Kansas City
Computer Measurement Group
Day 1
Focus on
Storage Management
October 8, 2003
Lunch
provided courtesy of VERITAS Software
Please RSVP using attached
RSVP form.
|
08:00 –
08:20
|
|
Continental breakfast
|
|
08:20 –
08:30
|
|
Welcome, announcements
|
|
08:30 –
09:30
|
|
Optimal Storage Architecture, Clustering,
and Design for Disaster Avoidance
Rob Peglar,
XIOtech
|
|
09:30 –
09:45
|
|
Break
|
|
09:45 –
10:45
|
|
Storage Transport Services and Business
Continuity using TCP/IP: A Case Study
Brian Heili,
CISCO
|
|
10:45 –
11:00
|
|
Break
|
|
11:00 –
12:00
|
|
Storage Management: The Focus for Storage
Professionals
Randy Kerns, Evaluator Group
|
|
12:00 –
1:15
|
|
Lunch
Presentation
by VERITAS Software:
Application Performance Management
|
|
1:15 –
1:45
|
|
Networking Session
Please take time
to network with vendors and other professionals.
|
|
1:45 –
2:45
|
|
Storage Resource Management –
Architecting a Storage Utility for Maximum Efficiency
Rick Harkins, VERITAS Software
|
|
2:45 –
3:00
|
|
Break
|
|
3:00 –
4:00
|
|
An Introduction to SAN Capacity Planning
Mark
Friedman, Demand
Technology Software
|
|
4:00 –
5:00
|
|
Midrange Server Resource
Mining
Rick Keck and Tom Hatcher,
Sprint
|
Kansas City Computer Measurement Group
Day 2
Focus on
Internet Performance
October 9, 2003
Lunch
provided courtesy of Wily Technology
Please RSVP using attached RSVP form.
|
08:00 –
08:20
|
|
Continental breakfast
|
|
08:20 –
08:30
|
|
Welcome, announcements
|
|
08:30 –
09:30
|
|
Men Are From Mars, Women Are From Venus,
Web Services Are From Betelgeuse
Denise P. Kalm,
BMC
|
|
09:30 –
09:45
|
|
Break
|
|
09:45 –
10:45
|
|
Virtual memory constraints in 32-bit
Windows
Mark
Friedman, Demand
Technology Software
|
|
10:45 –
11:00
|
|
Break
|
|
11:00 –
12:00
|
|
Managing J2EE Performance for the Production
Environment
Chris Farrell, Wily Technology,
Inc.
|
|
12:00 –
1:15
|
|
Lunch
Presentation by Wily Technology:
Managing High Performance Web
Applications
|
|
1:15 –
1:45
|
|
Networking Session
Please take time
to network with vendors and other professionals.
|
|
1:45 –
2:45
|
|
Zen and the Art of Application
Performance Management
Craig Hodgins,
Compuware
|
|
2:45 –
3:00
|
|
Break
|
|
3:00 –
4:00
|
|
Factors Affecting Web
Performance
Tim Goeke,
Global NetWatch
|
Abstracts – Day 1 – October 8
Optimal Storage Architecture,
Clustering, and Design
for Disaster Avoidance
Rob Peglar
Corporate Architect
XIOtech Corporation
This presentation will discuss optimal
clustered data storage architectures and illustrate solutions for disaster
avoidance, tolerance and recovery.
Strategies for avoidance and recovery are outlined. The impact of clustered storage on such
strategies is discussed, as well as data backup and restore. Replication of data is detailed and
elaborated upon as an optimal strategy.
Application best practice, as well as server architecture, coupled with
optimal storage strategy is outlined.
Overall architectural (server, network, storage) guidelines and best
practices are compared, along with the implications for enterprise and SMB
institutions, their staffs, and the financial/budgetary considerations when
using dynamic clustered networked storage architectures. A case study is given along with computation
center diagrams and infrastructure.
Biography
Rob Peglar is the
Corporate Architect for XIOtech Corporation. A 26-year industry veteran, he has global
corporate responsibility for healthcare technology, storage architecture,
strategic direction, and industry liaison for XIOtech.
He has served as a consultant to the Technical Council of the Storage
Networking Industry Association (SNIA), is XIOtech’s
principal member of the SNIA, the IP Storage Forum and the Shared Solutions
Forum, and is a former member of ANSI X3T9 and the IETF. He has extensive experience in the
architecture, design and implementation of large heterogeneous storage area
networks (SANs), virtual storage architectures, and is a frequent speaker and
panelist at leading healthcare, storage and networking industry-related
seminars and conferences worldwide. He
has led XIOtech in architecture to its current
industry-leading position in virtualized networked storage. He has a unique ability to synthesize the
business goals of healthcare, government, and corporate IT with strategic
storage architectures and technologies.
Prior to joining XIOtech
in August 2000, Mr. Peglar served as Senior
Consulting Storage Specialist for StorageTek for nine years, holding key field
and engineering management positions. Prior to StorageTek, he held engineering
and product management positions at McDonnell Douglas, Control Data Corporation
and its supercomputer subsidiary, ETA Systems, for fourteen years. His research
background includes I/O performance analysis, queueing
theory, parallel systems architecture and OS design, and virtual systems
optimization.
Mr. Peglar holds the
B.S. degree in Computer Science from Washington University, St. Louis Missouri, and performed
graduate work at Washington University’s Sever
Institute of Engineering.
Storage Transport Services
and Business Continuity using TCP/IP:
A Case Study
Brian Heili
Systems Engineer
Cisco Systems
Data distribution, data protection,
business continuance and disaster recovery strategies are critical components
of today's information-centric businesses.
The ability to efficiently replicate critical data on a global scale not
only ensures a higher level of protection for valuable corporate information,
but also promotes increased utilization of backup resources, reduces the impact
of a catastrophic failure at a single site, and lowers the total cost of
storage ownership. Using the IP network
in conjunction with FCIP holds out the promise of a far less expensive method
of reliably and accurately transferring large amounts of data over significant
distances.
Biography
As a Systems Engineer for Cisco Systems, Brian
is responsible for the development and integration of customer solutions for
storage networking and IP convergence.
With over 12 years of industry experience, he is also focused on the
design and deployment of large scale enterprise networks.
Storage Management: The
Focus for Storage Professionals
Randy Kerns
Senior Partner
Evaluator Group Inc.
Storage Area Networks have become more common
and the focus has now moved to storage management. Storage Management is being intensely
marketed and there are many companies offering products. The problem is, it is unclear how the
different storage management software products work together
or at least complement one another to help solve problems and allow
storage administration to be more effective.
There are many elements involved in storage management and how well they
integrate dictates the real value to a storage administrator. There is currently a great degree of
variability with solutions in the market today and many overlapping functions. Putting the storage management elements and
the types of Storage Resource Management into context is necessary to be able
to sort through this confusing maze.
Once understood, a more coherent plan for implementing storage
management can be formulated. This
session will explain these elements and the factors involved.
Biography
Randy Kerns is a senior
partner at The Evaluator Group and is responsible for Storage Area Network and
Network Attached Storage analysis and education as well as company and product
strategies.
He has over thirty years in the computer
industry involved in the development of storage products for both mainframe and
open systems. His background is in
product design and development. Randy’s
education includes a bachelor’s degree in computer science from the University of Missouri at Rolla and a
master’s degree in computer engineering from the University of Colorado. He has worked for IBM, Fujitsu, as Vice
President of Engineering at the Array Technology subsidiary of Tandem Computers
and as Director of Engineering for Enterprise Disk at Storage Technology
Corporation. Product development that
Randy has been involved in includes both disk and tape subsystems for those
companies.
Randy has made numerous presentations at
conferences and is the author of many industry articles and white papers.
An Introduction to SAN
Capacity Planning
Mark Friedman
VP and General Manager
Demand Technology Software
Emerging technology that allows the
construction of high performance storage area networks (SANs) requires
extending existing analytic frameworks that can accurately predict the
performance of complex disk subsystems. This paper focuses specifically on two
elements of emerging SAN technology: (1) the performance of Fibre Channel links
and connecting hubs and switches, and (2) the performance of in-band and
out-of-band SAN data management protocols. Traces and timings from benchmarking
tests conducted with an in-band SAN data manager product are analyzed and
discussed. Understanding this measurement data should enable existing disk
modeling approaches to be extended to encompass this new storage technology.
Biography
Mark is the co-author of Windows 2000
Performance Guide, published by O’Reilly in 2001. He is a
frequent contributor to CMG conferences on performance and tuning topics,
concentrating on Microsoft Windows NT since 1995. In 1996, he began developing
Performance SeNTry, performance monitoring software
for Windows NT, also known as NTSMF.
He is also currently a Vice President for
Storage Technology at Datacore Software, the parent
company that owns Demand Technology Software. He edited an industry newsletter
entitled "Mark Friedman on Storage
Management," published by Demand Technology. In 1994, he founded OnDemand Software, which developed and sold the
award-winning WinInstall software distribution
package for Windows networks. From 1987-1991 he worked at Landmark Systems
where he was the architect and the led the development team that built The
Monitor for MVS.
Storage Resource
Management – Architecting a Storage Utility for Maximum Efficiency
Rick Harkins
VERITAS Software
As we move towards utility computing, we need
to address the needs and concerns of our ‘consumers.’ When faced with the prospect of resource
control and sharing, end-users and managers alike hit the panic button. Storage service levels can be defined and
consistently delivered, while lowering hardware costs. In addition, the cost of storage can be
accurately allocated to the consumer, transforming that part of the IT
department from a cost center to a value center.
This presentation will show you what is needed
and the pitfalls to avoid on the path to a ‘storage utility.’
Biography
Rick Harkins is a Principal Systems Engineer with
VERITAS Software based in the Kansas City area. Prior to joining VERITAS Software five years
ago, Rick spent 10 years administering and managing systems at a local mutual
fund company and an Iowa
manufacturer. Rick started his career
working with the HP3000 and spent 10 years at Hewlett Packard as a Systems
Engineer, specializing in performance analysis and tuning of the MPE operating
system.
Midrange Server Resource
Mining
Rick Keck, Capacity Management Analyst
Tom Hatcher, Project Engineer
Sprint
Significant savings can be realized by mining
capacity from servers that currently reside in your environment. These servers, which range in size from very
small to very large, tend to be under utilized, very scalable, and good
platforms for new applications. The
fundamentals of mining capacity and methodologies to affect savings are
discussed. Server management
methodologies include active and passive consolidation, low utilization
elimination (right sizing), large asset monitoring and category reduction. The methodologies discussed in this
presentation apply primarily to UNIX based servers.
This presentation will focus on these two
important questions. First, “What tools
can be used to mine midrange server capacity?” Second, “What methodologies
exist for mining midrange server capacity?”
Biography – Tom Hatcher
Tom has twenty-five years of experience in
Information Technology, with the first seventeen years working with mainframes,
and the last eight years in the Open Systems environment, specifically capacity
planning. The focus of Tom’s last four
years has been on server consolidation and capacity mining. He was instrumental in developing Sprint’s
server consolidation process, which has saved approximately $35M since
1999. Tom organized and is currently
leading Sprint’s Server Consolidation Steering Committee, which is responsible
for establishing and maintaining Sprint’s server consolidation policies and
procedures.
Biography – Rick Keck
Rick has 25 years of experience in Information
Technology. He has a BS degree in
Computer Science and a Masters degree in Business Administration. His experience includes project management,
software design & development, web site administration, and IT
consulting. Currently he is a member of
the Capacity Management group developing models to make midrange server
application forecasts.
Abstracts – Day 2 – October 9
Men Are From Mars, Women
Are From Venus, Web Services Are From Betelgeuse
“The only way we will
really know the condition of Betelgeuse is to wait a few million years.”
Denise P. Kalm
BMC
Having trouble making robust legacy applications talk to hot new Web
applications? Trying to manage business
transactions with your partners when everyone is running their proprietary
software on different platforms? Simple,
flexible interoperability is the “holy grail” and happens to be analogous to
the challenges we all face when we try to communicate with the opposite
sex. Using the Mars-Venus theme, we will
explain the basics, value and challenge of creating or migrating to a Web
services standard-based Service-Oriented Architecture (SOA). We will extend that theme to communication
with people from other cultures/languages as an analogy to cross enterprises
B2B Web Services.
Biography
Denise Kalm has more
than 20 years experience in IT, most of that spent in the performance
management/capacity planning arena.
Starting out on Tandems, she branched out to CICS, MVS, UNIX and began looking at network performance for a large,
global bank before joining BMC Software, Inc as a software consultant. Her current role is in technical marketing
for performance products. Her last
position at the bank was as systems architect and relationship manager to
Credit Card Systems. Prior to entering
the IT profession, she was a biochemical geneticist. Her hobbies include freelance writing,
Jazzercise and scuba diving. Her first
novel, Lifestorm, was published last year and she is hard at
work on the next one.
Virtual memory constraints
in 32-bit Windows
Mark Friedman
VP and General Manager
Demand Technology Software
Many server workloads can exhaust the 32-bit
virtual address space in Windows Server 2003. Machines configured with 2 GB or
more of RAM installed are particularly vulnerable to
this condition. This paper discusses the signs that indicate a machine is
suffering from a virtual memory constraint. It also discusses options to keep
this from happening, including (1) changing the way 32-bit virtual address
spaces are partitioned into private and shared ranges, (2) settings that govern
the size of system memory pools, (3) hardware that supports 36-bit addressing.
Ultimately, running Windows on 64-bit processors is the safest and surest way
to relieve the virtual memory constraints associated with 32-bit Windows.
Biography
Mark is the co-author of Windows 2000
Performance Guide, published by O’Reilly in 2001. He is a
frequent contributor to CMG conferences on performance and tuning topics,
concentrating on Microsoft Windows NT since 1995. In 1996, he began developing
Performance SeNTry, performance monitoring software
for Windows NT, also known as NTSMF.
He is also currently a Vice President for
Storage Technology at Datacore Software, the parent
company that owns Demand Technology Software. He edited an industry newsletter
entitled "Mark Friedman on Storage
Management," published by Demand Technology. In 1994, he founded OnDemand Software, which developed and sold the
award-winning WinInstall software distribution
package for Windows networks. From 1987-1991 he worked at Landmark Systems
where he was the architect and the led the development team that built The Monitor
for MVS.
Managing J2EE Performance
for the Production Environment
Chris Farrell
Director of Technical
Marketing
Wily Technology,
Inc.
Complex Enterprise Java systems can be very
difficult to manage. When performance problems occur, any number of system
resources or components could be to blame: Servlets, EJBs, CPU overhead, memory utilization, application server
connection pools, the database connection or connections with any other
supporting systems. If not properly and quickly diagnosed, poor application
performance could negatively affect the organization's ability to serve its
customers and seriously impact its revenue stream. This presentation will
discuss a number of typical Java application performance problems and
approaches to identifying and eliminating them. You will learn that performance
problems can be address beyond a reactionary point of view as well as
preventing any future performance failures all together. This presentation will discuss the following
topics:
·
Challenges in the Java environment
·
Performance issue in deployment and production
·
Scenarios samples and best practices techniques
to manage performance
Biography
Chris Farrell has over 13 years experience in
the computing industry, including three years at Wily. Currently the Director of Technical Marketing,
Chris was previously Wily's Director of Product
Management, coordinating the development and initial launch of Introscope, Wily's flagship
product. Prior to Wily, Chris was the
Senior Product Marketing Manager for Ericsson mobile phones, directing Product
Marketing for the Ericsson Open professional tennis tournament. Chris spent ten years at IBM, most recently
as the Product Manager for the Wireless Web Software Suite and Segment Manager
for IBM Global Services Field Service Solutions. Chris' technical background includes BIOS
development in the IBM PC Company as the lead software engineer on several
ThinkPad systems and the first Intellistation. Chris
has a BSEE and MBA from Duke University.
Zen and the Art of Application
Performance Management
Craig Hodgins
Systems Engineer
Compuware
Application performance management is just as
important, if not more so, than systems performance management. APM is a
distinct and vital area of concern to IT organizations. This paper will present
a roadmap to help such organizations move through an application performance
maturity model that moves from chaos to kaizen, providing hints and tips at
each stage of progression. And maybe a little Zen, too….
Biography
Craig Hodgins has
been employed in the IT industry for over 22 years, 18.5 years with IBM Canada
and 3.5 years with Compuware. He has worked in various capacities, but
believes the most fun and interesting work is found in performance management. Craig is currently a Systems Engineer for the
STROBE product. Craig spoke at CMG 2002
in Reno, and has had two
articles published in Enterprise Systems Journal. Craig lives in Ontario, Canada with his wife
and three children.
Factors Affecting Web
Performance
Tim Goeke
Chief Technology Officer
Global NetWatch, Inc (GNW)
Web performance is affected by many
factors. How do the combinations of
browser, web server, and online application software combine to produce good,
or bad, performance? This talk will
focus on different aspects of high performance web sites, and different
approaches to achieving maximum performance and reliability when designing web
infrastructures and hardware.
Biography
Tim Goeke is Chief
Technology Officer at Global NetWatch. Prior to this position with GNW, Mr. Goeke was with US WEST as a Software Engineer and Prism
Resources as a Senior Software Engineer.
In both of these roles, Tim designed, developed and implemented various
Internet, Intranet, client/server and Windows based systems applications. While with Prism, he was an integral part of
the team that developed an online stock trading system, patent-pending Internet
Subscription Access system. Eventually,
the entire online trading system became the cornerstone of Ameritrade, Inc., a
leading online trading company. Additionally, Mr. Goeke
has held various designer, programmer and technician positions within the
computer industry. Tim holds a Bachelor
of Science Degree in Mechanical Engineering with minors in Computer Science,
Physics, and Math from University of Nebraska-Lincoln.