Big data analytics infrastructure for dummies

by Barry Schoenborn
BigDataAnalytics
Infrastructure
IBM Limited Edition

Big Data Analytics Infrastructure For Dummies,®
IBM Limited Edition
Published by
John Wiley & Sons, Inc.
111 River St.
Hoboken, NJ 07030-5774
www.wiley.com
Copyright © 2014 by John Wiley & Sons, Inc.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form
or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as
permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without the prior writ-
ten permission of the Publisher. Requests to the Publisher for permission should be addressed to the
Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011,
fax (201) 748-6008, or online at http://www.wiley.com/go/permissions.
Trademarks: Wiley, For Dummies, the Dummies Man logo, The Dummies Way, Dummies.com,
Making Everything Easier, and related trade dress are trademarks or registered trademarks of John
Wiley & Sons, Inc. and/or its affiliates in the United States and other countries, and may not be used
without written permission. IBM and the IBM logo are registered trademarks of International
Business Machines Corporation. All other trademarks are the property of their respective owners.
John Wiley & Sons, Inc., is not associated with any product or vendor mentioned in this book.
LIMIT OF LIABILITY/DISCLAIMER OF WARRANTY: THE PUBLISHER AND THE AUTHOR MAKE
NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETE-
NESS OF THE CONTENTS OF THIS WORK AND SPECIFICALLY DISCLAIM ALL WARRANTIES,
INCLUDING WITHOUT LIMITATION WARRANTIES OF FITNESS FOR A PARTICULAR PURPOSE.
NO WARRANTY MAY BE CREATED OR EXTENDED BY SALES OR PROMOTIONAL MATERIALS.
THE ADVICE AND STRATEGIES CONTAINED HEREIN MAY NOT BE SUITABLE FOR EVERY SITU-
ATION. THIS WORK IS SOLD WITH THE UNDERSTANDING THAT THE PUBLISHER IS NOT
ENGAGED IN RENDERING LEGAL, ACCOUNTING, OR OTHER PROFESSIONAL SERVICES. IF PRO-
FESSIONAL ASSISTANCE IS REQUIRED, THE SERVICES OF A COMPETENT PROFESSIONAL
PERSON SHOULD BE SOUGHT. NEITHER THE PUBLISHER NOR THE AUTHOR SHALL BE LIABLE
FOR DAMAGES ARISING HEREFROM. THE FACT THAT AN ORGANIZATION OR WEBSITE IS
REFERRED TO IN THIS WORK AS A CITATION AND/OR A POTENTIAL SOURCE OF FURTHER
INFORMATION DOES NOT MEAN THAT THE AUTHOR OR THE PUBLISHER ENDORSES THE
INFORMATION THE ORGANIZATION OR WEBSITE MAY PROVIDE OR RECOMMENDATIONS IT
MAY MAKE. FURTHER, READERS SHOULD BE AWARE THAT INTERNET WEBSITES LISTED IN
THIS WORK MAY HAVE CHANGED OR DISAPPEARED BETWEEN WHEN THIS WORK WAS WRIT-
TEN AND WHEN IT IS READ.
For general information on our other products and services, or how to create a custom For Dummies
book for your business or organization, please contact our Business Development Department in the
U.S. at 877-409-4177, contact info@dummies.biz, or visit www.wiley.com/go/custompub.
For information about licensing the For Dummies brand for products or services, contact
BrandedRights&Licenses@Wiley.com.
ISBN: 978-1-118-92136-4 (pbk); ISBN: 978-1-118-92311-5 (ebk)
Manufactured in the United States of America
10 9 8 7 6 5 4 3 2
Publisher’s Acknowledgments
We’re proud of this book and of the people who worked on it. For details on how to create a
custom For Dummies book for your business or organization, contact info@dummies.biz
or visit www.wiley.com/go/custompub. For details on licensing the For Dummies brand
for products or services, contact BrandedRights&Licenses@Wiley.com.
Some of the people who helped bring this book to market include the following:
Project Editor: Carrie A. Johnson
Acquisitions Editor: Connie Santisteban
Editorial Manager: Rev Mengle
Business Development Representative:
Sue Blessing
Custom Publishing Project Specialist:
Michael Sullivan
Production Coordinator: Melissa Cossell
Special Help: Rick Perret and
Herbert Schultz

Table of Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
About This Book......................................................................... 2
Icons Used in This Book............................................................. 2
Beyond the Book......................................................................... 3
Chapter 1: Getting to Know Big Data and Analytics. . . . 5
Looking at Big Data and Analytics............................................ 6
Volume............................................................................... 6
Velocity.............................................................................. 7
Variety................................................................................ 7
Examining Types of Analytics................................................... 8
Understanding the Business Need for BD&A.......................... 8
Competitive advantage.................................................... 9
Enhanced ROI.................................................................. 10
Improved customer experience.................................... 10
Introducing BD&A Infrastructure............................................ 10
Promoting success......................................................... 11
An approach you can follow......................................... 12
Chapter 2: Looking at Infrastructure for Big Data
and Analy tics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Looking at a BD&A Hardware Infrastructure........................ 15
Customer business goals............................................... 16
Needs analysis................................................................ 16
Introducing the Infrastructure Components of BD&A......... 16
Scalability of infrastructure........................................... 17
Parallel processing......................................................... 17
Low latency resources................................................... 18
Data optimization........................................................... 18
Understanding the Aspects of a Well-Architected
Infrastructure........................................................................ 20
Speed/performance........................................................ 20
Availability....................................................................... 20
Access.............................................................................. 21
Chapter 3: Looking at IBM Infrastructure Choices. . . . 23
IBM Servers................................................................................ 23
IBM Power Systems........................................................ 24
IBM System z................................................................... 25
IBM PureSystems............................................................ 25

Big Data Analytics Infrastructure For Dummiesiv
IBM Storage................................................................................ 26
Looking at storage basics.............................................. 26
IBM FlashSystem............................................................. 27
IBM System Software................................................................ 28
IBM Elastic Storage......................................................... 29
Platform Computing....................................................... 29
Chapter 4: Solving Big Problems with Powerful
Solutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Energy and Utilities.................................................................. 31
Business Need................................................................. 32
Solution............................................................................ 32
Fashion....................................................................................... 32
Travel......................................................................................... 33
Finding the solution........................................................ 34
Looking at the benefits.................................................. 34
Responding to customer behavior............................... 34
Healthcare.................................................................................. 35
Insurance................................................................................... 35
Overview.......................................................................... 36
Solution............................................................................ 36
Travel and Transportation...................................................... 36
Business need................................................................. 37
Solution............................................................................ 37
Chapter 5: Ten Questions LoB and IT Execs
Should Ask Each Other . . . . . . . . . . . . . . . . . . . . . . . . . 39
IT to LoB: Are You Aware of the Benefits of BD&A
Infrastructure?....................................................................... 40
IT to LoB: Are You Aware of BD&A’s Return On
Investment?............................................................................ 40
LoB to IT: How Can I Benefit from Bringing the
Back Office to the Front Office?........................................... 41
LoB to IT: Do I Need to Be On the Cloud?.............................. 41
LoB to IT: Should I Worry about Obsolescence?.................. 41
IT to LoB: Where and When Would Performance
Issues Occur?......................................................................... 42
IT to LoB: Are There QoS or SLA Concerns?......................... 42
IT to LoB: What Security Requirements Exist?..................... 42
IT to LOB: Does the BD&A Application Require
Mission-Critical Resiliency?................................................. 42
IT to LoB: How Will the BD&A Application Expand?............ 43

Introduction
The current form of analytics frequently involves an end-
user using a spreadsheet application. The trouble is that
information is often slow coming in and may not contain all
relevant data, and the result is unacceptable delays in getting
actionable information. That’s okay, but it probably won’t get
you where you want to go.
With Big Data and Analytics (BD&A), you get a new approach
to gathering, processing, and understanding information. You
may hear the term a lot, but many people still don’t fully know
what it is. Very broadly, BD&A is a combined hardware/software
architecture that gathers vast quantities of disparate data for
fast analysis. The analysis produces meaningful information. The
result is that you can make faster and better-informed business
decisions. The benefit is competitive advantage and (hopefully)
increased profitability for your operation.
And don’t forget infrastructure. There can be no doubt that
infrastructure matters. Take, for example, the many structural
failures throughout history: The Tacoma Narrows Bridge
(“Galloping Gertie”) in 1940 had poor girder design that
caused the bridge to buckle and fail under moderate winds.
Then there is the Leaning Tower of Pisa. Its builders failed to
address infrastructure, and that’s why the tower leans.
Fortunately, in the world of information technology, it’s
relatively easy to plan and implement appropriate infrastruc-
ture for BD&A. BD&A solutions produce positive, valuable
outcomes. They need to address three things: securing access
to all relevant data to get the best insights; making analytics
a part of everyday operations and business transactions;
and ensuring all employees can benefit from these insights
generated.

Big Data Analytics Infrastructure For Dummies2
About This Book
BD&A has several components: hardware, software, and ser-
vices. In this book, I emphasize hardware infrastructure —
processing, storage, systems software, and internal networks.
The reason for talking about infrastructure is that it’s the foun-
dation of a healthy BD&A solution. You can’t build anything
worthwhile without a good infrastructure.
If you’re reading this book, I assume that you’ve heard about
BD&A but want to understand a little more about what it is,
how it works, and how your company can benefit from it.
This book is meant for executives — decision makers — and
senior IT practitioners who want to put the concepts in front
of them.
The book talks about companies of different sizes in differ-
ent industries. Almost any business can benefit from a BD&A
implementation. Some people think that BD&A is only for
bigger companies, but smaller companies become big com-
panies by finding and exploiting the kind of competitive edge
that BD&A can provide.
Icons Used in This Book
You’ll find several icons in the margins of this book. Here’s
what they mean.
A tip is a suggestion or a recommendation. It usually points
out a quick and easy way to get things done or provides a
handy piece of extra information.
A warning alerts you to conditions that require extra care and
thinking. For example, you don’t want to omit critical steps in
evaluating your needs and planning your implementation.
Anything that has a Remember icon is something that you
want to keep in mind.
Technical Stuff contains information that’s interesting and
useful but not vital to understanding the topic. Info here may
include a brief history of a principle, the earliest practitioners,
or the origin of a word. It also showcases technical points.
You can either read these or skip over them.

Introduction 3
Beyond the Book
You can find additional information about BD&A (and about
IBM’s approach to it) by visiting the following websites:
✓ General information: www.ibm.com/big-data
✓ IBM BD&A Infrastructure: www.ibm.com/
big-data-infrastructure

Chapter 1
GettingtoKnowBig
DataandAnalytics
In This Chapter
▶ Getting a perspective on Big Data and Analytics
▶ Seeing the business need for Big Data and Analytics
▶ Introducing Big Data and Analytics infrastructure
Whether you’re a Line of Business (LoB) executive or a
senior IT manager, you’ve likely been exposed to a
great number of “new” new things. Analytics isn’t new; analy-
sis has been a component of planning for millennia. What has
always hampered civilization’s enterprises was insufficient
data and slow speed of getting it — but the need for intelli-
gence and planning goes back a long way.
What is relatively new is Big Data. It’s not an exaggeration
to say that data is now one of the world’s most valuable
resources. And there’s a lot of it. For example, consider the
Internet. During the Christmas 2013 shopping season, Amazon
sold 426 items per second — each transaction yielding a wealth
of data. But big companies like Amazon aren’t the only ones
getting benefits from Big Data and Analytics (BD&A). Smaller
companies without a big Internet presence have plenty of
useful data to analyze, too. There’s gold in BD&A if you’re will-
ing and able to mine for it. Check out Chapter 4 for examples
of smaller companies making a big impact.

Looking at Big Data
and Analytics
Executives often say, “If only I had known this (substitute
your own situation) and if only I had known it sooner, I would
have made a smarter decision.” That’s the problem BD&A
addresses. It’s the mechanism that lets you know more, and
sooner, about a lot of data. And, as a bonus, you can get infor-
mation you need from disparate and unstructured data.
Remember the three Vs of BD&A: volume, velocity, and
variety.
Volume
The first attribute of Big Data is volume. Big Data projects tend
to imply terabytes to petabytes of information. However, some
smaller industries and organizations are likely to deal with
mere gigabytes or terabytes of data.
Nevertheless, smaller organizations may still require intense
and complex analytics, based on less data. For example, a
manufacturer may have only a few retail outlets, but getting
and storing reliable sales data still is important. The same
manufacturer may have a simple supply chain, again requiring
acquisition and storage of reliable supplier data. Plus, data
is stored plenty of other places that don’t include your tra-
ditional computing system or your hard. Together this data
helps to paint a picture of what’s happening in the organiza-
tion, and what may happen in the future.
Terabytes and petabytes, frankly, aren’t a big deal anymore. At
the low end, a 1 terabyte external disk drive fits in a shirt pocket.
A 4-terabyte external drive has dimensions of 5.5” x 1.9” x 6.7”.
At the high end of the spectrum, ebay.com stores almost 90 pet-
abytes of data, which requires a good deal of storage hardware.

Chapter 1: Getting to Know Big Data and Analytics 7
Velocity
The second attribute of Big Data is velocity — the speed at
which information arrives, is analyzed, and is delivered. The
velocity of data moving through the systems of an organi-
zation varies from batch integration and loading of data at
predetermined intervals to real-time streaming of data. The
former can be seen in traditional data warehousing. The latter
is in the world of technologies such as complex event pro-
cessing (CEP), rules engines, text analytics, inferencing, and
machine learning.
Real-time or right-time isn’t just for air traffic controllers. It’s
about gathering data and insight as you run your business.
The moment can happen when a customer calls and you cap-
ture notes from the conversation. It can be click streams from
your website, hourly sales data, or anything that relates to the
speed at which you operate your business.
Variety
The third attribute of Big Data is variety. In the past, enterprises
had only to deal with a manageable number of data sources.
Times have changed. Today’s business environment includes
not only more data but also more types of data than ever before.
Disparate data is data from a variety of data sources and in a vari-
ety of formats, and is a major challenge that business analytics
and Big Data projects must contend with.
You can realize two kinds of cost-saving efficiency
enhancements:
✓ Operational efficiency: These gains are measured by
reduced costs to get the same or better results. With
BD&A, this is due to more efficient methods of data inte-
gration, management, analysis, and delivery.
✓ Business processes: These enhancements are measured
by the impact of new, better ways of conducting business,
including improvements to commercial transactions, sus-
tainable management of communities, and appropriate
distribution of social, healthcare, and educational services.

Examining Types of Analytics
After you get the data, you need to perform analytics to
extract its value. There are four principal types of analytics:
✓ Descriptive: Uses basic statistics or cool visualization
to characterize a set of data. The results may show aver-
ages, totals, frequencies, and perhaps a causal relation-
ship. The vast majority of analytics done today falls in
this category.
✓ Predictive: Helps you see what the future may hold. You
use statistical models to forecast a condition such as
revenue, profits, market-share, or operational outcome.
Predictive analytics is based on a modeled relationship
between a set of independent variables. Companies
deploy this type of analytics in different ways, but it’s
most often used for planning.
✓ Prescriptive: Takes predictive analytics (see the previ-
ous bullet item) to a new level by optimizing the best
outcome of a future prediction. It takes into account new
inputs or constraints that are specific to a given situ-
ation. It may include the next best action of changing
your credit or pricing policies or upselling/cross-selling
strategies.
✓ Cognitive: Uses techniques and a high-performance
infrastructure to extract relationship from diverse sets
of data.
Understanding the Business
Need for BD&A
An excellent technique for beating the competition and pros-
pering is reducing the time gap between awareness of a condi-
tion (the trigger event) and action on it. This is called time to
insight. Table 1-1 is a simple comparison of traditional time to
insight versus BD&A time to insight.
What remains after review of the information is making a deci-
sion and taking action.

Table 1-1 Traditional versus BD&A Time to Insight
Timeline Days (without BD&A) Days (with BD&A)
Event 0 0
Time to get data 14 1
Time to analyze 7 1
Time to insight 21 2
For convenience, the times in the table are expressed in days.
IBM can cite three much faster real-world examples:
✓ A global consumer packaged goods company reduced
real-time reporting from 6 minutes to 736 microseconds.
✓ A manufacturing company reduced customer reporting
from days to seconds.
✓ A financial company reduced cross-sell model calcula-
tions from 45 minutes to 5 seconds.
BD&A software, optimized on a well-tuned BD&A infrastructure,
provides three results of improved time to insight: competi-
tive advantage, enhanced return on investment (ROI), and
improved customer experience.
Competitive advantage
Competitive advantage occurs when an enterprise acquires or
develops attributes or processes that allow it to outperform its
competitors. It’s easy to define, but more difficult to get. Different
executives have their favorite items that give competitive advan-
tage. Broadly, they include one or more of the following:
✓ Acquire, grow, and retain customers
✓ Create new business models
✓ Transform financial processes
✓ Optimize operations and reduce fraud
✓ Improve IT
✓ Manage risk

Enhanced ROI
A universal business goal is increasing ROI — also known as
“getting more bang for the buck.” There are many ways to
measure ROI, including
✓ More profit through increased sales
✓ More profit through stable sales with less excess
inventory
✓ More profit derived from reduced operating costs
BD&A helps enhance ROI by putting more well-analyzed data
in the hands of decision makers at a fast rate.
Improved customer experience
With products becoming commoditized, price differentia-
tion is no longer sustainable for competitive advantage. Look
to improved customer experience to get an edge. Customer
experience is the sum of all experiences a customer has with
a supplier of goods and/or services, over the duration of their
relationship. You could also call this customer satisfaction,
and to a large degree you can manage and improve it.
In the past, brand loyalty could get you through, but that’s not
true anymore. That’s because your customer base’s choices
have widened, largely through the Internet.
You can use BD&A to learn what customers want and (therefore)
enhance their attaining it. You have to deliver on customer wants
and needs. Customer experience improves by your knowing what
attracts customers (products, interactions, purchasing, warran-
ties, “rewards,” and so forth).
Introducing BD&A Infrastructure
Broadly speaking, infrastructure is a combination of hardware,
software, and services. The foundation is processing and stor-
age hardware because the software has to live somewhere.
The combination ends with services, as the infrastructure
must stay robust, supportive, and scalable.

But first, here’s a difficult reality: It’s a certainty that your
current systems can’t keep pace with the growth of data and
neither can your IT budget. The volume of data grows fast, and
the parts needed to manage it grow increasingly complex.
To be a top-performing organization, you need to take a dif-
ferent approach to architecture, tools, and practices. To
outperform in your industry, try the following:
✓ Use analytics to improve the core competitiveness of
your business.
✓ Make speed a differentiator to reduce delay in decisions
and business processes.
✓ Monetize the data itself, creating new products and
services.
✓ Be more right, more often. A BD&A approach creates
more context — and more confidence.
Promoting success
The key items needed for success are managing risk, promot-
ing agility, acting strategically, and practicing forward thinking.
These are covered in this section.
How to manage risk
To manage risk, you need to
✓ Guard against poor decision making by sourcing the right
data. Source it at the right time, and integrate it.
✓ Protect against security and privacy risks, both internal
and external.
✓ Get the risk-opportunity equation right through proactive
management of your data. That includes regulatory risk
and organizational risk.
How to create IT agility
To create IT agility
✓ Build agility into your operation through optimized sys-
tems and the cloud. Doing so should relieve pressure on
your IT infrastructure.

✓ Adopt a new approach to the onslaught of data. There
are techniques for moving the analytics closer to the
data and analyzing in motion.
✓ Eliminate hidden costs by not starting piecemeal.
Instead, build from an orchestrated roadmap.
How to act strategically
Develop a curiosity-driven and evidence-inspired workforce. It’s
possible to infuse analytics into everything employees touch.
Follow these tips:
✓ Start with people. Encourage discovery and insight.
✓ Put analytics into key business processes.
✓ Deploy the full range of analytics: prescriptive, predic-
tive, and cognitive. See the section “Examining Types of
Analytics” in this chapter.
How to adopt a forward-thinking approach
Forward thinking is an important element to introduce into
corporate culture. Proactive is better than retroactive. A solid
BD&A infrastructure helps that happen. Check out these ideas:
✓ Use thinking approaches that will maximize impact while
balancing risk.
✓ Develop a shared set of trusted information to view the
business as a whole.
✓ Set the tone around privacy and policies for protecting data.
✓ Enable risk-aware decision making and model exposure
and understand variability of outcomes.
An approach you can follow
There’s a three-step procedure you can follow to get your
infrastructure in shape.
Master plan
Build against a master plan. Plan for all types of data and
all types of analytics, with an efficient IT Infrastructure as a
given. Collaborate. Cultivate new partnerships and roles. IT
and LoB executives must join to understand and develop the
master plan.

Infrastructure plan
Build an infrastructure designed to enable new levels of insights
derived from exploiting all relevant data. The platform should
be fluent in all forms of data and analytics: transactional data,
Hadoop data, and so forth.
Build a highly flexible, scalable IT infrastructure, tuned for
today’s big data environment and designed to capitalize on
the integration of social, mobile, and cloud technologies.
New approach
Take a new approach in thinking about the infrastructure for
BD&A:
✓ Enable shared, security-rich access to trustworthy infor-
mation while addressing ever-increasing data volumes,
greater variety, and increased scale.
✓ Build intelligence into operational transactions through
fast analytics and optimization of the solution stack for
different analytics workloads.
✓ Maximize right-time availability and insights to address
more users and more concurrency, changing demand and
resiliency at the point of impact.
The questions you must ask are
✓ Is my infrastructure architected for today’s challenges?
✓ Is it efficient and agile enough to support the transition
ahead?
✓ Will it meet the rising expectations of my customers,
employees, and business partners?
✓ Am I shifting my investment from maintaining my IT
infrastructure to pursuing new projects?
If you were unable to answer this last question with confi-
dence, you aren’t alone. IBM research shows that just one in
five IBM clients are able to answer this question affirmatively.
But the CIOs who run highly efficient data centers allow them
to allocate 50 percent or more of their IT budget to new proj-
ects. What could you accomplish if you could devote 50 percent
of your IT budget to innovation? What new opportunities could
you take advantage of? Don’t get mired in the status quo while
your competitor is out driving this type of innovation.

Chapter 2
LookingatInfrastructurefor
BigDataandAnaly tics
In This Chapter
▶ Taking a look at hardware infrastructure
▶ Examining infrastructure components
▶ Looking at three aspects of a well-architected infrastructure
Infrastructure for Big Data and Analytics (BD&A) is the
combination of components that produce an architecture
that delivers business outcomes. While developing this archi-
tecture isn’t rocket science, it is computer science.
You can’t build a house without a foundation, right? It’s the
same with BD&A. The finest software in the world won’t run
without optimal processing power and storage capacity as a
foundation. So in developing a BD&A solution, you start with
the foundation, which is the infrastructure. Infrastructure
must be an early part of your planning. Infrastructure matters.
Don’t leave it for last.
Looking at a BD&A Hardware
Infrastructure
A BD&A hardware architecture is largely is synonymous with
a hardware and systems software infrastructure. Granted,
the architecture is a hardware/software combination, but it’s
the hardware and systems software infrastructure that’s the
foundation. It refers to those items that deliver desired amounts
of data at desired speeds. A BD&A architecture is based on
customer business goals and mathematical analysis of needs.

Customer business goals
The executive knows best. There are often unmet needs for
fast acquisition and total storage of data that she can sense.
It’s rare to hear a customer say, “My computers run really fast,
and I have all the storage I need.”
Even if that were true, the executive must stay constantly alert
to the possibilities of increasing revenue, reducing expenses,
and competing more effectively. So chances are that the infra-
structure that meets business goals today won’t meet them
tomorrow.
Needs analysis
An old IT saying is that “it proceeds from need.” After the
customer declares her business goals, mathematical tools
address the infrastructure that meets those goals. For
example:
✓ If processing is slow, replacements or additions exist in
practically every family of computing hardware, from
blade servers to mainframes.
✓ If storage is inadequate, storage additions exist in all sizes,
from simple disk collections to massive disk arrays. Storage
capacity planning is relatively easy to do, and executives
can find options for both large one-time upgrades and grad-
ual buildouts.
✓ If moving data is slow, you have numerous network options
with predictable increases in speed (reductions in latency).
✓ If data can’t easily reach the company, new, innovative
ways (particularly mobile apps and the cloud) get the
information to the data center.
Introducing the Infrastructure
Components of BD&A
This section addresses four core infrastructure capabilities
critical for optimal BD&A infrastructure: scalability, parallel
processing, low-latency resources, and data optimization.

Chapter 2: Looking at Infrastructure for Big Data and Analytics 17
Scalability of infrastructure
Your BD&A needs will change. Presumably, the reason is that
your needs grow as your enterprise grows. To accommodate
this grown, the infrastructure must feature scalability and
resilience.
Scalability, simply put, is the ability of a system to process a
growing amount of work. In a well-planned BD&A architecture,
scalability is a planned-in component and should pose little
difficulty for an enterprise. Well-thought-out scalability makes
it relatively easy to add processor and storage power.
Hardware requirements are derived from numbers about your
current data utilization, your predicted data growth, com-
plexity of the analytics and underlying algorithms, and your
desires and decisions about business goals. Some analyses
help predict the hardware that meets the goals.
Parallel processing
There’s parallelism in processor design as in system level
software. Of course, you need an intelligently designed traffic
cop that understands the workloads so data and processing
instructions can be intelligently threaded through the hard-
ware layer — from processor to memory to storage. Greater
core and thread densities can help boost performance in data
and processing as more users begin to scale.
The same can be said for systems level software and for Big
Data in particular. File system parallelism is coming of age. Big
Data may need to be accessed in multiple ways from multiple
places and at multiple rates of speed.
Take a simple example from back in the pre-computing days.
Would you want a filing cabinet that has one long ten-foot drawer
or ten one-foot drawers to store all your information? What if five
people needed to access information at the same time?
The infrastructure design challenge is in determining where
and how to store data for the best performance as data
scales. High-performance parallel file systems spread the
data around intelligently so it can be quickly accessed and
updated at high speed.

Low latency resources
High speed (often referred to as low latency) is becoming a real
differentiator for companies that are leveraging analytics. Low
latency is pure speed — speed by design. One of the marvels of
modern computing is the tremendous increases made in speed,
which embraces CPU speeds, DRAM speeds, disk speeds,
and transmission lines (think optical fiber cable). Together,
hardware systems are orders of magnitude speedier than their
predecessors.
In computing, speed is also architected in the system software.
For example, flash memory is very fast, and the price is coming
down. It provides real benefits in accelerating BD&A. Also, new
forms of resource management software enable scale-up/scale-
out architectures to perform at lightning speed. Speed can also
be delivered when you optimize the hardware and software
together, either at the middleware or application level.
You can be assured that infrastructure latency is capable of
meeting your business goals, both simple and elaborate — if
you have the right hardware infrastructure in place.
Data optimization
In the world of BD&A, data management takes on a whole new
meaning. What data do I keep? Do I need to store it in a differ-
ent location? What are the costs of dealing with the data? How
quickly and when do I need it? All good questions that require
new capabilities.
Developing the right data architecture enables data indepen-
dence and provides a simple, resilient, and agile environment
to support BD&A.
Big data isn’t just about “social media” Hadoop data. It’s about
all your data and how internal and external data sources
provide you with the best insights. Consider the following:
✓ More data = better context = better confidence
✓ Not all data needs to be stored on expensive media
✓ BD&A is dynamic — and your storage needs have to self-
tune with your ongoing requirements

Network infrastructure
A network is a combination of points
or nodes and the lines that connect
them. In the world of IT, the definition
holds up, but has several variants.
The variations depend largely on the
perspective of the people discussing
the network — and they’re all legiti-
mate. In all cases, they contribute to
the BD&A architecture.
The variants include the following:
✓ Operating departments: In a busi-
ness or government environment,
network likely refers to the con-
nection between desktop com-
puters and processors. That why
it’s called a Local Area Network
(LAN).
✓ Data center: Network often refers
to the system of redundant full
crosspoint switches that move
data between storage and serv-
ers, and vice versa.
✓ Failover: There’s much more
to disaster recovery (DR) than
local backups or offsite back-
ups. Backups aren’t all that
helpful when a flood or earth-
quake hits. Sophisticated com-
panies have robust connections
between regional data centers.
Should disaster strike, process-
ing will “fail over” to another
data center over a Wide Area
Network.
✓ Branch offices: Many businesses
have a central data center and
branch offices, but banking surely
offersthemostdramaticexample.
Bank of America has 5,600 bank-
ing centers and 16,200 automated
teller machines. That implies an
enormous network.
✓ Mergers: Bank mergers are a
special source of concern. There
are security issues, compatibility
issues, and generally no (or few)
branch closings. IBM assisted
Bilbao Bizkaia Kutxa (BBK) when
it acquired CajaSur. Both insti-
tutions were about the same
size. That implies an enormous
network.
✓ Internet: Large Internet retailers
have robust data center net-
works, but also get information
from and send information to
many users connected over the
Internet.
Some important capabilities to consider in the design of your
data architecture include
✓ Data virtualization so you can focus less on the data and
more on the workload
✓ Automated data tiering and compression so you can
store more and use more for less money

✓ Analytics in a flash; speed up to 12 times faster and more
can be achieved by using the latest innovations in flash
storage systems
✓ Platforms that promote diversity — diversity of work-
loads and data types and data demands that can be
tuned and managed automatically
Understanding the Aspects of a
Well-Architected Infrastructure
The following sections describe the three aspects of a well-
architected infrastructure: speed, availability, and access.
Speed/performance
BD&A provides insightful answers, but you need the horse-
power to get the job done. The infrastructure must have the
power to meet current needs and grow to meet expanding
needs. An application may have a key requirement to meet
certain thresholds. For example, for a customer service
analytics application, you can’t wait two minutes to get an
answer; the answer has to be available in seconds.
Throughput often consists of a series of analytical steps output
to a real-time transaction process. Understanding how data
and insights flow and how consistently they flow at speed is an
important essential in planning.
You need to balance cost versus speed and the value of busi-
ness outcomes. In addition, also consider future growth. A well-
architected solution should achieve the best balance among all
requirements.
Availability
Resilience is the ability of a system to absorb or avoid damage
without suffering complete failure. A well-planned BD&A archi-
tecture allows for processor, storage, and facility failures (with
protection via “failover” mechanisms), software failures, and at
the least a strong backup and restore capability.

Ensuring availability begins with evaluating from where the
data comes and to where the insights flow. If insights are
important to your real-time operations and processes, resil-
ience (and disaster recovery) is vital. Solutions may range
from simple rebuilds of a failed disk to simultaneous asyn-
chronous data replication at a remote site.
Evaluation of insights includes what ones are mission criti-
cal, what ones are critical, and what ones are non-critical.
Understanding and differentiating insights is important for
putting resiliency where it’s most needed. The more support-
able and manageable a system is, the easier it is to maintain
availability.
Access
Understanding data access is the basis for three kinds of
efficiency:
✓ Data locality: It’s vital to know where the data is coming
from, where it needs to go, and whether there’s any
pre-processing required.
✓ Data life cycle management: Be aware of how often data
need to be accessed, how much data there is, and the
cost of maintaining it. There may be governance issues,
and data may need to be retained to keep models fresh.
✓ Data deduplication: There may be some excellent lever-
age in storage, achieved by eliminating the same data
stored in multiple places.

Chapter 3
LookingatIBM
InfrastructureChoices
In This Chapter
▶ Seeing a world of infrastructure options
▶ Examining the storage universe
Apowerful infrastructure goes a long way toward improving
success in implementing Big Data and Analytics (BD&A).
The alternative is failure, which usually isn’t an option for enter-
prises. Chapter 2 discusses infrastructure capabilities in a gen-
eral way, but this chapter gives you a look at the IBM families of
server, storage, and system software products.
Many products from various sources appear at first to meet
the BD&A challenge, but that’s not the whole story. You’re
better served by using products that have superior features
designed into them. So, in addition to brute force processing/
storage power, look for products with intuitive management
that require a minimum of manual intervention and result in
reduced administrator time.
IBM Servers
You may want to become familiar with key IBM product line
names. They’re the core of a BD&A infrastructure and fall into
the worlds of processing and storage.
In all cases, infrastructure needs for BD&A meet six critical
business needs. Examples are acquire, grow, and retain custom-
ers; transform financial management processes; optimize oper-
ations and reduce fraud; manage risk; improve IT economics;

and create new business models. For each entry point, certain
infrastructure design points (speed, access, availability) are
more important than others. Certain design points are better
enabled by specific hardware and software infrastructure capa-
bilities and architectures.
Some deployments are more aligned with scale-out server
solutions, while others are used with scale-up solutions:
✓ Scaling out (scaling horizontally) means adding more
nodes to a system, such as adding a new computer to a
distributed software application. Hundreds of small com-
puters may be configured in a cluster to obtain aggregate
computing power that often exceeds that of computers
based on a single traditional processor.
✓ Scaling up (scaling vertically) means adding resources
to a single node in a system. It typically involves adding
CPUs or memory to a single computer. Such vertical scal-
ing of existing systems also enables them to use virtual-
ization technology more effectively, as it provides more
resources for the hosted set of operating system and
application modules to share.
Choose a scale-in, scale-up, or scale-out server infrastructure to
support the complexity and breadth of analytic workloads. IBM
provides an integrated, high-performance infrastructure includ-
ing core server, storage, networking, and systems software tech-
nology. You can accelerate the flow of data and insights, provide
shared and highly secure access to all types of data where it
resides, and significantly improve right-time availability. The
core systems of the BD&A infrastructure are covered in this
section.
IBM Power Systems
IBM Power Systems are highly virtualized, based on IBM next-
generation POWER8 multicore processors. Designed around
the world’s first open server ecosystem, they bring together
the computing power, memory bandwidth, and I/O in ways that
are optimized for innovative data hungry applications. Power
System’s support Linux open standards, as well as the industry
leading AIX UNIX. The platform enables choice in design and
deployment to meet a wide range of data center requirements.

Chapter 3: Looking at IBM Infrastructure Choices 25
There are multiple deployment styles for business analytics,
including server nodes for business analytics grids and scal-
able server nodes to analyze data stored in data warehouses.
The proximity of data stored near (or on) these systems
enhances overall throughput, as do fast server I/O links con-
necting to external storage and network switches.
These capabilities have made Power Systems the platform
that has driven innovative BD&A solutions, such as the IBM
Watson solution that’s now commercially available for clients
looking to take their analytics to the next level by leveraging
cognitive computing.
IBM System z
IBM’s mainframe is System z. Much of the world’s transactional
and financial data still reside on mainframe systems, and the
System z BD&A solutions are centered on the concept of moving
analytics processing closer to the data, integrating and manag-
ing the full variety, velocity, and volume of data that comprises
advanced analytics business solutions. This approach relies on
the structured data already residing on System z but also incor-
porates feeds from unstructured and streaming data sources.
IBM System z mainframes deliver the ability to easily handle mas-
sive data volumes while also scaling quickly and cost-effectively
as needs arise — efficiently protecting and governing data and
delivering leading-edge computing performance.
IBM System z is highly virtualized through logical partitioning
(LPARs) and supports several operating systems, including
System z z/OS environment and Linux distributions running
on System z Integrated Facility for Linux (IFL) specialty pro-
cessors. This solution can dramatically reduce downtime (by
over 99 percent) and provides the highest level of support in
mission critical applications. The built-in support for Linux
opens System z to a wide variety of business analytics use
cases, supporting workload consolidation and linking to other
servers (for example, Power Systems blades).
IBM PureSystems
IBM PureSystems is a series of integrated systems that allow
choice and flexibility in server deployments, including both
x86 and non-x86 server nodes. The nodes are connected by

a high-speed fabric and built-in switches. These integrated
systems have built-in security features. Flexibility in server
deployments means that you can add nodes later on as
demand for computing resources grow. A single console man-
ages shared system resources.
Three major areas of focus in the IBM PureScale design are as
follows:
✓ Scale-in systems design: The system’s overall design
automates integrated server, storage, and networking
resources for both physical and virtual environments.
✓ Patterns of expertise: The systems embed software tools
that automate configuration, upgrades, and application
requirements.
✓ Cloud-ready integration: Designed to be deployed to
support private self-service cloud infrastructure and
software. That enables flexible scale-up and provisioning
of workloads.
IBM Storage
IBM has a portfolio of storage capabilities for BD&A deploy-
ments. In this section, you take a look at those and get the
lowdown on storage basics.
Looking at storage basics
IBM storage solutions are designed to save data center real
estate through the use of dense storage and power efficiency.
Essential storage technologies are built in and tightly inte-
grated. Essentially, they have the following features:
✓ Self-optimizing
✓ Cloud-agile
✓ Reduced space requirements

✓ Intuitive management
✓ No manual intervention required
✓ Reduced administrator time
IBM storage solutions use techniques such as virtualization,
deduplication, data compression, automated tiering, and cata-
loging. Take a look at the nearby sidebar “A brief glossary of
terms” for more information.
Like servers, storage can be scaled up and scaled out. Check
out the section “IBM Servers” earlier in the chapter for more
information on scaling up and scaling out.
IBM FlashSystem
Huge amounts of data are generated by sensors and comput-
ers in addition to data created by people. It’s hard to keep up
with data growth. And then, without high-performance stor-
age, it can be impossible to extract meaningful insights from
rapidly accumulating data in areas such as customer satisfac-
tion, operational efficiency, financial processes, risk, fraud,
and compliance management.
The latest innovation from a BD&A perspective is the use of
flash technology for storage. You can boost business agility
and insight with flash storage for analytics. That’s where IBM
FlashSystem comes in.
IBM FlashSystem is fast, so it unlocks storage performance
to give analytics applications the speed they need. But speed
is just one component of effective analytics applications.
Storage must also meet three key requirements in order to
effectively support analytics workloads:
✓ Compelling data economics
✓ Enterprise resiliency and
✓ Easy infrastructure integration

FlashSystem delivers advanced technologies in each of these
areas, so organizations can unleash the power of their data for
critical business insights.
IBM System Software
IBM provides two key system software offerings designed
for BD&A: IBM Elastic Storage (formerly General Parallel File
System) and Platform Computing.
A brief glossary of terms
When you talk about infrastructure,
youcanacquaintseveralusefulterms:
✓ Virtualization: Creates a virtual
(rather than actual) version of
something. This goes back to
the days of creating mainframe
partitions, so multiple jobs could
run simultaneously on the same
machine. Hardware virtualization
is creating a “virtual machine”
that acts like a real computer with
an operating system. Hardware
virtualizationisawayofimproving
efficiency.
✓ Deduplication: Eliminates dupli-
cate copies of the same data.
You get two benefits from data
deduplication: a single version of
the truth and excellent leverage
in storage.
✓ Data compression: Encodes
information using fewer bits than
theoriginalrepresentation.That’s
a space saver. Compression
helps reduce resource usage,
such as data storage space or
transmission capacity.
✓ Automated tiering: Automates
the movement of data across
different types of storage devices
and media (tiering). The most
important data goes to flash or
cacheddisks.Lessimportantdata
goes to less expensive, slower
devices,suchasopticaldisksand
tape drives.
✓ Cataloging: Comprises views
and tables that contain infor-
mation about all the objects in
the database — sometimes
referred to as metadata. As the
database changes, the catalog
changes. You can leverage this
in many ways, including report
generation, automation of
required database changes, and
checking compliance with data-
base naming conventions.

IBM Elastic Storage
IBM Elastic Storage is a high-performance, cost-effective stor-
age methodology. It started out as a clustered file system and
evolved into much more. Today it’s a full-featured set of file
management tools, including advanced storage virtualization,
integrated high availability, automated tiered storage manage-
ment, and performance to effectively manage very large quan-
tities of file-based data.
IBM Elastic Storage is designed to support a wide variety of
application workloads and has been proven extremely effec-
tive in very large, demanding environments. It is installed in
clusters supporting BD&A, gene sequencing, digital media,
and scalable file serving. All indications are that the future
of BD&A includes bringing more unstructured and file-based
data into the application.
An Elastic Storage cluster can be a single server, two servers
providing a high-availability platform supporting a database
application, or thousands of servers used for applications
such as the modeling of weather patterns, or supporting the
needs of advanced analytics leveraging file-based data. IBM
Elastic Storage was designed to support high-performance
workloads and has been proven effective for a variety of BD&A
applications.
Platform Computing
When you have big data problems to solve, you need to maxi-
mize the potential of your computing power and the support-
ing infrastructure to accelerate your applications at scale,
extract insight from your data, and make better decisions
faster. IBM Platform Computing provides that by pooling your
computing resources, managing them efficiently across mul-
tiple groups, and getting the most out of your IT investment.
For those looking to accelerate Hadoop performance, IBM
Platform Computing’s Symphony delivers peak performance
as measured by Terasort and UCB SWIM results. Symphony is
a low-latency grid management platform for SOA and distrib-
uted applications. It supports diverse compute and data inten-
sive applications.

Chapter 4
SolvingBigProblemswith
PowerfulSolutions
In This Chapter
▶ Seeing how businesses deploy infrastructure
▶ Learning about problems addressed and improved outcomes
achieved
IBM has deployed infrastructures for clients in a range of
industries around the world. In this chapter, you see sev
eral areas that are affected by infrastructure and how it drives
new revenue (or increases revenue), services customers
better, and reduces the costs of IT.
Energy and Utilities
Oncor Electric Delivery Company, LLC, is an electric distri
bution and transmission company in Texas with more than
three million customers and a service area covering nearly
117,000 square miles. Oncor began its advanced meter program
(digital smart meters) in 2008 to collect data on usage, load,
and availability. Oncor recently completed the deployment of
3.25 million meters serving the customers of north and central
Texas. Last year, customers in Oncor’s service territory showed
during the company’s biggest energy saver contest that by
using the information from Oncor’s advanced meter, they could
reduce their electric usage and bills by 25 percent or more.

Business Need
The Oncor smart meter deployment creates immense volumes
of data, including an average of 2,900 records a month for
each of its 3+ million customers. The company needed a more
efficient and cost-effective alternative to its existing storage
setup.
To accomplish its goals Oncor, partnered with IBM and IBM
Business Partner Ecologic Analytics, along with Advance
Metering System provider and IBM Business Partner Landis +
Gyr, to implement a Meter Data Management System (MDMS)
and deploy smart meters across its grid — meters that would
collect energy usage and event data automatically.
The meters provided consumers the data needed to become
active participants in their power consumption management
and enabled individual households to reduce electrical con
sumption by an estimated 5 to 10 percent.
Solution
Oncor deployed IBM Power Servers and System z with IBM
ProtecTIER.
Fashion
Bernard Chaus (Chaus) is an American clothing company
that retails consumer products. The clothing line is made to
enhance the dynamic woman’s personal style. The company
is involved with wholesale distribution and services. Chaus
found its sales and merchandising tracking efforts were
weighted toward time-consuming report creation and data
issues. That, combined with the increasingly competitive
nature of the business, prevented Chaus from effectively ana
lyzing and addressing its selling opportunities.
Chaus enlisted SKYPAD from IBM Premier Business Partner
Sky IT Group to help analyze and address selling opportuni
ties. SKYPAD is delivered via a Web portal and integrates
weekly electronic data interchange (EDI) and spreadsheet
feeds from each retailer selling various Chaus products.

Chapter 4: Solving Big Problems with Powerful Solutions 33
EDI is a particular set of standards for computer-to-computer
exchange of information.
By using SKYPAD, Chaus has the ability to track daily which
clothing items are selling (and which aren’t) at the retailer
level. This unique insight is achieved by merging information
from retail outlets around the world augmented by Chaus
Enterprise Resource Planning (ERP) data.
This solution for Bernard Chaus runs on IBM infrastructure
hosted by Sky I.T. Group. It offered Bernard Chaus the analy
tics capabilities it needed to make sense of sales data and gen
erate business insight. The hosted infrastructure behind these
IT capabilities included an IBM POWER server, which Sky I.T.
Group uses to run the SKYPAD application (using data col
lected from the ERP and retail point of sale systems) and IBM
DB2 software running on IBM BladeCenter servers to power
the SKYPAD database.
Henry Thervil, president of sales for Vince Camuto, Bernard
Chaus commented that “The sheer fact that we have IBM
hardware and software with SKYPAD has given us the analytics
capabilities to gain a big advantage over our competitors.”
Travel
Go Voyages, an online travel company in Paris, France, offers
a full range of services including accommodations, cruises,
short- and long-term vacation packages, and insurance. Go
Voyages has 450 employees with websites in English, French,
German, Italian, and Portuguese.
The company needed the ability to manage annual web-based
business growth of up to 30 percent with intensive behavioral
analytics and was looking for a solution. Part of the Go Voyages’
business model is to analyze customer behavior based on
response to specific travel offers and discounts and the
volume of data threatened to overwhelm the existing systems.
Go Voyages wanted to create a solution that would solve the
immediate challenges and scale continuously to deal with future
growth.

Finding the solution
Go Voyages chose IBM Power and System x servers for both
their front office and back office systems. The IBM solution
allows Go Voyages customers to compare prices from their
travel providers and to propose the best price to online cus
tomers. For example, during the air travel disruptions caused
by the Iceland volcanic ash clouds, Go Voyages was able to
find flights for stranded clients when many other travel com
panies could not.
Looking at the benefits
Being able to connect to different global reservation systems
means that Go Voyages can offer a far greater choice to its
consumers, and the IBM Power servers provide quick cache
services for more than two million daily unique web searches.
Even with 700 query parameters, the new IBM solution is able
to process customer requests twice as quickly as in the past.
The insight generated from behavioral analysis is used by
the Go Voyages marketing team to develop better marketing
promotions. This helps ensure that more customers complete
their ticket bookings, generating increased sales. As men
tioned, sales are rising by 30 percent annual — 25 percent
more than the industry average.
Responding to customer
behavior
Go Voyages has deployed a business intelligence solution,
built on IBM hardware and software technologies, to create
and support a comprehensive data storage environment for
the company’s travel data and payment records, and enables
customer data to be integrated and analyzed using the soft
ware’s predictive modeling capability.
For example, customers who want to purchase an airline
ticket from Paris to New York may be discouraged by the high
price. Using behavioral analytics, based on the IBM solution,
Go Voyages prompts the customer with alternative routes

with a short layover — and with substantial savings. The
analysis allows Go Voyages to tailor the response to flight
inquiries and to increase the probability of ticket sales.
Healthcare
Park Nicollet Health Services provides healthcare services
through its network of 25 clinics, six urgent care sites, and one
hospital. From its headquarters in St. Louis Park, Minnesota,
the organization provides inpatient and outpatient services,
emergency care, pharmacy services, and health and wellness
programs. Although many other healthcare organizations
operate hospitals as separate entities from physicians, which
can inhibit information sharing, Park Nicollet set out more
than two decades ago to create an integrated care network
designed to encourage collaboration.
Park Nicollet wanted to improve the quality of care for its
patients by implementing a new electronic medical record
system and new servers capable of supporting advanced ana
lytics. So the company deployed an Epic system supported by
IBM Power Systems.
The IBM solution achieved the following:
✓ Employed advanced analytics and reduced the heart-
failure mortality rate of patients by about 20 percent
✓ Reduced rack-space requirements by more than 75 per
cent compared to previous servers
✓ Reduced failover time from more than two hours to less
than two minutes
Insurance
The Swiss Re Group, headquartered in Zurich, is a leading
worldwide wholesale provider of reinsurance, insurance, and
other insurance-based forms of risk transfer. Its global client
base consists of insurance companies, mid-to-large-sized cor
porations, and public sector clients. From standard products
to tailor-made coverage across all lines of business, Swiss Re
deploys its capital strength, expertise, and innovation power
to enable the risk-taking on which enterprises and progress in
society depends.

Overview
With 2.5 billion transactions and records dating back nearly
30 years, Swiss Re needed a solution with the flexibility and
capacity to manage its growing data workloads. The company
wanted to acquire insights from multiple international locations
of claims information from numerous insurance companies.
These insights would increase the understanding of risk, the
agility in identifying profitable segments, and the ability to
make decisive decisions quickly by internal business users.
The company needed a system that would also scale to meet
future data and analytics needs.
Solution
Swiss Re chose to deploy IBM zEnterprise System with IBM
DB2 for z/OS to perform data analysis and reporting from
a central location, with IBM DB2 Analytics Accelerator
(IDAA), powered by IBM Netezza technology, to deliver faster
responses to individual analytic queries.
With the solution in place, the company benefitted from
✓ Increased satisfaction among internal business users
by speeding report generation by 70 percent through
faster query response times with IBM DB2 Analytics
Accelerator for z/OS
✓ Off-loaded workloads, reducing processing costs
✓ Transparently implemented deployment without pro
gram modification, enabling fast and cost-effective
installation
Travel and Transportation
Stuttgarter Straßenbahnen AG (SSB) provides passenger
transport services in the greater Stuttgart region of Germany.
The company is owned by the State Capital Stuttgart and acts
as the region’s main public transport operator. SSB manages
an extensive network of light rail lines, bus routes, and a rack
railway, carrying a total of more than 171 million passengers
annually. In all, the company operates approximately 450
buses and trains, and employs more than 2,900 people.

In addition to its public transport duties, SSB acts as an IT
service provider for other public transport companies and
medium-sized companies from various industries. Its IT service
business specializes in SAP ERP hosting, and the company also
provides general data center services and runs online shops,
marketplaces, and business-to-business portals for its clients.
Business need
Slow response times or lack of application availability could
have a huge impact on SSB’s ability to properly coordinate its
activity and deliver safe, efficient urban transport. It could also
cause significant disruption to its clients’ operations, leading
to lost revenues if clients chose to switch hosting providers.
The company wanted to ensure fast access to enterprise data,
in order to provide its own executives and its clients with the
information they needed to perform timely, accurate analysis
and reporting and guide better decision-making.
Solution
SSB chose IBM Power Systems servers, running IBM AIX, to
deliver the performance it needed for its business-critical sys
tems, including an extensive SAP ERP application landscape.
To manage growing quantities of data and provide fast access
to vital information, the company extended its existing stor
age environment, based on IBM System Storage SAN Volume
Controller and high-performance IBM FlashSystem technology.
The solution allowed
✓ Quicker payouts with optimized commission calculations
in SAP Revenue Management and Contract Accounting
(SAP RM-CA), reducing runtime from five hours to just
40 minutes, a time savings of more than 86 percent
✓ HR staff new opportunities to work better with data by
accelerating complex payroll processing from six hours
to six minutes, a time saving of more than 98 percent
IBM FlashSystem storage is expected to deliver a full return
on investment within five years and offers a total cost of oper
ation that is 50 percent less than traditional storage solutions.

Chapter 5
TenQuestionsLoBand
IT ExecsShouldAsk
Each Other
In This Chapter
▶ Looking at common BD&A questions
▶ Providing additional info for the LoB executive
▶ Preparing the IT manager for talking with the LoB executive
Big Data and Analytics (BD&A) offers numerous advantages
for an enterprise. However, different Line of Business (LoB)
executives have different business processes and are likely to
have different perceptions about the value of BD&A and BD&A
infrastructure:
✓ To a senior executive responsible for sales and marketing
(whether it’s the light truck division or domestic airline
ticket sales), BD&A can be a nimble trend spotter that
builds sales and enhances the client experience.
✓ To a senior executive responsible for national or world-
wide finance, BD&A provides a rapid view of what is and
what may be financially beneficial for the firm.
✓ LoB executives are cost conscious, and when there’s a pro-
posed improvement in infrastructure that costs money, the
IT executive has some explaining to do.
✓ As an IT person, you may need to do some selling to get
buy-in. The core principles for these internal sales are
the same as for client sales — emphasize benefits, not
features, and put costs in proportion.

This chapter gives you ten questions LoB and IT executives
should be asking each other (and also be prepared to answer).
IT to LoB: Are You Aware
of the Benefits of BD&A
Infrastructure?
Use case after use case shows the benefits of BD&A. Those
cases address many industries, and your operation is likely
very similar to one of them. The IT department and IBM (along
with partners, sometimes) can analyze operational problems.
There are metrics that will predict the benefits the enterprise
is likely to see. Take a look at Chapter 4 to see five use cases
from different industries.
IT to LoB: Are You Aware
of BD&A’s Return On
Investment?
In relative terms, BD&A infrastructure is a bargain. The infra
structure deployment for the company is scaled to the company.
When the company does a thorough cost/benefit analysis, it can
factor in increased productivity (and hence lower overall labor
costs), manufacturing efficiencies, marketing timing efficiencies,
and so on. It’s a game of chess, where you need to see the whole
board. For example, wouldn’t the company save money if it
made only products that customers wanted and none that they
didn’t want?
Change in IT is continuous change, and a flexible, agile infra-
structure minimizes longer-term changes and impacts. Overall,
that can improve return on investment (ROI). Besides, in abso-
lute terms, most change has produced lower costs. Consider
Moore’s Law, predicting regular increases in processing power
with lower costs. The same is true of disk storage. A terabyte
of storage was once obtained with 30 hard drive costing
$400.00 each — or $12,000.00. These days that same terabyte
costs about $230.00. For more information on enhanced ROI,
see Chapter 1.

Chapter 5: Ten Questions LoB and IT Execs Should Ask Each Other 41
LoB to IT: How Can I Benefit from
Bringing the Back Office to the
Front Office?
Back office operations have traditionally been thought of as
non-customer-facing operations. But in fact, certain operations,
such as procurement, shipping, or billing, affect customer sat-
isfaction. And increased customer satisfaction can produce a
competitive advantage. There is back office data, which, when
aggregated and available for analysis, can create useful market-
ing information for those in the front office.
LoB to IT: Do I Need to Be
On the Cloud?
The cloud has its advantages, and IBM has use cases that sup-
port that. Just the same, you can achieve business goals with a
robust BD&A infrastructure and no cloud. Chapter 2 mentions
that considering the cloud should be part of needs analysis.
Chapter 3 discusses IBM systems that are “cloud-ready.” Check
out those chapters for more information.
No matter what you do, make sure you and your IT team
understand the infrastructure implications early.
LoB to IT: Should I Worry
about Obsolescence?
Not at all. Business intelligence (BI) and predictive analysis
have always been components of business. The concept of
BD&A is the next step in logical evolution from slow, scattered
“little data” to a big picture vantage point. Particular items go
obsolete, but constant reinvention replaces them — sometimes
at a lower cost than the items they replace. That’s the way a
flexible, agile infrastructure works.

IT to LoB: Where and When Would
Performance Issues Occur?
Given the data that must be gathered and analyzed and the
speed that’s required, think about where you may have per-
formance issues. These could pertain to the amount of data
required, the number of people or processes using the outputs,
and where the data resides. For example, a performance issue
may be the need for a sales/service agent to have up-to-date
information in seconds after a customer query. Having the data
could magnify upsell/cross-sell opportunities.
IT to LoB: Are There QoS
or SLA Concerns?
An application (or the workload it must process) could run
the risk of falling short of contracted Quality of Service (QoS)
or Service Level Agreement (SLA) goals. This situation can be
prevented by deploying a robust BD&A architecture. Refer to
Chapter 2 for a brief discussion of speed/performance.
IT to LoB: What Security
Requirements Exist?
Is the data exposed to a security risk? Are the outputs or the
delivery of insights also exposed? The company usually wants
a solution that provides security for all these. Chapter 1 gives
you four key items that promote success in managing risk.
IT to L0B: Does the BD&A
Application Require Mission-
Critical Resiliency?
Consider the impact of a failed system. What would happen if
the application stopped working for ten seconds, ten minutes,
or ten hours? See Chapter 2 for more on resiliency and failover.

Chapter 5: Ten Questions LoB and IT Execs Should Ask Each Other 43
IT to LoB: How Will the BD&A
Application Expand?
A well-architected BD&A application expands gracefully to
meet future needs. Also, it should be able to leverage other
data and business processes in the future. See Chapter 3
for more info on IBM storage. “Future needs” embraces the
growth of an application to manage increased volumes and
faster speeds in a primary application. The LoB executive
should consider other company operations, probably with
new data sources and different users, that would benefit from
expansion of the BD&A application.

Big data analytics infrastructure for dummies

More Related Content

Viewers also liked (14)

Similar to Big data analytics infrastructure for dummies (17)

More from Luke Farrell (14)

Recently uploaded (20)

Big data analytics infrastructure for dummies