Salon 5

CONF: BOFsgo to top

Room: Salon 6

CONF: BOFsgo to top

Conference Room C

Thursday, November 6, 18:30 - 19:45

CONF: Dinnergo to top

Dining Room - Lobby Level

Thursday, November 6, 20:00 - 22:30

CONF: Networking Sessiongo to top

Lounge/Lobby Bar - Lobby Level