Cloud computing-made-easy

Cloud Computing Made Easy
by Landis & Blacharski

Copyright  2010-2013 by Virtual Global, Inc.

All rights reserved. No portion of this publication may be reproduced,
stored in a retrieval system, or transmitted in any form by any means-except
for brief quotations in printed reviews-without the prior written permission
of the publisher.

ISBN 978-1482779424

By Cary Landis and Dan Blacharski
www.saasmaker.com
www.techie.com

Table of Contents

I. INTRODUCTION ......................................................................................7
II. WHAT IS CLOUD COMPUTING? .........................................................9
What does the cloud mean to your partners, employees, and customers? . 11
What is "as a service"?................................................................................ 12
III. WHY CLOUD COMPUTING? .............................................................. 13
Cloud computing for end-users ................................................................... 14
Cloud computing for system administrators ............................................... 14
Cloud computing for software developers .................................................. 15
Cloud computing for IT buyers, corporate and federal ................................ 16
IV. CLOUD COMPUTING UP CLOSE ...................................................... 17
A. Cloud Infrastructure (Infrastructure-as-a-Service, or IaaS)........... 18
So where is all this infrastructure? ................................................................. 20
Virtualization ................................................................................................... 20
B. Cloud Platforms (Platform-as-a-Service, or PaaS) ........................ 21
Cloud Platforms as Middleware ...................................................................... 24
Type 1 PaaS: Runtime Platforms (BYOC) ......................................................... 25
Type 2 PaaS: Software Development Platforms ............................................. 26
What PaaS is not ............................................................................................. 27
C. Cloud Software (Software-as-a-Service, or SaaS) ......................... 28
V. OTHER CLOUD OFFERINGS.............................................................. 28
Cloud Brokers ............................................................................................. 29
What are "Web services"? .......................................................................... 29
Supercomputing-as-a-Service ..................................................................... 29
High Performance Computing as-a-service (HPCaaS) .................................. 30
VI. CLOUD AS-A-NECESSITY ................................................................... 31
Custom software on the cloud .................................................................... 33
Why cloud computing is already becoming mainstream ............................. 34

What does cloud computing mean to me? .................................................. 35
SOHO and small business ............................................................................ 36
VII. SECURITY CONCERNS WITH THE CLOUD ................................... 36
VII. CLOUD STRATEGY ........................................................................................ 38
A growing expectation .................................................................................... 38
The personnel side .......................................................................................... 39
VIII. Market Disruption ............................................................................... 39
VIII. MISCONCEPTIONS .............................................................................. 44
Top Ten Misconceptions about Cloud Computing ....................................... 44
1. The cloud is just a return to centralized computing. ...................... 44
2. The cloud is not secure. .................................................................. 44
3. The cloud isn’t ready for enterprise users. ..................................... 45
4. You lose control with the cloud. ..................................................... 48
5. It’s the same thing as utility computing or grid computing. ........... 48
6. It’s only for low-end consumer applications. .................................. 49
7. It’s too isolated from my other data and applications. ................... 49
8. We won’t need PCs any more with cloud computing. .................... 50
9. Reliability will be a problem. ........................................................... 50
10. The cloud will give you performance problems. ......................... 50
IX. THE “PEOPLE CLOUD” ....................................................................... 51
THE END OF THE COMPANY AS WE KNOW IT .................................................... 54
The Virtual Company .................................................................................. 55
"Jobs for Americans" .................................................................................. 57
THE NEXT WAVE OF COLLABORATION............................................................... 58
X. GROUNDBREAKING CLOUD APPLICATIONS ............................. 61
Healthcare applications (Health-IT)............................................................. 61
Government: NASA and Nebula .................................................................. 64
XI. THE OPEN CLOUD ............................................................................... 66
Is open software free software? ................................................................. 66
XII. SECURITY AND RISKS ........................................................................ 67

Cloud computing provides superior physical security ................................. 69
THE FALLACY OF DIRECT CONTROL .................................................................... 71
Alternative Delivery Models ....................................................................... 74
Accessing the cloud ..................................................................................... 74
XIII. THE FUTURE OF CLOUD ................................................................... 76
Reactive Experimentation ........................................................................... 78
What’s Next? .............................................................................................. 86
About the Authors ...................................................................................... 88

I. Introduction
The cloud isn’t just a fad anymore – it’s the future of computing for the next
20+ years.

J
ournalists, pundits, and industry analysts are fond of referring to the
cloud as a technology revolution. That depiction is deceiving and
makes it seem as though the cloud appeared overnight out of some
mysterious laboratory. In fact, many of the underlying technologies
have been developing incrementally for several years.
The importance of the cloud lies in the fact that it is not just
technology—cloud is in fact, a business model, which is based on a
particular set of technologies. As the technologies evolve, they are having
a profound impact on the future of cloud computing, as we know it. The
"revolutionary" aspect of cloud derives from the fact that cloud
transcends the underlying technology that drives it—ultimately bringing
about major changes in how organizations make operational and strategic
decisions. Changes that led to the cloud were incremental—but the
consequences to the computing industry have been groundbreaking! For
example, consider virtual machines. Naysayers of the cloud may rightfully
point out that virtual machines have existed since the late 1960s.
However, only recently have the business models and enabling
technologies matured to a point that makes it possible to sell, manage, and
use virtual machines over the Internet almost as easily as buying a pair of
shoes on eBay. The same types of things are happening in the world of
software and computing platforms on the cloud.
This immediately brings up several important questions, which deserve
thoughtful answers: “What is cloud computing all about?”, “Why should I
care?”, and most importantly, “How does it affect me?”.
In short, cloud computing is completely real and is affecting almost
everyone. In this day and age, we have all become stakeholders in the
computing movement, and we are all affected when major changes occur.
Remember how things changed when the Internet came along? Changes
in computer technology seem to move at lightning speeds. It wasn't that
long ago that desktop computers had 20MB hard drives and people relied
on floppy disks for storage. For that matter, it wasn't that long before the
floppy disk era that there were no desktop computers, and computing
involved cardboard punch cards fed into a hopper.

7

It should be no surprise that
another evolution is upon us
once again, as there have been “Everything we think of as a
several since the dawn of the computer today is really just a device
information age. In this book, we that connects to the big computer that
choose the term “era” because we are all collectively building”
cloud computing is more than a
- Tim O’Reilly, CEO,
fad. Rather, we are entering the
type of radical shakeup that only O’Reilly Media
comes around once every 25 or
so years in the computing industry: a disruptive shift in the underlying
platform-of-choice. It is the realization of the Web as an ultimate
computing platform with unlimited power and size.
Cloud computing has been recognized as the fastest growing technology
movement. Forrester Research estimates that the global cloud computing
market will grow from $40.7 billion in 2011 to $241 billion in 2020.1 We
are likely looking at another decade of consistently strong growth,
followed by another decade of persistence.
Why else should you care? Lots of reasons, depending on who you are. If
you’re just writing a document or working from home, then you can
probably find online apps to do the trick without buying expensive office
software. If you’re an IT guy, even better—the cloud makes computing
easier to manage, drives down costs (as compared to PCs and dedicated
servers), and allows end-users to gain access to a broader range of
applications and services. Sure, PCs and dedicated servers have served us
well, but not without problems. They crash; they require us to buy,
manually install, upgrade, and uninstall expensive software; they become
bloated, slow and loaded with viruses. Wouldn’t it be so much better if
someone else could take care of all the hassles? With cloud computing, we
“rent” only what we need and somebody else manages the dirty work. Ask
any IT person about their work schedule. You'll find out quickly that
expectations and workload often exceed the reasonable amount of time
anybody really wants to work. More importantly, ask the CFO who signs
the paychecks. Do they want to cut costs? Absolutely! Cloud computing
will do it—cutting costs while giving the IT staff a break at the same time.

1http://www.zdnet.com/blog/btl/cloud-computing-market-241-billion-in-
2020/47702

8

At this juncture, cloud computing is more than a fad – it’s here to stay.
The cloud has become the computing platform of the future. There is a
popular quote attributed to Thomas Watson, founder of IBM: "I think
there is a world market for maybe five computers." That quote assumed
that computers were
only for the very
largest customers. “We are at the beginning of the age of planetary
We've come a long computing. Billions of people will be wirelessly
way since that interconnected, and the only way to achieve that
speculation, and the kind of massive scale usage is by massive scale,
general trend has brutally efficient cloud-based infrastructure.”
been to move
computing into the --Dan Farber, Editor-in-Chief, CNET News
hands of everybody
from big business
users, all the way down to preschool children. Cloud computing continues
that trend by bringing greater levels of access to high-end applications and
data storage, as well as new techniques for collaboration to the smallest
mom 'n pop businesses, telecommuters, and independent work-at-home
contractors.
Mr. Watson got many things right and to his credit once again, what if his
quote was saner than we once imagined? The term "cloud" refers to the
computing power that is available across the Internet. In a sense, the
cloud is rapidly transforming a worldwide network of computers into the
largest single [virtual] computer in the world.

II. What is cloud computing?
If the term “cloud computing” sounds confusing, then you are not alone.
Cloud computing sounds like a very fuzzy term, and like a literal cloud in
the sky, you can't really put your finger on it. It may help to understand
WHY cloud computing is so hard to understand:
 First, cloud computing is an extremely broad term. It’s as
broad as saying “desktop computing” (i.e. the PC), which
encompasses everything from the microchip to the Windows
operating system to the software. As we will learn in this book,
cloud computing encompasses all the same elements as the
desktop.
 Second, you cannot touch the cloud. Desktop computing is
easy to understand because you can see, touch, and feel your PC.

9

The cloud is real, but it is abstracted to the point where you
cannot see it, so it is harder to imagine.
 Third, the term is tainted by the “me too” marketing buzz.
The term "cloud computing", for a variety of very good reasons,
has become very popular, and there are plenty of new and
established IT companies that want to jump on the bandwagon,
often incorrectly labeling anything to do with remote computing
as the "cloud".
This book is entitled “Cloud Computing Made Easy”, so let us defer the
formal definition for now and begin with a simple working definition.
Here goes:
Cloud computing is “pay-as-you-go computing on the Internet.”
The cloud provides an attractive alternative to traditional desktop
computing. It is not simply about Web software like many people
wrongfully believe. Many major software vendors are saying, “We do
cloud computing too!” simply because their software works over the
Internet. Far be it from the authors of this book to disagree with some of
the biggest technology companies in the world; however, we will disagree
nonetheless. Web-enabled software is wonderful and very useful - but it
has been around for a long time. It’s nothing new in itself.
In reality, the term “computing” encompasses other forms of computing
beyond software. In many ways, cloud computing is strikingly similar to
desktop computing in that it encompasses the same three basic elements:
hardware (infrastructure), operating systems (platforms), and software.
The main difference is that, with cloud computing, all three elements are
"rented" over the Internet, rather than being managed locally.
Let’s take a closer look at the working definition above:
“…pay-as-you-go computing on the Internet”
What does it mean to say "pay-as-you-go computing on the Internet"? We
simply mean that you can log onto a website to do whatever you might
normally do on a PC or local server. For example, instead of paying $500
for a box of software, you might rent it over the Internet for a few bucks
a month, or instead of buying a $5000 server, you might rent it over the
Internet for a few bucks a day. You can essentially rent and manage all
your hardware over the Internet, configure computing environments
and/or run software. Cloud computing lets us do all of our computing on
the Internet as a viable alternative to buying, installing, upgrading,
uploading, downloading, backing up and otherwise managing physical

10

hardware, operating systems and software. It does not require a big
upfront capital investment because you only rent what you need, and only
as much as you need. With cloud computing, your PC is mainly used as a
way to run a Web browser, store some files, and occasionally print – that’s
all. The actual processing and computing is done by software on remote
servers that may be scattered across the Internet, thus the word “cloud.”
Incidentally, Google's Kevin Marks describes that the word cloud "comes
from the early days of the Internet where we drew the network as a cloud
. . . we didn't care where the messages went . . . the cloud hid it from us."2
The internet therefore gave us the first cloud, which centered around
networking. Later, data abstraction added another layer to it. Today, the
cloud abstracts the entire environment: infrastructure, platforms, data, and
applications.
The terminology is admittedly confusing. In cloud jargon, the term “as a
service” loosely refers to the ability to rent something over the Internet on
a ‘pay as you go’ (as needed) basis. You never actually buy or own
software on the cloud, but rather you pay to use the service.
The terms software, operating systems and hardware are confusingly described
as Cloud Software (or Software-as-a-Service), Cloud Platforms (or
Platform-as-a-Service) and Cloud Infrastructure (Infrastructure-as-a-
Service). To make matters worse, the acronyms SaaS, PaaS, and IaaS are
often used. Since this is Cloud Computing Made Easy, we’ve adopted the
lesser confusing terms: Cloud Software, Cloud Platforms, and Cloud
Infrastructure, though we will occasionally reference the other terms.

What does the cloud mean to your partners,
employees, and customers?
If we put aside the National Institute of Standards and Technology
(NIST) definitions for a moment, perhaps more important is what people
you do business with believe the cloud is, and what it will do for them.
People do have their own, vastly differing, definitions of the cloud, and
for the most part those definitions revolve around the value and the
changing business model, rather than the underlying technology. It is
important to understand what those perceptions are when making a cloud
deployment.
The cloud is then, from this perspective, the following:

2Dan Farber. "Defining the cloud." Video interview, Cnet.com.
May 7, 2008.
11

1. A business model enabled by technologies that allows for economic
value in terms of lower-cost software and infrastructure; from the
vendor's perspective, the ability to offer a more flexible pricing model that
appeals to a broader audience.
2. A business model enabled by technologies that facilitates rapid
development and deployment. From a users perspective, this means
gaining access to services quickly on an as-needed basis; from the
developer's perspective, it means rolling out services more quickly.
3. A business model enabled by technologies that is much more elastic—
almost infinitely so—that allows for much easier scaling in either
direction.
When the underlying technological definitions are stripped away, this is
what the cloud delivers—and ultimately, what it is.
What is "as a service"?
In cloud terminology, the phrase “as-a-service” is extensively used, which
simply means that a given cloud product (whether infrastructure,
platforms, or software) is offered in a way that it can be “rented” by
consumers over the Internet. By “rented,” we are implying that you pay
only for what you use. It is often described as an “on demand” service
because it is available whenever you need it. There are two immediate
advantages to the as-a-service model; first, up-front costs tend to be
substantially less; and second, it affords a greater level of easy scalability.
For example, if you store large amounts of data on premises, you’ll
probably buy extra servers and storage (over-provision) to make sure that
a shortage does not occur; then when you do reach capacity, you must
spend time purchasing and installing more. If you use storage-as-a-service,
on the other hand, the need for over-provisioning is eliminated, and you
simply purchase as much as you need on an ongoing basis, and the actual
provisioning of it is transparent.
There are several methods of offering a cloud product as-a-service:
The most familiar model used by cloud software is a per user/monthly
subscription. For example, a software provider may offer its collaboration
product over the Internet for $30 per month for each user. Another
approach is the advertising supported model, in which the offering is free,
but you need to stare at advertisements. In such cases, the vendor receives
revenues from the advertiser, rather than from the end-users. Facebook is
a popular example of the seemingly free, but ad supported model.
Likewise, cloud platforms employ both the per user/monthly and ad
supported models, as well as more creative models, such as assessing a fee

12

per record. In truth, cloud services may use any unit of measure to track
usage.
For example, cloud infrastructure may offer CPU time on a per hour
basis, assessing for storage usage, as well as assessing for data transfers per
gigabyte, often with differing rates for uploads versus downloads.
Amazon’s Elastic Computing Cloud (EC2) is a great example. Amazon
EC2 offers a console for creating virtual machines on a per hour basis,
with additional fees assessed for data transfers and storage.
NIST takes it a step further by asserting that true cloud offerings provide
certain expected characteristics, which may not have been present in
earlier Web-based software. These include such things as on-demand self-
service, resource pooling, and rapid elasticity. Naturally, on-demand
simply implies that the service is available to turn on or off as needed.
Resource pooling means that multiple users share a bank of servers
(including storage devices and other computing resources) over the
Internet, as an alternative to using dedicated servers. Lastly, rapid elasticity
means the cloud offering can be dramatically scaled up and down as
needed. As an example, let’s pretend that a person launches his own
dotcom and next week he is scheduled to appear on CBS News. Should
he buy ten new servers just in case? No! If he takes advantage of cloud
infrastructure, he can offer his software as-a-service, and scale it up and
down as needed. With as-a-service, you only pay for what you use, and
you can use as much as you want.

Cloudwashing
Beware - the phrase “as a service” is often abused by vendors. It has
become the emperor’s clothes of cloud computing.
Many vendors are trying to circumvent millions of dollars in reengineering
costs by sticking the letters “aaS” at the end of their antiquated offerings.
It may not really be cloud, but it’s a heck of a lot cheaper than
reengineering. Although this next point should go without saying, here’s a
helpful hint: If software isn’t truly available in as a service, then it’s not.

III. Why Cloud Computing?
To the casual end user who is just trying to get some work done there
may seem to be little difference between cloud computing, desktop
computing, and any other type of computing model that has been floated
around over the past few decades. He or she may even use the same types
of software applications to do the exact same types of things. That’s the

13

point! Cloud computing offers a better way to do the same types of
things.
So then, why is cloud computing any better than ordinary desktop
computing?
The answer depends on who you are.
Cloud computing for end-users
As an end user, cloud computing lets you run software applications and
access data from any place and time, and from any computer; without the
need to ever install, upgrade, or troubleshoot software applications
physically on a local desktop or server. This is one of the most important
elements of cloud computing and why it has become so popular today. In
a sense, cloud computing outsources the technical hassles to someone
else.
Cloud computing also makes it easier to do work anytime and from
anywhere, often referred to as “ubiquitous.” The old model of working
involved going to the office from 8:00 to 5:00, and getting on a plane and
taking a business trip or two every year. If we did work from a location
outside of the office, then when we returned to the office, time had to be
spent synchronizing the ad hoc work done at home with the in-office
systems. Today's model of working is different. We can get just as much
done at home or on the road as we can in the office. We can connect
instantly to the office from anywhere in the world, gain secure access to
our applications and data, and in short, get things done in a way that was
never before possible.
Cloud computing for system administrators
Keep in mind that almost all PC owners have become system
administrators in a way, unless we’re fortunate enough to have access to a
teenager to install and manage things for us. If your PC has ever crashed
and wasted your day, then you’ll understand the benefits of somebody else
doing the dirty work. The problems can get out of control inside big
companies, which manage thousands of software configurations, and pay
employees whether their PCs work or not. The superiority of the cloud
model comes in when we start to realize that desktop applications are
more or less static, and cloud applications can be continuously refined.
Desktop applications must be physically installed on a PC, upgraded
periodically, have patches applied when they become available, and re-
installed when the user moves to a new desktop, or when the old one
crashes. The cloud model eliminates those inconveniences. Need a new

14

PC? Just buy one. You can still access your cloud applications without
having to re-install anything. System administrators, who may need to
manage hundreds, or even thousands of desktops, remote devices,
servers, storage arrays and other equipment, quickly get bogged down—
and the cloud model makes their lives easier.

Cloud computing for software developers
There is an even bigger advantage on the development end. Because the
applications are delivered from a common code base from a central
location, upgrades to the application, patches and fixes can be pushed out
to the user transparently. Desktop applications require the user to actively
install a patch, or at least, allow an auto-connection to take place.
Microsoft Windows uses the auto-update feature, which has become very
useful and convenient, for example. However, it still requires patience on
the part of the end user, who must wait for the upgrade to come in over
the Internet, and then must re-boot the system for it to take effect. A
cloud application, since it does not exist on the desktop, does not have
that requirement. All upgrades take place on the back end, requiring no
intervention, action, attention, or patience from the end-user. This makes
it much easier for developers to continuously upgrade their applications,
and to push those upgrades out to users on a real-time basis. Going a level
deeper to the platform stage, cloud computing gives developers another
critical advantage. Since the platform provides developers with a common
set of cloud services that have already proven to be robust, all applications
are that much more stable—and quicker to completion as well.
Cloud-based deployment environments, inexpensive and robust software-
as-a-service, and easy to deploy platform-as-a-service that serves as a
powerful means to create apps quickly, all are factors that lend themselves
to rapid development, rapid prototyping, and the concept of "minimum
viable product" (MVP).
The concept of MVP is deceiving. It does not, as the name might imply,
involve producing applications that are substandard; rather, it is the
driving force behind the lean software movement.
Much software today is released incrementally, and competitive pressures
require development firms to issue that first release as soon as possible—
before the next person comes out with the same thing. Waiting until a
software product is perfect, includes every conceivable feature consumers
will want, is completely bug-free, and is a recipe for never releasing at all.
A quick and early release—typically called the "beta"—gives the
developers a chance to gauge consumer feedback, and learn what's most
important.

15

The need to produce software applications quickly is driven by the
available tools to do so. This is where platform-as-a-service really shines.
PaaS can be used to create sophisticated, robust, and complex software
applications, but it does so quickly; allowing the developer to meet those
competitive requirements, prioritize the addition of product features, and
get a product out the door before the competition.
The lean software movement is about producing high-quality applications,
iteratively, with an accelerated go-to-market strategy. Using PaaS as a lean
software tool allows for the developer to more heavily rely on consumer
feedback as part of the development process, and assist the decision-
making process in pinpointing the best features to assign priority.
Cloud computing for IT buyers, corporate and
federal
The critical advantages listed above have not been lost to corporate users.
The ability to lessen the workload on system administrators and
developers alike lets companies save dollars spent on labor. In short, your
company can do more with less, and with greater efficiency. Besides the
labor advantage, companies will also gain an advantage in terms of
reduced capital expenditures. Why? The cloud not only reduces time spent
on admin duties and development, it also addresses the physical
infrastructure itself. Companies taking full advantage of cloud computing
will enjoy a reduced need for servers and storage arrays—providing
another source of savings (and in turn, reducing the system admin
overhead even further).
In the corporate world, one of the most important parts of business is
improving the bottom line. That's done either through increasing revenue,
or by decreasing costs. When decreasing costs, the ideal scenario is to do
so while still maintaining the same or better level of efficiency the
company enjoyed before the decrease in costs; cloud computing provides
the answer to that need. Let's look at a few of the dollars-and-cents
statistics:
Enterprise software represents an enormous expense, as some $800
billion a year is spent on purchasing and maintaining software. The bulk
of that—or about 80 percent of the $800 billion—is spent not on the

16

actual purchase of software, but on installing and maintaining it.3 The
federal government alone spends $70 billion a year on IT systems, much
of which goes toward enterprise systems. Most servers operate at only
about 15 percent capacity at most times, and over-provisioning is
regrettably common. Virtualization, an important element of cloud
computing, allows the data center operator to make full use of server
capacity. Enterprise cloud platforms can save even more.
The advantage to individuals, small businesses, and large enterprises
which buy software is obvious. The cost of software represents a major
expense for businesses of all sizes. The presence of cloud computing
options has allowed many small and midsize businesses to gain access to
important features of high-end, enterprise-class software that would not
otherwise be available. As a result, a major barrier to success has been
dissolved, and the saga of million-dollar price tags for enterprise software
is nearing an end. Large corporations will save money; and smaller
companies will gain the advantage of being able to access more software
resources, which were
previously unavailable due
to either high cost, or the "If you move your data center to a cloud
software simply being provider, it will cost a tenth of the cost."
unavailable for smaller
implementations. - Brian Gammage, Gartner Fellow

IV. Cloud
Computing Up Close
We’ve informally defined cloud computing as “pay-as-you-go computing on the
Internet.” A more purist definition of cloud computing is one that
differentiates true cloud computing from mere software on the Web.
After all, Web software has been around for more than a decade and
cloud computing is relatively new, so how can they possibly be the same
thing, right?
Rather, true cloud computing takes advantage of new enabling
technologies and cloud constructs, which are making the movement
possible. In essence, Web software has been around for years, but until

3Peter Mell and Tim Grance. "Effectively and securely using the
cloud computing paradigm." National Institute for Standards and
Technology, 10-7-09.

17

recently, it has been prohibitively costly for the masses to develop and
host – often costing millions of dollars and taking years to develop and
implement. Cloud computing changes all that by incorporating
virtualization technology that allows the physical infrastructure to be
rented for mere pennies compared to the old ways of engineering Web
software. It further changes the equation by providing cloud-specific
platform toolkits to accelerate development.
The National Institute of Standards and Technology (NIST) puts it this
way:
"Cloud computing is a model for enabling convenient, on-demand network access to a
shared pool of configurable computing resources (e.g., networks, servers, storage,
applications, and services) that can be rapidly provisioned and released with minimal
management effort or service provider interaction."4
NIST also categorizes cloud computing into three “as a service” offerings,
namely infrastructure, platforms, and software, which are broken down in
more detail here:
A. Cloud Infrastructure (Infrastructure-as-a-
Service, or IaaS)
In the old days, if you needed a server, you might spend between five and
ten thousand dollars or more upfront and then you’d pay a techie $80,000
a year plus benefits to manage it. Nowadays, with cloud infrastructures,
you can actually buy a “virtual server” over the Internet almost as easy as
signing up for an email account. The server never arrives at your
doorstep. Instead, it stays out on the cloud where you can log on and
manage it anywhere and anytime over the Internet. For the techies
amongst us, that means that you can buy and manage processing time
(CPU time), storage, network capacity, and other fundamental computing
resources without shelling out big bucks upfront. You only pay for what
you use.
If you’ve never used cloud infrastructure, then the very notion of using a
server over the Internet may sound crazy. You’re probably asking the
question, “How can I do that?” Well, actually, it’s easy to rent computers
over the Internet. The easiest way to learn is to actually do it. If you go to
Amazon.com’s EC2 website, you can launch and manage a real live server

4Peter Mell and Tim Grance. The NIST Definition of Cloud
Computing. Version 15, October 7, 2009.

18

(well, actually a small virtual server) for an hour for about 20 cents. What
a bargain! You will specify a server name, the type of operating system
and other details to create your server instance. Then, you can log on
using a remote desktop or visit the website, just as if it were a real live
server.
How it Works
We use the term “virtual server” because you’re not really renting a
physical box. That is, you can’t actually walk into a room and touch your
dedicated machine. Instead, it’s all managed by “virtualization” software,
such as VMware.
In the old days, one operating system would run on one physical box. For
example, you would buy a Windows server that contained a copy of
Windows, or a Linux server that contained a copy of Linux. It was one-to-
one.
Virtualization, on the
other hand, lets you
run multiple operating
systems on the same
box.
In the early days of
virtualization, this was
a handy trick. For
example, virtualization
made it easy to test
new software on
multiple operating
systems without
needing multiple
physical boxes. Virtualization also made it easy to run Windows and
UNIX programs on the same physical box, such as when a program was
only available for one operating system.
Then, one day, somebody realized an even bigger trick. With
virtualization, it’s possible to sell the same physical machine multiple
times. That is, a data center can run 10 copies of Linux on one box, and
then sell it over the Internet like different servers. Voila, the basic concept
behind cloud infrastructure (infrastructure-as-a-service).
In reality, cloud infrastructure is not limited to a single server, but rather
relies on a shared pool of servers, whereby any one user can scale up to
take advantage of extra computing power when needed. It works because

19

servers are mostly unused anyway, so there’s always some extra
computing power available if the pool is large enough. It works somewhat
like a bank that loans the same money 10 times over. In theory, it
wouldn’t work if everyone demanded their money (or computing power)
at the same time.
So where is all this infrastructure?
The cloud's infrastructure consists of actual, physical hardware that is
complemented by a delivery mechanism. The main difference is that it
exists outside of the user's immediate grasp, and its existence has been
abstracted to the point where its exact location is both unknown and
irrelevant to the users. When you work in the old way, you know that your
application is located on the PC in front of you, and the data is held on
the data server in the room across the hall. With cloud computing, you
don't know whether your applications and data are in a data center in Des
Moines or Delhi, and it doesn't make a difference.
This is an important point. One of the biggest objections to cloud
infrastructure is that you can't put your finger on it. There is a certain
satisfaction to walking into your server room, and being able to point to a
rack of servers and storage arrays with a glorious mass of cables coming
out of the back, and saying "that's where our data and applications are."
But, that satisfaction is an illusion; in reality, there is no inherent
advantage in being able to reach out and touch your own hardware
infrastructure. If it works, it works; it doesn't really matter where it lives.
In fact, if your infrastructure is elsewhere at a hosted facility, then you are
gaining a strategic advantage of having somebody else who specializes in
such things manage it for you.
An ordinary computing infrastructure may consist of several physical
pieces of hardware and cables that you must maintain and allocate. A
cloud infrastructure consists of a pool of highly abstracted and scalable
infrastructure devices existing in multiple provider data centers, connected
over virtual private Internet connections, where a trusted third party is
charged with maintenance and allocation.
Virtualization
Virtualization is the behind-the-scenes enabling technology that makes
cloud infrastructure possible. Just as you can drive a car without
understanding how the engine works, so too is virtualization. You don't
need to know how virtualization works to use it, since it involves the
infrastructure, which the end user no longer has to worry about with a
cloud implementation. The concept of server virtualization allows many

20

"virtual servers" to run on a single physical server as if each one were a
separate device. Storage virtualization works the same way. Both types of
virtualization essentially decouple the function from the underlying
hardware, and virtualization has become a common enterprise technology
for saving money and making better use of existing resources. It's not
uncommon for a large enterprise to employ this technique in its own data
center, and the technology is in common use by cloud providers—
delivering an efficient method for them to service the needs of multiple
clients on a cost-effective and secure basis.
Why does virtualization make better use of resources? Because of
technologies like thin provisioning. In storage for example, traditionally a
volume was created for each application. This storage volume was always
over-provisioned, to ensure that adequate storage would always be
available; as a result, it was common for storage environments to be
making use of only 30 percent or so of available storage. Virtualization
uses thin provisioning to allocate storage virtually, instead of absolutely;
freeing all of that previously trapped storage space. Server virtualization
also makes better use of resources by not requiring separate, dedicated
hardware servers for each application. In virtualizing servers, it allows a
single physical server to act as multiple virtual servers, each one separated
absolutely by a virtual division that isolates each one. The "walling off" of
the virtual servers within the single physical server addresses the obvious
concern that there would be some potential for somebody else on the
same physical server to access your data.
This is an important concept for cloud computing, since cloud computing
is all about abstraction. For the end user again, the concept of applications
and storage is abstracted to the highest degree—and in many cases, the
user will not even be aware of where the actual application or data is
located. It doesn't matter. Virtualization works hand in hand with cloud
computing to provide the abstraction that is necessary for both.
Virtualization provides a type of technology that allows applications to be
moved around freely onto different devices that exist in the cloud,
transparently to the end user.

B. Cloud Platforms (Platform-as-a-Service, or
PaaS)
For most people, the term "cloud platform" is even fuzzier than cloud
computing as a whole. Yet, when cloud platforms are properly
understood and embraced, they potentially offer the greatest impact over
any other aspect of cloud computing.

21

Cloud platforms can drive down software engineering costs tenfold,
reduce time to market, improve profit margins, and lower risks. They can
promote higher levels of security and system interoperability, and can
allow system integrators to enter into new markets within days, instead of
years. They can dramatically lower the skill requirements needed to create
new software applications, so that entrepreneurs are empowered to serve
their customers, and customers are empowered to serve themselves. In a
nutshell, cloud platforms takes cloud computing to the masses.
Let’s start with a simplistic understanding of the term “platform” for
computing, and then we will expand our definition to the realm of cloud
computing.
A platform generally refers to a “prefab” software architecture upon
which you can build computing solutions. It provides core software
functionality, which would otherwise need to be engineered from the
ground up. Can you imagine building an oven every time you wanted to
cook dinner? Probably not. Fortunately, the oven is already built;
otherwise, your meals would cost $500 each.
Likewise, the cloud needs platforms to do a lot of the grunt work, which
otherwise needs to be engineered into every software application from the
ground up at great expense. Cloud platforms serve as a launch pad for
cloud software, providing “prefab” functionality such as a user interface,
user sign up and administration, role-based security, federated search,
multi-tenant data management, and so on. If you’re asking the question,
“What the heck is multi-tenant data management?” then exactly! You’re
seeing the point. It’s complicated stuff, and you wouldn’t want to program
it into every app. Unfortunately, most cloud developers are not taking full
advantage of platforms.
By their very definition, cloud platforms are offered “as a service”,
meaning that you can use them over the Internet with no need to install,
upgrade, or host. Cloud platforms are readily distinguished from other
platforms, which require installations, upload, downloads, and managed
hosting. As-a-service means that cloud platforms are easy to use. More
important, if you build cloud software on top of a cloud platform, then
your solution is inherently cloud-enabled, taking advantage of underlying
cloud infrastructure, elasticity, and as-a-service models.
Cloud platforms also may include online tools and APIs that make it
easier for developers to build on top of the platform. When choosing a
cloud platform, it’s important to make sure that the API is open, allowing
for integration with 3rdparty, open source and legacy software, and web
services; otherwise, you could become overly locked into the platform

22

provider for all your needs. This is referred to as an “open platform”
versus “proprietary vendor lock-in.”
Major Benefits of Cloud Platforms
On the surface, it’s easy to think that cloud platforms are for software
developers, but it’s the IT buyers who are suffering the most from
astronomical software engineering costs and delays.
For IT buyers, investors, and developers the advantages of cloud
platforms are tremendous. Creating a cloud application from the ground
up is a complex process, involving not just ordinary coding, but also
adding a layer of abstraction, and incorporating a far-flung
communications layer, as well as security protocols. If every SaaS provider
had to create each of these things from scratch, then cloud-based
application development would be hindered, and limited to only the larger
software companies. Cloud platforms address this problem by allowing
developers to build cloud applications on top of an existing architecture
that includes core functionality. In essence, developers can use platforms
to get their software to the “80-yard line” without programming and avoid
reinventing the wheel.
The benefits are many:
 Lower costs – In some cases, a cloud platform can reduce costs
by 80% or more, because non-core code is already engineered.
 Lower risks – Likewise, a cloud platform can reduce risks by as
much because common functions are already tested, sometimes
over a period of years.
 Faster time-to-market – Cloud platforms dramatically reduce
time-to-market because they serve as a launch pad for software
engineering efforts
 Higher profit margins – Software developers and system
integration firms can deliver more for substantially less, thus
higher margins on fixed price contracts.
 Rapid prototyping – Create and deploy concept applications
without writing code.
 Higher security and interoperability – NIST indicates that the
cloud suffers from major security issues, largely because vendors
are implementing disparate and unproven security models. Cloud
platforms provide a common, proven security model. If cloud
software uses the platform, then it is inherently secure.
23

As an added advantage, cloud platforms shield both software engineers
and end-users from the behind-the-scenes complexities of the entire
cloud. Dan Tapscott, the author of Wikinomics, talks about the growing
complexity problem this way, “the Web look[s] increasingly like a
traditional librarian’s nightmare --- a noisy library full of chatty
components that interact and communicate with one another.” He is
referring to the cloud as a cluttered hodgepodge of Web apps and
services, each with their own logins, data sources, and security/resource
functions. In the absence of cloud platforms, we are recreating the wheel
millions of times over. In a few years, the redundancies will drive up costs
by billions within federal IT systems, health-IT systems and other
enterprise IT systems that rely on cloud services. All these IT systems will
struggle with disparate security models and interoperability concerns.
Unfortunately, cloud platforms remain vastly underutilized. That’s why
some enterprise software systems unnecessarily cost millions of dollars
and take years to implement, only to eventually fail! The resistance to
platforms is sometimes baffling. It’s almost as if the logic is to stick with
what we know, even though it doesn’t work. Ironically, some software
integrators are creating totally proprietary stovepipes from the ground-up,
just to avoid platforms. As a result, IT buyers are paying more than twice
as much for their systems, and being locked into developers. Instead, they
should be taking advantage of open APIs that are available with some of
the more open cloud platforms.
Cloud Platforms as Middleware
It may also help to think of a cloud platform as the middle layer of a
three-layer cake, in that it rests between the hardware and the software.
Sure, you can remove the middle layer, but in doing so you’re also
removing a lot of important “cake” that somebody needs to bake from
scratch. In the case of software engineering, that’s some expensive cake.
This means that software can be built without using cloud platforms, but
the costs of doing so can be detrimental; creating a barrier to entry for all
but the largest development shops. You see – platforms actually do a lot
more than just provide core functionality for software. They also lower
the time and risks of engineering software dramatically because the
platform engineers have already worked out the devils in the details on
their own dollar.
Platforms also reduce the software footprint and maintenance costs,
because the responsibility for maintaining platform code is essentially
outsourced to a platform provider, who achieves economies of scale by
maintaining one system.

24

Web software that is created without platforms is considerably more costly. On the other
hand, cloud platforms provide core functionality that dramatically reduces time, risk,
and development costs.

Types of Platforms
Cloud platforms come in many shapes and sizes, depending on the
application at hand. Arguably, Google is currently dominating the
consumer app platform, whereas Facebook is dominating the social
networking platform, and Salesforce.com is trying to make a footprint as
an enterprise software platform. Several other players, Yahoo included,
offer ecommerce platforms, which have driven down the time, risk, and
cost of ecommerce solutions.
The term "platform" as it relates to cloud computing is often misused to
refer to customizable software. Software that can be customized is simply
that: customizable software. Similarly, cloud infrastructure vendors
sometimes promote their products inaccurately as platforms. A platform,
on the other hand, is something entirely different. A cloud platform rests
between the physical infrastructure and customizable software.
There are, at the highest level, two categories of platform-as-a-service:
Hosting and deployment platforms, and software development platforms.
While there may be different terminology applied by various research
groups, and these groups may break down the definitions to greater levels
of granularity, at the highest level the categories remain fundamentally the
same.

Type 1 PaaS: Runtime Platforms (BYOC)
Cloud platforms today are moving towards more of an end-to-end
proposition, from concept to deployment. Still, the majority of PaaS
offerings from large software vendors—though evolving—still focus on
providing auto-scaling and runtime services only. This basic type of
runtime platform usually support advanced hosting, auto-scaling and
easier deployment. Runtime platforms do nothing to help create the actual
coding and applications. It is certainly more than ordinary hosting—it is
enhanced at the infrastructure level. Nonetheless, it still presents a "bring
your own code" proposition. This type of PaaS is extremely useful for
deploying quickly and scaling, but is not truly end-to-end because it does
not assist in actually creating the code for the app. The development side
is a legitimate and essential part of PaaS for several reasons, most

25

importantly; cloud apps need to be developed with tools that specifically
take the cloud into consideration. A true end-to-end PaaS tool will offer
development facilities that assist in the creation of cloud apps, with cloud
functionality built in.
Runtime PaaS does offer significant advantage, especially for
environments with lots of customers in that it makes it easier to deploy
and scale software for a large number of users—since it essentially takes a
single, common code base and pushes it out to tens of thousands of users
on a subscription basis. While it offers nothing in the way of true
customization, it does offer the ability to allow selection of pre-configured
options.
Further, it goes beyond traditional hosting, in that it allows for additional
features for scaling, performance monitoring, provisioning virtual machins
up and down as needed, load-balancing, and consumable database
resources. Without these features, providing scalable SaaS applications
would be difficult.
A few examples of this type of PaaS include: Google AppEngine,
Amazon Elastic Beanstalk, Microsoft Azure, Salesforce, Heroku, Engine
Yard, and OpenShift. Heroku, an integrated hosting environment for
deploying scalable Ruby on Rails, Java, and other web apps, was one of
the first cloud platforms. It's open source and supports Facebook apps.
On the downside; however, it is opened by Salesforce, so despite its open
source underpinnings, its future is still unknown, and its success is tied to
a single company.
Google AppEngine, used for developing and hosting web apps in a
Google-managed data center, sandboxes apps and runs them across
multiple servers. It offers automatic scalability, and is free up to a certain
level with incremental fees for additional services. While it is backed by a
strong vendor (Google), it is still highly proprietary and non-portable.
Microsoft Azure, used to build, deploy, and manage apps at Microsoft-
managed data centers, can also connect on-premise applications with each
other. While it does offer robust services and strong backing, it is also
highly proprietary and non-portable.

Type 2 PaaS: Software Development Platforms
The second type of PaaS, which is less prevalent but will emerge as the
dominant model in the near future, provides more of an end-to-end
toolset that provides online facilities for building and deploying software
on the cloud.

26

Going far beyond the first type, this implementation is true end-to-end
PaaS. It provides the following:
 Online tools for developing and deploying SaaS
 No uploading or downloading or desktop development
 Push button deployment with no server-side configuration
 May co-exist with hosting/runtime environments. For example,
SaaS Maker™ runs on Amazon EC2, or may be deployed on
Microsoft Azure to build and deploy applications to massive scale
Examples of end-to-end PaaS systems include SaaS Maker, AppDeployer,
along with Force.com, WordXpress, and Cordys. Force.com is used
specifically for creating SaaS apps on the Salesforce infrastructure, and it
was one of the first major PaaS offerings on the market. Though it does
offer mature tools and is well suited for creating apps that are related to
Salesforce automation, it is proprietary and requires a commitment to
Salesforce programming, data objects, and infrastructure. It can't be used
to publish apps as a true cloud service on-demand for subscription-based
revenues.

What PaaS is not
We've outlined what PaaS is, and what people perceive it to be. Now what
it is not:
Concepts similar to PaaS have been around for a long time, but PaaS in its
full glory is a new and disruptive offering. Traditional middleware or
databases are not PaaS. If it’s not offered “as a service,” then it’s not.
Beware of cloudwashers who try to rebrand traditional platforms as cloud.
Serving the Enterprise
Of all types of platforms, enterprise business platforms may provide the
greatest value in the near future, simply because enterprise business
systems are extremely expensive – sometimes costing tens of millions of
engineering dollars. By enterprise business system, we are referring to the
types of scalable multi-user / multi-tenant cloud-enabled software that
government agencies and Fortune 500 companies spend millions on,
sometimes without blinking an eye. For enterprise business systems,
platforms offer such great benefit only because the engineering costs are
otherwise so high, sometimes representing more than 95% of the total
cost of ownership.

27

C. Cloud Software (Software-as-a-Service, or
SaaS)
The most important, and most visible of the three elements is Cloud
Software, without which there is no need for Platforms or Infrastructure.
Again to stress, cloud computing and cloud software, i.e. software-as-a-
service (SaaS), are not the same things. The two terms are often used
interchangeably, but it would be incorrect to do so. Rather, cloud software
is just the software part of the cloud computing triad. It is without a
doubt though, the most visible part, since it faces the end user. In a purist
sense, true cloud-enabled software refers only to that software which
intentionally takes advantage of the other cloud computing technologies:
namely cloud infrastructure and cloud platforms.
The overall market for SaaS subscriptions, compared to on-premises
software, is still young, though it is a rapidly growing niche. Because it is
the most visible part of cloud computing though, it will be SaaS that
drives the growth of cloud computing. Already we are seeing this growth,
driven both by startups moving into SaaS offerings from the ground up,
as well as established IT giants like Google and Microsoft moving into the
SaaS market. Almost every major software vendor today has at least a test
SaaS program in the works, and many have them in the market already.
When we talk about software-as-a-service, it usually means that the
software being delivered has a common code base that is delivered to
multiple users. However, this does not preclude customization. Using a
common code base for SaaS applications has a big advantage, in that it
allows the SaaS provider to continuously refine the program, and push
those refinements out to each user on a timely basis. This not only makes
for a more robust piece of software, it also allows the cost to be shared
between many users. However, customization is still allowed. Each end
user may for example, be able to choose from multiple software
components to create a SaaS application that very specifically meets their
own precise needs; and of course, just like most types of on-premises
software; SaaS applications allow each end user to apply their own user
preferences and custom configuration.

V. Other cloud offerings
This section addresses other considerations of cloud computing, which
didn’t fit nicely into a category, but are too important to overlook.

28

Cloud Brokers
Cloud Service Brokers (CSBs) are receiving a lot of attention lately.
Brokers act as intermediaries between buyers and sellers of cloud services.
For example, a buyer may go to a single broker to buy virtual machines
from three different providers. If you buy many cloud services, then
brokers can make your job a lot easier because you only need to deal with
one seller for everything you buy. In theory, brokers will become
increasingly important as the cloud becomes increasingly convoluted. A
value-added broker may do more than aggregate and resell services by
offering shared services such as a single sign-on, or provider-to-provider
migration services, etc.. They may also offer online price comparisons and
other niceties.
If the notion of brokers is confusing, then consider the travel industry as a
simple analogy. Instead of visiting dozens of different airlines, hotels, and
rental agencies, we can save a lot of time and money by visiting a travel
agency that brokers the various services. Cloud brokers work in similar
ways, except they sell virtual machines, for example, instead of selling
airplane tickets and car rentals.
At the time of this writing, the brokerage market is primarily focused on
aggregating and reselling infrastructure services. They do little to help with
buying software and platform services. This model will likely mature with
time.
What are "Web services"?
The term "web services" is a bit oversimplified, and it implies that it is just
a service that you access over the web. In reality, as a formal definition,
web services are usually considered the domain of web programmers, not
end users. It is a programming technique that involves use of remote
subroutines, which can be called over the cloud, such as making a
calculation or authenticating users. In the case of cloud computing, web
services allow programmers creating cloud programs (SaaS) with ways to
manage the cloud infrastructure, or integrate with other cloud programs.
Using technologies such as SOAP, XML or WSDL, web services simply
provide an ability to allow programmers to use other peoples' offerings
over the Internet.
Supercomputing-as-a-Service
Typically thought of as the domain of wild-eyed scientists working on
large-scale projects that are far beyond the scope of ordinary business,
supercomputing occupies a mysterious place in the computer business.
But, let's draw a comparison—as recently as the 1970’s, computing in

29

general was thought to be the exclusive domain of a handful of extremely
large companies and government agencies. Computers weren't for
ordinary people, or even for small companies. But, look where we are
today. The room-sized computers of the '60’s aren't even as powerful as a
simple netbook. Supercomputing today is in the same place that general
computing was fifty years ago.
When supercomputing meets the cloud, its power becomes available to a
much broader audience. And, that's what is already happening. Maybe you
can't have a supercomputer in your home—at least not yet—but you can
access one over the cloud. Companies like Exa sell their supercomputing
processing power over the cloud, and companies that don't necessarily
have big budgets can harness the power of supercomputing
environments. There are already a small number of companies that offer
supercomputing as a cloud option, including the venerable Amazon,
whose MapReduce offers supercomputer-like capabilities to crunch large
data sets in Amazon Web Services.
High Performance Computing as-a-service (HPCaaS)
Along with supercomputing, the cloud is also changing the face of high
performance computing (HPC).
Supercomputing has always been expensive, often costing tens of millions
of dollars. Nonetheless, they’re viewed as unavoidable by many members
of the scientific community. In recent years, grid computing has gained
attention as a possible alternative. The notion with grids is to take
advantage of otherwise idle CPU time that’s available on millions of
computers. With grids, special software divvies up and “outsources”
calculations to several computers in parallel, such as to PCs that act
somewhat as mini-servers. Historically, grids were manually orchestrated,
and relied on, other people’s computers, which raise questions about
security and privacy.
With cloud infrastructure, we have already learned that servers can be
allocated dynamically as needed (as in "thin provisioning"), rather than
paying for unused computing power. Then, this begs the ten million dollar
question: Why can’t I just harness the power of 100 servers when I need
it, run a calculation, and then shut them down? That way, I wouldn’t need
to buy a supercomputer, right?
That’s exactly what HPC as-a-service does. Special HPC cloud software,
including open source software like Univa UD, makes it possible to turn
computing nodes on and off as needed, while orchestrating intensive
calculations on those nodes. With cloud HPC, the concept is that a

30

supercomputer never rests idle, doesn’t become comparatively outdated in
a few years, and has no hard limits on scale. The future of cloud HPC is
yet to be determined. In the meantime, it will be fun to keep an eye on
how the technologies mature for adoption by the serious scientific
community.

VI. Cloud as-a-Necessity
Technologies become essential when they become a part of the very
fabric of society. They become essential when they become disruptive.
There are a great many new technologies that appear every year, and many
of them are technologies designed to make things simpler, cheaper, and
more convenient. Yet, most of them do not fall into the category of
disruptive technology—or a technology that results in far-reaching and
important changes in the way people work, think, do business, and
communicate. Cloud computing is one of those disruptive technologies.
 Cloud computing changes the way we work. The very nature
of what a "job" is, is changing. We work from home. We work as
contractors. We telecommute, work from on the road, and
increasingly, pay no attention to the physical boundaries of the
corporate brick and mortar walls.
 Cloud computing changes the way we think. Old barriers are
being broken down. We are no longer afraid to think outside the
box because the box no longer exists.
 Cloud computing changes the way we do business. The
collaborative technologies that are enabled by the cloud let us
take advantage of outsourcing, focusing on our core goals while
letting other experts take care of what they do best on our behalf.
 Cloud computing changes the way we communicate. Is it
necessary to get on a plane, or drive across town for a meeting?
Increasingly, the answer is no. New types of communication
allow us to work closely with partners, remote employees, and
suppliers around the world as if they were right there in our
office.
Cloud computing is destined to become part of our everyday lives,
because it is more than technology. It is not just software that is delivered
from a remote server over the Internet. Cloud computing represents a
new way of thinking and doing that has become essential to stay
competitive and efficient in today's economy. Here are just a few of the

31

drivers that highlight why cloud computing has grown in importance so
quickly:
 Explosion of data. We are truly in the "information age" today.
That means we rely on it, but it also means that there is a lot of
it.
 Renewed focus on collaboration. So what do we do with all of
that data? Information is usually more valuable if it is strategically
shared, not only within the company, but also with partners,
suppliers, outsourcers, and other stakeholders all around the
world.
 Economic necessity. Companies face the continual need to cut
costs, especially during the worst economic recession since the
'30’s. But, even apart from the recession, global competition and
other factors have led companies to embark on major cost-
cutting initiatives. This involves both implementing new methods
and cutting staff.
 Entrepreneurial activity. The economic recession has a positive
impact on entrepreneurial activity. The result is that there are
more small companies today than ever before, and those small
companies need access to resources at low cost. Cloud computing
allows these small entrepreneurial ventures to gain access to the
services they need and flourish.
 Outsourcing. Outsourcing and cloud computing go hand-in-
hand. The outsourcing trend is driven by the economic necessity
described above, and it flourishes because of the intense amount
of entrepreneurial activity that we are seeing from two
perspectives. Many of the small entrepreneurs that are launching
their companies today are outsourcing providers and the demand
on the part of larger existing companies for cost-cutting further
drives the need for outsourcing. Cloud computing provides the
framework for outsourcing to exist.
 Teleworking and telecommuting. Yes, people are working at
home, and companies are allowing it, in part out of the effort to
keep costs in check. Cloud computing has provided the
framework to allow a new era of working at home to become
reality.
With so many factors coming into play at once, we are seeing a "perfect
storm" that can have only one result: Cloud computing becomes

32

pervasive. In every one of the above drivers, cloud computing is what
makes it happen.
Custom software on the cloud
The simplicity of the cloud, ever since the beginning of the cloud
movement, was built around the fact that a single cloud provider could
push out a software-as-a-service application. It was built from a common
code base and deployed from a single cloud data center over the Internet
to thousands of users, which could have de facto customization by
selecting, or de-selecting, a pre-configured set of options.
That model was, and still is, valid. It delivers easy access, simple
maintenance, low cost, and a variety of other options.
Because of the obvious advantages of that model, until recently the cloud
focused mostly on delivering those types of commodity infrastructure and
software services, such as computer services, storage, email, and
collaboration, which require very little customization. For those types of
commodity services delivered over the cloud, offering customization
simply through offering available "on" or "off" options has been more
than adequate, and has delighted end users and IT managers alike.
However, custom software has still been a per-case project, and when it
comes to a custom installation, the cloud has often not been part of the
discussion. No more.
This is changing now, thanks to the continuing evolution of PaaS.
Software developers and integrators can use these increasingly robust and
feature-rich development and deployment platforms to build complex,
sophisticated software using a cloud-based development system, and then
deploying them on the cloud.
Those mega-million dollar custom installations, which often required
more millions for private data centers, are rapidly becoming outdated, and
the next wave of cloud computing will go beyond commodity products
and deliver highly custom services.
Self-healing nature of cloud computing
A true cloud computing architecture is self-healing, which promotes
higher uptime, and less likelihood of failure. Self-healing is really nothing
new; it is based on technologies that are often used in large enterprises
and data centers. It simply means that should a failure occur, technology
and protocols are in place and will automatically correct that failure in real
time. This is the heart of disaster recovery, and is part of the cloud
computing model.

33

For the provider of the cloud architecture, this means having redundant
data centers and automatic failover. For the user of cloud services, it
means having constant access and guaranteed uptime to applications and
data, without having to worry about recovering from data loss or disaster
recovery.
Why cloud computing is already becoming mainstream
Why do people use cloud computing? The Pew Internet & American Life
Project5 noted several reasons: 51 percent of users who take advantage of
cloud computing do so because it is easy and convenient; 41 percent do
so because of the advantage of being able to access data from any location
and any computer; and 39 percent do so because it promotes easy sharing
of information. The advantages below all point to mainstreaming of the
technology.
 Collaboration
 Scalability
 Better performance
 Reliability
 Simplicity
The last point, simplicity, is perhaps one of the greatest driving forces of
the cloud. Let's face it, there is an element of laziness involved, and that's
okay. Workers everywhere want their jobs to be easier. Cloud computing
provides that. Working at home in the past, may have required a user (or
the user's admin) to pre-load software into the user's home computer, and
install special logins for accessing the corporate server. More often than
not, that burden just led people to inaction, which resulted in fewer
telecommuting opportunities. Cloud computing simplifies the entire
process by removing the need for client software and by abstracting the
data and application servers. Simply put, if it's easy, workers will go for it.
And in the end, that helps the corporation get things done.
Business users, consumers, and software developers alike, ignore cloud
computing at their own peril. Remember when Windows first came out,
and there was still a large contingent of people who insisted on sticking
with the command-line interface? Those who resist the cloud model are in
the same category today. Cloud computing and SaaS is increasingly
impossible to ignore.

5 Pew/Internet & American Life Project, Op. cit.
34

Why? Everything in computing has led to this moment. Web 2.0
technology first gave us a little taste of what true interactivity and
collaboration over the Internet could do for us. While earlier Internet sites
gave us information on static web sites, Web 2.0 raised the bar with blogs,
social networking, instant connectivity, and a new level of interactivity
over the web. Instead of just reading a web site, we could interact with it.
We could send feedback. Take polls. Search for products we like, compare
prices, and see what other people thought. We could hold web
conferences and use things like shared whiteboards. These Web 2.0
innovations put us all in the mindset of free collaboration, unfettered by
physical boundaries. Web 2.0 made it possible for the first time to
hold a productive conference, for example, between people in
Chicago, Delhi, and London. We have gotten accustomed to Web
2.0 innovations and cannot go back to the way it was and we want
more. Cloud computing was the next logical step.
Cloud computing has also gone mainstream because of the
presence of a robust infrastructure. Virtualization technology has
come to the fore, and this too serves a major role in letting vendors
deliver SaaS services and in letting companies gain access to
infrastructure services without large capital expenditures.
What does cloud computing mean to me?
Cloud computing doesn't work unless every stakeholder has something in
it for them. Every party involved can benefit, if it is implemented
correctly, from the end user, the entrepreneur, the CEO who wants to cut
costs, the project manager, IT people, and third party providers.
If you think that cloud computing doesn't affect you, think again. A recent
study by the Pew/Internet and American life Project reported that 69% of
all Internet users6 make use of some sort of cloud computing service and
that number is growing. Do you use one of the free public email
platforms like Hotmail or Gmail? Take advantage of one of the many
online file storage services? Store your vacation photos online? Then you
use cloud computing. The applications go far beyond these three simple

6Pew/Internet & American Life Project. "'Cloud Computing'
takes hold as 69% of all internet users have either stored data
online or used a web-based software application." September,
2008.

35

examples, but the trend is noteworthy. Most people use cloud computing,
even if they have never heard of the term.
These simple consumer applications of cloud computing also highlight an
interesting trend. The most successful technologies are those that have
penetrated both the consumer and the business markets. Cell phones,
once tools of the rich and famous, are now used by everybody and can be
had at any department store for fifteen dollars for a basic model. Social
networking tools started out as consumer-based applications used for fun
and friendship, but are now widely used as vital tools for business
marketing and project collaboration. Now, cloud computing also carries
equal weight in the consumer and business realms. If cloud computing
hasn't touched you yet, chances are, it will in the near future.
SOHO and small business
There has been an interesting trend in software applications, which may
well be ending. That trend is to deliver larger and more feature-rich
productivity applications with every feature that can be imagined. In the
early days of software, this was a good approach, but today, it's rapidly
reaching the status of bloatware. After a point, software in general reaches
a point where it contains everything users need to function; anything else
is just fluff. Yet, traditional software vendors thrive on that fluff to give
them a marketing advantage.
Is it necessary? Not always. Word processing and spreadsheet programs,
for example, contain far more features than most ordinary users take
advantage of. We may even apply the 80/20 rule here as a casual
observation: 80% of users only take advantage of 20% of the features.
The existing office applications delivered on a SaaS basis by Google,
Microsoft, and others, are less feature-rich than shrink-wrapped offerings,
but they contain enough features to be perfectly serviceable by most
users. The advantages of easy maintenance and low cost will drive more
users to adopt it.

VII. Security Concerns with the Cloud
This isn’t to say that cloud computing is perfect. It’s not. Nothing really is.
Industry mover-and-shaker Guy Kawasaki, in his noteworthy "How to
change the world" blog, once wrote, "If a company waits—for example,
the engineers convince management to add more features—until
everything is perfect, it will never ship, and the market will pass it by."
Technology, computer science, and science in general is all about
approaching absolute truth and perfection, but as any scientist will tell

36

you, it's an elusive goal. That said, the cloud has eliminated a lot of the
common worries about achieving perfection, or at least close-to-
perfection, in the technology industry, by providing an environment
where much of the work is done for you and the worries about perfection
are passed on to someone else.
Early on in the cloud movement, one of those worries involved security
challenges, and the first cloud models did indeed suffer on that front.
Many software-as-a-service products were built for functionality and a
rapid go-to-market strategy, with security often bolted on later as an
afterthought. Those vendors that did consider security did so with
disparate security approaches, raising critical questions about where data
was hosted, concerns over international privacy laws, exposure of data to
foreign entities, nonstandard authentication, and leaks in multi-tenant
architectures.
The term “multi-tenancy” is a much talked about topic of debate these
days with regard to cloud computing. It refers to a software architectural
principle whereby several clients (users or groups of users) share a single
instance of cloud computing software. The architectural problem being
solved is that many users (and groups of users) often share a single
instance of software applications on the cloud. If you are creating a
ground-up software program using traditional development approaches,
this often presents a programmer’s nightmare. For example, imagine a
simple healthcare application with a financial team and a cardiology team.
How do you program the application so that cardiology users cannot see
financial data, nor can the financial team members see cardiology data?
Now, imagine that some cardiologists can see different data than other
cardiologists; imagine that managers need to search across both teams.
What happens if someone writes a report that exposes unauthorized
information? This is an extremely simplistic example. In the real world,
there are hundreds of scenarios that must be considered for every
application to ensure that data is never wrongfully exposed.
As the cloud matures, it is becoming more secure than its SOA
predecessor. For example, emerging cloud platforms are making it easier
than ever to implement multi-tenant architectures in software systems.
The cloud isn’t perfect (nothing ever is), but market forces have
demanded that it be secure. Because the cloud has been so widely
accepted, it would simply be impossible for cloud providers to exist were
they not paying extremely close attention to security. Cloud data centers
are some of the most secure facilities in the world, from the physical layer
on up. Service level agreements are in place. Competitive realities dictate
that if a cloud provider doesn't provide a secure environment, they simply

37

won't stay in business because there are thousands of others that do. Even
more critical on the cloud security front is the emergence of platform-as-
a-service, which in its true form builds in robust security from the ground
up. SaaS applications are created with security in mind from the very
beginning—rather than at the end.
There are objections, to be sure. There are objections to any disruptive
technology. People resisted graphical operating systems. They resisted
cellular phones, the Internet, even computers as a whole, but each of
these disruptive technologies won out in the end and our lives are better
for it.

VIII. Cloud Strategy
First of all, the question "Should we move to the cloud?" is not the right
question to be asking. If you are a large enterprise and don't have a formal
cloud strategy, chances are you have an informal one, but just don't know
it yet. It is likely that multiple departments, workgroups, and internal
fiefdoms have already started experimenting with individual cloud
projects. Your competitors are already moving to the cloud, and your
partners and customers expect you to be there. The question is not
whether to move to the cloud, or even when to move to the cloud. The
question is how to do it.
A growing expectation
The benefits of moving to the cloud are indisputable. But aside from the
list of hard benefits, one of the biggest "soft" benefits of a cloud
migration is simply meeting the expectations of employees, stockholders,
partners and customers.
Chances are, those who do business with you already have a cloud
strategy,and that strategy involves not only using cloud internally, but also
interacting with partners via the cloud and taking advantage of third party
cloud services. If you're not part of the program, you risk being left
behind.
Just what are those expectations? In section two, we discussed how
employees, partners, stockholders, and customers don't necessarily see the
underlying technology of the cloud—what they see is what it can do for
them. To those constituents, that is what the cloud really is—not the
technology, but the value. These are the expectations: A better economic
model with superior pricing, faster delivery of services, and scalability. A
big part and most immediate aspect of a cloud strategy is considering

38

how your implementation aligns with the expectations of others with
whom you do business.
From a longer-term perspective, cloud strategy has to take into account
the disruptive nature of the cloud how this will change the business
culture, as well as the operating model. Along those lines, strategy has to
accommodate the inevitable resistance that will be met, how to ease the
transition, and encourage buy-in at all levels.
The personnel side
According to Robert Half Technology, IT professionals and CIOs worry
about in order: Understaffing, talent poaching and retention, data security,
technology upgrades, and keeping up with innovation. Most of those
concerns have to do with people, and how difficult and costly it may be to
retain them.
Talent retention is a major pain point for any organization, and the cloud
shifts this burden away from the enterprise and towards the cloud
provider. Having the right people—and knowing that your cloud
providers have the right people as well—is a big part of your strategy.

IX. Market Disruption
"That's all right. These things gotta happen every five years or so, ten years. Helps to
get rid of the bad blood. Been ten years since the last one."
--Quote from Clemenza in "The Godfather"
The cloud is highly disruptive for the computing industry.
These recent evolutions have made it possible for businesses to consume
virtual machines on mass scales as a viable alternative to buying and
managing expensive servers. As a result, the business of selling servers is
being turned upside down. A harsh truth is that the cloud is uprooting
some of the biggest tech companies in the world: software development
companies, PC manufacturers, and data centers just to name a few.
Meanwhile, the cloud is creating enormous new opportunities for
forward-thinking individuals and businesses to enter new markets and do
a lot more with less. We are beginning to see hundreds of billions of
dollars in impact to tech companies.
Such a disruptive shift doesn't happen just because someone invents a
new piece of technology. The disruption happens not because of the
hardware and software, but because the cloud changes how people do
business with major implications across all industries.

39

Here are just a few examples:
 Cloud has created hundreds of new business categories. For
example, the mobile app market has created thousands of small,
entrepreneurial startups, which market their apps on cloud-based
app stores for customers to find, download, and sometimes use
over the cloud.
 The creation of those new business categories is one of a handful
of triggers that is already creating a new wave of innovation and
entrepreneurship that we haven't seen since the dotcom boom of
the nineties. This current "boom", which has already begun, will
be longer lasting than the earlier one, and will yield positive
changes that will forever change how people work, live, and play.
 The older model of pre-packaged, shrink-wrapped apps is nearly
obsolete. Software today is moving rapidly towards a universal
model of cloud deployment. Instead of buying a set of CDs for
your desktop productivity apps, you will have a cloud
subscription. This is already happening among the major
providers such as Microsoft, with a cloud-based Microsoft Office
package.
 Complicated, on-premise PBXes that require in-house telephony
expertise are also rapidly going out of style, in favor of cloud-
based VoIP systems with a virtual PBX model.
Remember when we moved from host computers to PCs? Now, cloud
computing is shifting that computing power back to hosts again. Only this
time things are different because those hosts have become abstract and
are scattered all over the Internet… all over the world. That is to say,
computing power is being shifted to the “cloud”. Such a shift to cloud
computing would not have been possible until now because the enabling
technology did not yet exist. Broadband connectivity now makes cloud
computing a realistic possibility for not just larger companies, but for
small businesses, SOHO operations, and individual consumers. These
users now have the fat pipes they need to access the cloud, and they have
access now to applications and services that they couldn't begin to access,
or afford, just a few years ago. The possibilities are growing even faster as
the US government undertakes its rural broadband initiatives, which in
turn will push the potential of the cloud further to the masses.
An even greater impact in the emergence of cloud computing may be that
it inspires a new wave of entrepreneurship. Nowadays, thanks to the
cloud, nearly anyone can launch a genuine global business for mere

40

pocket change. Venture funding has given way to back-pocket funding,
and startup entrepreneurs no longer need $100,000 to hire a system
administrator, or to buy new business software and servers. Today’s
emerging entrepreneurs can do everything over the Internet, and without
the burden of huge up-front capital expenditures. With cloud computing,
they can do more than collaborate. They can participate.
Businesses don't thrive in an environment of stagnancy. Standard business
advice in the '50’s may well have been, "Find what works and stick with
it", but even during the Eisenhower years, it was bad advice.
Good business requires evolution and occasionally revolution.
Suppose, for example, you're using a particular piece of enterprise
technology. It works, it gets the job done, and everybody in the company
knows how to use it. There will always be resistance to change, and when
your CIO comes in and says, "Let's put in this new thing" the greatest and
most immediate reaction will be, "But what we have now already works!
Why mess with success?" Simply put, the answer is that your competition
is using something that works better, and they're about to eat your lunch.
Sometimes, moving to the Next Big Thing isn't even a matter of getting a
competitive edge over the other guy, it's just a matter of not falling behind
the curve.
Every now and then, a major market disruption comes to town, and as
Clemenza from The Godfather would have said, it's gotta happen. In the
case of technology, it's been nearly 25 years.
Today, the market disruption isn't happening on just one front—it's a
perfect storm of multiple things. Mega-billion dollar giants are struggling
to retain and reinvent their identities. Emerging startups are starting to
nibble at their jugulars.
Market disruption is happening at the highest levels. Just in the past year,
we've seen CEOs and other members of the executive suite fired, hired
from other companies, and fired again. Companies have been sued, major
lawsuits are going on, and stocks skyrocketed, and subsequently,
plummeted. There is bad blood between the biggest and the baddest.
Within HP, for example, their botched $11.1 billion acquisition of
Autonomy caused some major unrest, and naturally, some people got
whacked, including both top executives on the deal—former CEO Leo
Apotheker and CTO Shane Robison. Shareholders have naturally filed
suit against the board, current CEO Meg Whitman, and audit firms
Deloitte and KPMG.

41

Everybody on top of the corporate ladder understands very well that
being on top today is by no means any guarantee of being on top
tomorrow. For every industry leader, there is a corresponding hole in the
desert and a shovel waiting. At one time, BlackBerry was the undisputed
king, only to be displaced by Android and iPhone. Facebook came along
and quickly gunned down MySpace.
It's not just companies that come and go, it's technologies too—
particularly when those technologies underlie major changes in the
business model. PaaS is inherently disruptive to the business models of
some of the biggest software vendors and system integrators in the world.
To understand the future of PaaS, we must understand the historical
dynamic of disruptive innovation. Small companies are typically the best
innovators. Big companies rarely innovate well. Their R&D projects
typically face scrutiny by their revenue-producing peers who still make
money from the old way of doing things and who fund the seemingly
risky ventures. That’s why big companies watch small companies closely,
and then buy them when they’re successful. This business model has
become the most popular way for today’s big companies to outsource
their innovation without funding internal R&D projects.
This is exactly what we’re seeing with PaaS: The game changing
innovation is happening within the small business community. That’s
because small businesses have everything to gain and nothing to lose by
making things better; and PaaS (along with the rest of the cloud) gives
those small businesses the opportunity to do things that simply would not
have been possible otherwise. To see real innovation, look at companies
and initiatives like Virtual Global, WorkXpress, and Heroku. The latter
was purchased by Salesforce.com in 2010, again demonstrating that
history repeats itself. One thing for sure is that the market leaders today
will not be the leaders tomorrow. This is a true shakeup.
In looking at trends of cloud adoption, we're already seeing a disruption
on a massive scale that rivals the invention of the printing press. The
impact on people and business is likely a heyday for modern psychologists
and authors of great works like Malcolm Gladwell’s Tipping Point, Clayton
M. Christensen’s Innovator’s Dilemma, and Geoffrey A. Moore’s Crossing the
Chasm.
In 2010, we saw the first signs of early adoption. In 2011, we saw hype,
fear, excitement, and panic at the same time! In US Government, we saw
IT Buyers waiting for the cloud movement to blow over, followed by
reality setting in quickly. In 2012, we are seeing intense planning and
more early adoption.

42

The low hanging fruit of services like virtual machines, cloud storage,
collaboration, and cloud email has already become commoditized. Despite
this, cloud is nowhere near the end of its cycle. There is another market
getting ready to hit the cloud: custom software.
So far, there hasn't been very much discussion about custom software for
the cloud, though as PaaS offerings mature and provide for a greater and
easier level of customizability in development of cloud apps, this is already
changing. Software developers and integrators will turn their attentions
towards more customized cloud offerings in the coming year.
This will be a disruption on top of a disruption. Custom apps have always
been an expensive proposition with rollout times of often a year or more,
and this model is coming to an end with PaaS. More innovative,
customized SaaS offerings will emerge at a fraction of the price. In
2013,we will see tens of thousands of custom software applications
moved to the cloud, and PaaS is the doorway.
In the interest of fun, here are a few historical examples of highly
disruptive technologies:

Innovation Disrupted market

Telephones Telegraphy

Automobiles Trains

Light bulbs Gas lighting

GPS Paper maps

Wikipedia Encyclopedias

Personal
Mainframes
computers

IaaS Traditional servers

Traditional software
PaaS development and desktop
operating systems

SaaS Software in a box

43

X. Misconceptions
Now that we've decided what it is, let's look at how cloud computing has
become misunderstood. The very word “cloud” gives way to many fuzzy
definitions. In reality, cloud computing is just as solid and reliable as any
other type of computing; the technology just refers to a mechanism to
connect infrastructure, applications, and platforms over a remote network,
typically on virtualized off-site servers, over a secure IP connection.
With that in mind, let’s look at what the cloud is . . . and what it isn’t.

Top Ten Misconceptions about Cloud Computing
1. The cloud is just a return to centralized computing.
The old days of dumb terminals connected to a centralized mainframe
limited our computing power to one provider. With cloud computing, we
can access the computing power of millions of providers from anywhere
at any time, and for a fraction of the cost of host computing.
In the pre-desktop days, computing, applications, and data storage was
centralized. People used dumb terminals attached to mainframes. The
terminals themselves didn't hold the applications or data. Cloud
computing does have that in common with the old mainframe model, in
that the individual endpoints, in this case PCs or laptops, also do not hold
applications or data. However, there are some crucial differences. The
dumb terminal used in centralized mainframe computing had no
processing power; the PC does. Centralized mainframe computing
connects to a central computer and storage device and cloud computing
may connect to several computers and storage devices in a "virtualized"
fashion.
In addition, the level of access is much broader. The centralized model
required you to log in from one of the dumb terminals on the network.
Today, users enjoy the unique ability to log onto their applications from
any location, anywhere in the world, from a wide variety of devices,
including desktops, laptops, or even smartphones to access applications
"in the cloud" and data that may reside in a remote data center.
2. The cloud is not secure.
Truth-be-told, in-house systems are often less secure because they use
unproven home-grown security models. Cloud applications developed
with cloud platforms use a common security model, which lends
additional security from the ground up; cloud providers will often pay
more attention to issues such as physical security and access controls.

44

In fact, the cloud does have several security advantages. According to
NIST, these cloud computing security advantages include:
 Shifting public data to an external cloud reduces the exposure of
the internal sensitive data
 Cloud homogeneity makes security auditing/testing simpler
 Clouds enable automated security management
 Redundancy / Disaster Recovery
All four points are well taken. Cloud providers naturally tend to include
rigorous cloud computing security as part of their business models, often
more than an individual user would do. In this respect, it's not just a
matter of cloud computing providers deploying better security, the point
is rather, that they deploy the precautions that individual companies
should, but often don't.
3. The cloud isn’t ready for enterprise users.
Enterprise software need not cost millions of dollars, or take years to
implement. CIOs are increasingly demanding more affordable
alternatives. Some of today’s popular cloud systems host tens of millions
of users. The biggest concerns of enterprises rolling out mission-critical
apps are flexibility, scalability, and availability. The cloud has resolved
those concerns.
The very term “enterprise software” is yet another one of those fuzzy
techie terms that is usually glossed over. Therefore, let’s break tradition
and start with a working explanation: As the name implies, an enterprise
software system is one that is engineered in such a way as to
accommodate a very large, nationally dispersed and/or global business or
organization (an enterprise). As such, enterprise software must be able to
efficiently scale to handle tens of thousands, or even millions of users, and
very large data sets that often exceed several terabytes.
We seem to have accepted as a painful truth that enterprise software
systems are supposed to cost millions of dollars and require years to
implement. Without seemingly batting an eye, Fortune 500 companies
and federal agencies alike are laying out tens of millions of dollars for
enterprise software systems. The federal government alone spends $70
billion a year on IT systems, many of which includes enterprise solutions.
If you’d like to see examples, simply visit fedbizopps.gov any day of the
week and look at some the awards that are being made to IT companies.

45

For good reason, many people still believe that enterprise software should
cost a lot of money and time. After all, as compared to small business
software applications (think of QuickBooks), enterprise software systems
are much more complex, with sophisticated architectures that make it
possible to accommodate such large numbers of users and scalable data
sets. Buyer beware that a lot of software vendors will promote their
solutions as “enterprise-enabled” simply because they use a scalable SQL
backend database engine; however, this is a misuse of the word. A true
enterprise software system is engineered in every respect to accommodate
massive scalability, by accounting for multi-tenancy, front-end scalability,
dynamic provisioning, and backend SQL databases with load balancing
thoughtful indexing. This is not an eBook about enterprise engineering,
so we won’t go into any detail, except to explain by examples of the
problem being solved.
Example #1: What if Facebook bogged down and was too slow after the
first 1000 users had signed up?
Example #2: What if you needed to select from a pick-list of a million
users every time you wanted to send an email to someone?
Example #3: What if the human resource team could look at the finance
team’s data?
Needless to say, all of these things would be problems, and all of these
problems need to be solved. However, the shocking truth is this:
Enterprise software need NOT cost millions of dollars! In many cases,
buyers are unnecessarily investing millions into enterprise software
systems that could have easily been implemented for less than a tenth the
time, risk, and cost as compared to a few years ago. How can this be,
right? Well, cloud platforms are driving down the costs of engineering by
offering robust enterprise architectures as-a-service. At the same time,
cloud infrastructure is driving down the costs of scalable storage and
computing power by providing those things as a service.
Together, cloud platforms and cloud infrastructure are leveling the playing
field. Today, approximately 80% of enterprise software revenues go to
IBM, Oracle, or SAP. What will happen when millions of entrepreneurs
are suddenly able to enter the same software markets with their credit
cards? In a few years, we will soon see.
The appeal to small businesses, SOHO businesses, and individual users is
obvious. This is where the early adopters are for the most part. The most
obvious successes are the office suites delivered over the Web, such as
Microsoft's line of subscription-based software offerings, as well as

46

Google's Google Apps collection. Cloud computing, and the Software-as-
a-Service applications that run on top of it, give these smaller users a
quick and easy way to get up and running, access high-end applications,
and avoid up-front costs that are typically associated with software.
But, when it comes to enterprise businesses—typically defined as
companies with over 1,000 employees—those cost concerns weigh less
heavily than they do for smaller, cash-strapped businesses and individual
consumers. Nonetheless, cloud computing applications are starting to
target the mid-size and enterprise market. Traditional enterprise vendors,
whose installations often run into multiple millions of dollars, are now
starting to offer SaaS versions of Enterprise Resource Planning (ERP)
software and other mission-critical applications.
For these larger companies, although the cost factor isn't the only
consideration, it does nonetheless weigh in. A SaaS implementation of
what would otherwise be a massively expensive project allows these large
companies to move more gradually into ERP software, without having to
implement a full-fledged "forklift" installation all at once. The risks of
failure are therefore much lower. Oracle, SAP, BEA, and other well-
established enterprise software vendors are all experimenting with this
type of implementation.
Large-scale enterprise applications tend to be all encompassing.
Enterprise Resource Planning (ERP) systems attempt to be everything to
everybody within the company, and if it's done right, it succeeds. But, by
trying to do so much, there is a great deal of complexity that comes with
it. It's usually not possible to put in an ERP system "off the shelf." It
requires customization and integration, and that's where the big costs
come in. Beside big costs, the extreme amount of customization can lead
to very long rollout times, errors, or even complete failure of the
implementation.
In fact, although an ERP application is large and complex, the bulk of the
cost is seen in the customized development that each implementation
requires. Here's where cloud computing lends a major advantage to
enterprise users.
As described earlier, cloud computing is made up of three components:
Infrastructure, platform, and software. The infrastructure costs (physical
servers, networking, and connection) are minimized with a cloud project,
since the enterprise company no longer has to host their own servers for
their massive applications. The platform portion of the cloud makes
development easier and more robust; the software portion of the cloud

47

means the enterprise user no longer has to pay for continued maintenance
and upgrades, and still gets to enjoy robust software applications.
4. You lose control with the cloud.
The very word “cloud” implies outsourcing to an unknown vendor;
however, this is a misnomer rather than reality. You can use cloud
technologies internally, or outsource to a well-established vendor who has
been offering reliable service for years. With the cloud, you can gain
more control through a web-based control panel, while letting go of day-
to-day maintenance. “The fallacy of direct control” (see section XI in this
book) posits that it is more efficient to retain control over those things
that matter, while freeing up your time by leaving the details to a third
party expert provider.
5. It’s the same thing as utility computing or grid
computing.
Grid computing was an early predecessor that virtually clustered
computing resources to serve a single purpose. Cloud computing has
matured to serve multiple clients and multiple tasks simultaneously. Utility
computing, by the same token, has evolved. Today’s cloud delivers all
three major elements as a service: architecture, platform, and software.
Utility computing is another term that's been widely used to describe
shared access, but this buzzword also serves to muddy the waters. "Utility
computing" is a term that actually predates "cloud computing", and there
is some debate on whether the two are actually one and the same. In
general, as we have described here, "cloud computing" refers to a broader
set of services (architecture, platform, and software), while "utility
computing" is generally thought of as "as a service" computing, or only
the last of those three elements. Utility computing is not "the cloud", but
it runs on the cloud. In that light, utility computing can be defined as the
same thing as Software-as-a-Service.
One common argument that explains the difference is to say that cloud
computing affords a much greater level of abstraction, while utility
computing allows users to retain a greater degree of control over the
physical infrastructure.
In that regard, is it necessary to retain that control? Do you need to know
where the servers are located and be able to tweak and fine-tune the
applications? For most purposes, usually not. That is the greatest
advantage of cloud computing; the end user need not worry about the
infrastructure, or indeed, even be aware of where it is located.

48

Cloud computing may also be confused with "grid computing", but there
is a fundamental difference . Grid computing is typically thought of as a
collection of resources, such as computer servers, which may be in
different locations, but are virtually clustered together to address a single
problem or serve a single client or single purpose. Cloud computing on
the other hand, while organized in more or less the same way, serves
multiple clients and multiple purposes simultaneously.
6. It’s only for low-end consumer applications.
Cloud computing has gained popularity in many consumer areas, but has
also gained widespread acceptance in business applications, including
productivity suites, online backup and storage, and collaborative
environments; with applications of enterprise-class software already being
delivered over the cloud.
Cloud computing is in fact used in many consumer-facing applications,
such as free email (Hotmail, Gmail), Instant Messaging, and online file
and photo storage. But, at the same time, it has gained widespread
acceptance in many business applications, ranging from productivity
applications (business application suites), to online backup and storage,
and collaborative environments.
Small businesses especially have been early adopters to the cloud model,
since the economic advantage allows these smaller businesses to take
advantage of applications and Software-as-a-Service offerings that would
otherwise be too costly for an on-premises installation.
The advantage has not been lost to larger enterprise users however, and
this represents the next wave of cloud users. Having been proven in the
consumer and the small business realm, the natural progression is to larger
corporate users.
7. It’s too isolated from my other data and applications.
Cloud computing applications are easy to integrate with the rest of the
enterprise and already have several integration tools on the market to
make it happen.
This is another misconception based on early cloud projects that have
long since evolved. In fact, since a great many networks run over an IP
backbone to begin with, cloud computing is natural to integrate with the
rest of the enterprise. Since cloud-based "virtual storage" is rapidly
becoming the standard as well; using an application in the cloud need not
isolate the application and the data from the rest of the business.
Integrating and sharing the data is straightforward.

49

Integrating the cloud applications themselves with existing applications
running within the business however, may be a bit more difficult—but
this too, is being done already. Can you run a cloud-based application that
has hooks into an on-premises application? Sure. While in some cases it
may take some custom integration work, there are already several tools on
the market that specifically meet this need. SaaS integration tools have
already proven to be quite useful and robust in this regard.
8. We won’t need PCs any more with cloud computing.
Cloud computing is a broad concept with many elements, and powerful
desktops are a central part of the cloud model.
"But what about my PC?" you ask. Desktops and laptops continue to
evolve year after year, becoming ever more powerful, able to hold more,
do more, and connect faster. You can run some pretty powerful apps on a
standard desktop computer, so why do we even need to move those apps
somewhere else? Nobody wants to work on a dumb terminal any more,
but don't worry. Nobody's going to take away your PC and replace it with
a dumb terminal.
For many users, desktop computing may be just fine for years to come.
PCs continue to serve a valuable role, having become the basis of a very
large industry. Many desktop computers, desktop operating systems
(which we equate to a "platform" in the cloud computing paradigm), and
desktop applications are quite productive, useful, and robust. Cloud
applications will become another powerful tool in your toolbox, but the
difference between today's cloud computing and yesterday's old "dumb
terminal" model is that the apps are running on a more powerful client,
which has multiple capabilities.
9. Reliability will be a problem.
Virtualization and platform technologies are almost as old as computing
itself. What’s new is the ability to market the capabilities. Cloud
technologies can provide superior reliability with service level guarantees.
10. The cloud will give you performance problems.
Performance is seldom a problem with cloud computing. Latency can be
minimized by selecting a provider with a data center in your own region,
and by reviewing the provider’s upstream carriers and service level
guarantees.

50

XI. The “People Cloud”
Jobs 3.0 and Decentralization of the Workplace
Cloud computing is more than a technology - it's also a game-changing
business process. The reason it has gained so much traction is because of
what it enables for entrepreneurs. A complete discussion of cloud
computing must go beyond the technology that underlies the cloud
process, to include a discussion of the greater question of what drives
cloud computing, and what the social and macroeconomic impact of it
may be.
In this book, we choose the term “people cloud” to illustrate how our
workforce is scattering in many of the same ways that computing
resources have scattered across the cloud. Even more, we are managing
people resources over the Internet in many of the same ways that are
managing computing resources. Because of cloud computing as an
enabling technology, we are seeing an explosion in entrepreneurship and a
decentralization of larger companies.

The workplace has evolved from everyone in one place, to a scattered workplace, and
finally to one in which the physical roof is replaced by a virtual roof. The virtual
workplace is again reconnected.
We've already seen major changes in the workplace. Companies have
embraced a model of decentralization in favor of outsourcing and
offshoring. Web 2.0 technologies have enabled a greater level of
collaboration, which means two things: First, people no longer need to be
physically present at the office and can instead work from home, or
anywhere else in the world. Second, this new level of collaboration allows
companies to collaborate with smaller providers anywhere in the world to
get the job done.
Job 1.0
The 1950's mindset of the corporation as a sort of benevolent father is
obsolete. That older (and short-lived) way of doing business is what we
refer to as "Job 1.0". During that time, we saw the corporation as a
benevolent institution, which looked out after our own well-being. We
had an expectation of a 30-plus year career with a single company, an
opportunity to rise from within the ranks, and a relative amount of job

51

security. The prevailing philosophy was that company growth was
equivalent to the company's apparent physical size. A company during this
time that had 1,000 employees was considered to be more successful than
one with 100 employees. Companies embraced the in-house strategy with
a vengeance, and larger firms did everything from running their own
internal print shops, to hosting their own cafeterias for workers.
This is, in some ways, reminiscent of the old-fashioned "company town",
which went so far as to even provide rental housing for its workers
(usually substandard), a company grocery store, and so forth. In the
process, companies kept employees beholden to the company, and
usually in debt to it.
Job 1.0 came with three illusions: That physical growth of a company's
mass was equivalent to success, that the "everything in-house" business
model was beneficial to employees, and that it created a heightened sense
of job security and loyalty.
Job 2.0
The "dotcom boom" broke down the illusions of Job 1.0. During this
incredible time of entrepreneurism, the notion that a company with 1,000
employees is better and more successful than one with 100, or even one
with 10, started to break down. New technologies allowed companies to
do more with less. One clerk with spreadsheet software could do the work
of ten people in the pre-desktop days.
What's more relevant is that Job 2.0 started to break down the illusion of
a single-company career as being beneficial. The desire for a 30-plus year
career with a single company became less desirable, and employees
became freer to move from job to job in search of greater opportunity.
The dotcom boom ushered in a new era of mobility in the workplace, and
at the same time, made it more acceptable and possible for someone to go
out on their own and start an entrepreneurial venture.
Still, "Job 2.0" operated under the concentrated model of corporate
communities. Silicon Valley flourished, contained an incredible
concentration of talent, and more high-tech companies in one small
region, than anybody could imagine. That's because while the concept of
"job" had evolved, the concept of "company" had not yet shifted.
Many high-tech companies during this time were short-lived, but
nonetheless contributed to the collective wisdom by creating new
technologies that are still used today in the latest iteration of "Job 3.0".
Job 2.0 re-set the tone, breaking down the expectation of a 30-year single-
company career, providing the technology for a dramatic change in how

52

business processes are accomplished, and overcoming the '50s mindset
that prevented people from switching jobs, or leaving a job to go out on
their own entrepreneurial venture.
Job 3.0
So now we come to the latest version of what a job really means. Today,
two factors are driving a permanent shift in employment patterns:
 Modern communications technologies and cloud computing
 High unemployment and a huge recession
These new technologies mean that we now have the technological
wherewithal to move away permanently from the centralized model of
work and employment. Collaboration no longer requires a physical
presence, and this means companies can do more with less. It means that
companies are keen to outsource many of the functions that were once
done in-house. This means, in turn, that those functions are being done
by people working at home, or for small companies, people that specialize
in specific areas.
When we speak of the macroeconomic realities and how they too have
enabled the cloud computing shift, we mean that the recession has made
companies take a long, hard look at how they get things done. Companies
are looking for new ways to become more efficient, and they are looking
for technologies that enable them to do more with less. It's not just a
matter of getting new features or capabilities—it's a matter of economic
survival. The economic downturn transformed cloud computing from a
"nice to have" into a "must have".
The age of the cubicle is over. There are naturally some jobs that must be
maintained on-premises, but increasingly, it is just as feasible to
accomplish many tasks off-site, either through telecommuting or
teleworking arrangements, or outsourcing to a third party provider.
Today, the link between corporate size and corporate success is upside
down. It is possible for a ten-person company to be more successful and
productive than a 1,000 person company. Taking that to the logical
conclusion, the possibility of a successful, one-person company is now
much more realistic than it has ever been.
The notion of working at one's own home has gone through a lot of
iterations over the centuries. In the Middle Ages, it was the standard, as
craftspeople and tradesmen plied their trades out of their own workshops
behind their homes, but the Industrial Age brought us a new normal.
Working outside the home became the standard, and people started to see

53

working at home as less desirable. Today, the pendulum shifts once again,
as new technology makes it possible to conduct business from any place
in the world.
The idea of working from a lounge chair on the beach in a tropical island
isn't that far-fetched. For those who don't have a tropical island handy, at
least working from home is more possible. When you call into any large
company's Customer Service department for example, more often than
not, you are either speaking to someone on the other side of the world, or
someone who is wearing a bathrobe, sitting in their own kitchen with a
laptop and broadband connection.
Ultimately, Job 3.0 has led to the decentralization not only of the
workplace itself, but of the workplace community. We no longer need
Silicon Valley. It is no longer necessary for all those high-tech companies
to be physically present in the same little section of central California.
Silicon Valley has made itself obsolete and what's more fascinating is that
it has made the very idea of what a "company" is obsolete as well.
The end of the company as we know it
Cloud computing technology and outsourcing have an obvious symbiotic
relationship, and one cannot exist without the other in the real world.
Outsourcing becomes much easier and more realistic when there is cloud
computing; and cloud computing becomes much more than just a
theoretical technology when outsourcing functions as a practical
application of it.
What is a job and what is a company? Those questions seem simple to
answer, but the answer isn't always evident. Today, the answers are
changing rapidly. In the last chapter, we talked about what a "job" really is
and what it is becoming; now let's talk about what a "company" is. Sure,
in business school they taught you all about corporate structures, and how
a corporation is an entity unto itself, but that's not what we're talking
about.
A "company" has always been traditionally seen as an entity engaged in
commerce, which has members (owners and employees) that carry out the
tasks related to the company's commercial endeavor. A larger company
has "divisions" of employees, which may carry out tasks such as
accounting, human resources, information technology, customer service,
sales, and marketing. Seen in this way, a company is a very defined sort of
organization that is self-contained. In a limited sense, every company has
some interactions with other companies, as the company will take on

54

suppliers, vendors, customers, and partners, but still it stood on its own as
an island.
A company today, or "Company 2.0", operates a little differently. It is still
an entity engaged in commerce, but it is no longer dependent on its
internal departments and employees to carry out those tasks. Instead, a
company's set of tasks is condensed down to its core mission, with all
others being carried out by other companies. As such, the "corporate
walls" have broken down and collaboration has built up. When a manager
gets his or her weekly reports, they may not come from inside. Customer
service may be taken care of by a company in Mumbai. IT is taken care of
by a managed service provider in San Francisco, and marketing functions
are handled by a handful of small and creative companies that collaborate
with each other even further to accomplish the goals of the main
company.
Web 2.0 technology, outsourcing trends, cloud computing, competitive
pressure, and other macroeconomic realities all have converged to make
these major changes. Is a company with 1,000 employees more successful
than a company with ten employees? The answer is no longer obvious. In
many cases, the company with ten employees may be able to accomplish
the same thing, reach the same sales goals, and carry out the same tasks as
a much larger firm with many more employees.
The Virtual Company
Taking the concepts described in the previous section a step further, we
can easily see the shift that has occurred from a workplace organizational
structure that was several layers deep to one that is leaner in nature, but
incorporates a "cloud" of virtual extensions. In the past, for example, a
hierarchical business would include internal departments for data entry,
payroll, public relations, IT programming, and so forth. In addition, the
same business would retain functions like data storage,
telecommunications, web hosting, and server farms internally.
The inherent inefficiencies of this hierarchical model are obvious.
The boundaries of the actual "company" have become permeable to the
point of being nearly invisible. As a result, we are already seeing the
emergence of the "virtual company." These companies exist in reality
today all over the world. A "virtual company" has no corporate walls at
all. It may be organized to formally have only one or two employees, yet it
may have dozens, or hundreds of people working towards its main
commercial goal. The CEO's office may be a spare room in her house; the
"Marketing Department" is actually a virtual group of creatives working in

55

spare rooms of their own, servicing not only the primary virtual company,
but several others as well. The entire network of people—we can no
longer call them "employees"—are connected in real time through
modern collaborative technology, and the entire IT infrastructure exists in
the cloud. Virtual private networks (VPNs) ensure that every party can
connect to the applications and data they need on a secure basis, from any
location and from any machine. At any given time, the Public Relations
manager may be working out of a Starbucks, the tech support guy is
sitting in his kitchen wearing a headset and nibbling on leftover pizza as
he doles out advice, and the Vice President of Operations is keeping
everything flowing smoothly from a bungalow on the beach in Thailand.
Indeed, it is very possible that most of the members of this "virtual
company" have never even met face-to-face. They don't need to.
Why does a company outsource?
A company engages in outsourcing because it brings cost savings and
efficiencies and because it has the technological framework to do so
efficiently through innovations in cloud computing.
The bigger question is, does it really make a company more cost-effective
and efficient? Since traditionally we think of achieving gains as something
that is done through control, but this is not always correct. Outsourcing
actually serves the broader goal of efficiency by breaking down those
artificial corporate barriers, exposing processes so that they are more
transparent and responsive to the corporate entity, and eliminating
unnecessary layers of corporate bureaucracy.
A highly vertical company, which tends to do all functions in-house, will
out of necessity have enormous layers of bureaucracy. Processes are
bogged down. Reporting may not be responsive enough. Individual
fiefdoms within the corporation may have conflicting goals and may be so
caught up in their own domain that they neglect the greater goal of the
corporate entity. When a company is so large and organized vertically in
this way, it may very easily lose focus and lose its ability to respond to the
market quickly and efficiently. As such, the economic advantage is not the
only advantage—a less integrated company will simply be able to respond
better, maintain its core focus better, and spend its money better.
What makes a good company to begin with? A company that specializes
in something; a company that does something, or produces something,
better than anybody else. When a company starts devoting large amounts
of energy and resources to tasks that are not related to that thing it does
better than everybody else, then that company's energy starts to dissipate.
More importantly, those peripheral tasks aren't being done as well.

56

For example, a company that makes pizza may make the best pizza in
town. They're good at it. That's their "thing". But, there are other things
they're not as good at. Good pizza makers aren't necessarily good
marketers, and so that pizza company outsources the marketing function
to another company, which is very good at what they do.

"I've been outsourced!"
That's become a common cry of the working person in today's world.
And yes, it's real, and it happens every day. A company decides to
outsource a particular function, and the internal staff are let go.
Opponents of outsourcing incorrectly assume that when a job is
outsourced, it is lost; and therefore contributes to the overall rise in
unemployment and contributes to the overall detriment of the economy.
This is not the case. Outsourcing does not necessarily result in job loss. It
results in job shifts. A programmer today, for example, will gain greater job
security and higher pay by working for a programming job shop, rather
than an internal corporate IT department.
The proven benefits of outsourcing are undeniable, and the market reality
is that it is here to stay. The appropriate response is to see it not as a
challenge from a work perspective, but as an opportunity, as indeed it has
proven to be. Opportunities exist as a direct result of outsourcing—
opportunities for employment at outsourcing service bureaus,
opportunities for individuals to work independently, or opportunities to
start their own entrepreneurial service bureaus.
"Jobs for Americans"
Another common battle cry of opponents of outsourcing; this complaint
assumes incorrectly that when a job is outsourced, it is outsourced
overseas. While many tasks do indeed go to India, Vietnam, Russia, and
other third world and emerging nations, plenty of those tasks do stay at
home.
Why they do stay at home is simple supply and demand. Because we are
in the era of Job 3.0, there are more former employees who have moved
towards being independent contractors, freelancers and consultants,
telecommuters, teleworkers, and work-at-homers. The supply of domestic
third-party businesses offering services to other companies that wish to
outsource has increased, making it very easy indeed for companies to take
advantage of outsourcing (through cloud computing technology), while
still keeping jobs within the geographical boundaries they call home.

57

There is no question that some of the tasks go offshore, but that is one
cloud that has a silver lining. Here's why:
India has built much of its economy on IT outsourcing to US and
European companies. Manufacturing is now frequently sourced to
companies in China, and more recently, Vietnam. When this occurs, the
immediate result is the loss of an American job. Those who do not favor
outsourcing because of social and political reasons fail to look beyond that
immediate impact. Yes, there is an immediate job loss, but what is the net
result?
 First, the company doing the outsourcing can cut costs, which is
important to its long-term survival. The company's outsourcing
strategy will help it to remain profitable; and therefore, be able to
stay in business and keep in place the jobs that still exist. In other
words, outsourcing certain processes and jobs allows for the
retention of other jobs.
 The third world or emerging nation raises its own standard and
increases the ranks of its own middle class. That country then
becomes a more viable trading partner for the United States and
Europe. The citizens and workers of those emerging nations,
having raised their own standards of living due in large part to the
presence of outsourcing, gain an appetite for consumer goods—
and more often than not, consumer goods that are made, sold,
and distributed by Western companies. Walk into any shopping
mall in downtown Bangkok, Beijing, or Moscow, and you'll see
aisles full of Western chains, Western brands, and Western
products. Raising the standard of living of emerging nations
therefore creates a positive feedback, giving greater opportunities
for Western companies to supply and trade with emerging
nations, while creating more jobs in the process.
Strategies of isolationism have always been popular with a certain subset
of the citizenry, but they have always been a failure. Is it really necessary
to keep opportunities from third world countries and keep them
impoverished in order to promote our own success and wealth?
Absolutely not. Free trade and outsourcing benefits everybody.
The next wave of collaboration
Cloud computing isn't just about delivering software-as-a-service. Cloud
computing has given us a whole new way of collaborating.
What did collaboration mean in the 1950s? It meant walking over to the
conference room and chatting about a project with your colleagues over

58

coffee. Maybe it meant picking up the phone and calling your supplier to
discuss your needs, or getting on an airplane and flying across the country
for a face-to-face meeting.
But, collaboration means more than talking, it means sharing information
and data. In the 1950s, sharing information and data was a very physical
and labor-intensive process. Shared data may have come in the form of
reams and reams of printed reports that were sent via special courier, but
still, collaborating on a project required a lot of face-to-face interaction.
The only way for two people in different offices to work on a design, a
spreadsheet, or a document together was to sit down in front of the same
document at the same time, in the same room.
The information age has given us a lot of things, but chief among those is
a whole lot more information. That translates to enormous amounts of
data. Now office politics is a strange thing, and people tend to be
protective of their own work areas, their own projects, and their own
information. This tendency, along with a lack of collaborative technology,
led to "data silos", or independent areas of data that were useful for
specific purposes and in specific areas, but were inaccessible outside of a
very narrow scope. The existence of data silos meant that there was a lot
of redundant effort going on in a big company.
Two things have happened to break through that data silo mentality. First,
the sheer explosion of data has made it impractical; and second, cloud
computing and collaborative technology has made it possible to open up
those silos and allow for data sharing to take place. Data that was
previously held in independent silos in corporate fiefdoms throughout the
enterprise must be shared and constantly revised and updated by many
different people in different locations. This sharing isn't possible without
some sort of collaborative technology that exists "in the cloud", which
allows for instant and easy communication regardless of location, easy
sharing of data, and easy collaboration on projects.

Progressive collaboration
As we have progressed from "Job 1.0" to "Job 2.0" and "Job 3.0",
collaborative technology has progressed in the same way. Those ERP
applications we talked about earlier represented some of the first attempts
at large-scale collaboration, and while they did provide for some
collaboration, data sharing, and a unified view of information; the costs
were enormous.

59

Middleware offered another approach to collaboration, although this too
had limitations that were based on each proprietary middleware platform.
Integration afforded through middleware platforms is often limited to
certain applications or data types.
Other types of Web 2.0 collaboration overcame the data formatting
limitations of middleware, and tools like wikis portals or mashups stepped
in to allow for greater access through a common Web-based interface.
Cloud computing moves collaboration a step further, and brings together
the benefits of all three. Like Web 2.0 mechanisms, access can be nearly
universal. Participants can collaborate from virtually anywhere, and
depending on the cloud application and the interface, may not even have
to have any special client-side software installed. Like middleware, cloud
computing creates an environment where applications and data can be
more easily integrated. Like ERP applications, it can be used to present a
unified view of information.
The New Openness
The reality of corporate life, for anybody that has ever worked in an
office, is that any large company tends to develop "islands" over time.
Fiefdoms. Independent areas of domain into which others dare not tread.
As corporate people, we tend to be protective over our own areas of
work. Although this seems to be human nature, it is counter-productive to
the corporate goal.
This tendency results in a lack of cooperation and results in redundancy
of work. Early on, redundancy was often necessary simply because of
technological limitations. The Operations Department, Customer Service
Department, and Accounting Department all need information on
customers, and all three probably had their own databases of customer
information, which were completely stovepiped and inaccessible to
anyone outside of that specific department. Relational database
technology and simple networking made that model unnecessary, yet it
still exists. Companies that have overcome this mentality operate more
successfully. Simply put, it's often efficient to share data. Of course, all the
usual security precautions, authentications, and authorizations are in place,
but the data gets shared with who needs to see it.
Let's extend that to a broader view. We've seen that companies are no
longer constrained by physical boundaries, and a company's mission is
better carried out by an interacting subset of many different companies,
individuals, teleworkers, and partners. This too can present an
information bottleneck. Yes, we have networks for sharing information

60

within the corporate boundaries, but what about outside the corporate
boundaries. This too, is starting to melt down. For example, some of the
country's largest retailers have supplier networks that allow vendors to
connect securely and directly into the retailer's inventory database. The
vendor can see when a particular product is low, and trigger a
replenishment order automatically. A cloud-based approach to data and
applications allows for data and applications to be shared whenever
appropriate, with whomever appropriate.

XII. Groundbreaking cloud applications
Cloud computing is a classic "disruptive technology" that is destined to
change long-standing processes across all industries. Two of the most
groundbreaking cloud-based applications that will occur over the next few
years are in healthcare, specifically in electronic health records and
healthcare informatics, as well as in government applications.
Healthcare applications (Health-IT)
The current administration continues to be aggressive in pursuing
healthcare reform. Aside from the issue of universal health insurance
coverage, which has gained the greatest coverage in the media, the
reforms include much more under the hood. Most notably, this means
implementation of electronic health records, and the creation of a
nationwide health care infrastructure that would make it easier for
healthcare providers to share and access patient records.
Part of this plan would create a National Health Information Network
(NHIN), which is a broad, interoperable platform for sharing electronic
health information. The NHIN would connect providers, insurers, and
emergency responders.
According to the Department of Health and Human Services, the
government's health care informatics plan's goals include:7
“Medical information will follow consumers so that they are at the center
of their own care
Consumers will be able to choose physicians and hospitals based on
clinical performance results made available to them
Clinicians will have a patient’s complete medical history, computerized
ordering systems, and electronic reminders

7 US Department of Health and Human Services.

61

Quality incentives will measure performance and drive quality-based
competition in the industry
Public health and bioterrorism surveillance will be seamlessly integrated
into care
Clinical research will be accelerated and post-marketing surveillance will
be expanded.”
The concept of electronic medical records (EMR) and patient health
records (PHR) is one that has long been discussed, and is already in use in
other countries. There is no doubt that it will be part of the current
administration's broad health care reform initiative, and there are
already legislative incentives in place to encourage health care
providers to get with the program. EMR doesn't just mean that the
hospital puts your patient records in their computer—it means a
new level of sharing. This of course, is within the HIPAA
regulation framework and assumes a rigorous level of security, but it
allows for a cloud-based infrastructure to exist for EMR. The
benefits are obvious. A patient's medical records would be available
to any authorized health care provider, anywhere in the country.
You could travel anywhere you want, and your records would go
with you. Any authorized provider could access your records in case
of an emergency. Already, there is a common but limited version,
which has shown great benefit—many of the large drugstore chains
keep customer records in a secure database, so that you can go to
any branch, anywhere in the country, and receive your prescription.
The database also includes relevant information such as drug
interactions and allergies. This is only the tip of the iceberg.
Ultimately, this limited drugstore application will be integrated with
all other healthcare providers. What's the result? It could save lives.
Error rates would be reduced, and caregivers will have more
information at their disposal when making critical decisions about
your care.
This technology, based strongly in cloud computing technologies, is
rapidly gaining momentum. The RAND Corporation, in testimony
presented to the Senate Finance Committee, highlighted just a few
of the potential benefits of a cloud-based healthcare IT (HIT)
system rolled out on a national scale:

62

"The hope of many is that the broad adoption of HIT systems with
the aforementioned functionality in the United States will transform
health care in terms of making it more efficient and effective
simultaneously. Efficiency would be enhanced by reduced test
duplication, improved drug utilization, better scheduling, reduced
paper record handling, and improved claims processing and billing.
Effectiveness would be enhanced by reduced errors (reduced
handwriting-based errors, for example), reminders to improve
preventative care, decision support for better evidence-based
practice, improved management of chronic illness, and improved
continuity of care for those patients seeking care away from their
primary provider (such as was needed to support the mass
evacuation that occurred after Hurricane Katrina). Effectiveness
would also be enhanced by the quality of care assessment such
systems would make possible and by improvements in the evidence
base for best practices derived from the analysis of large electronic
medical record databases."8
The potential social benefit is clear. Not only would individual
hospitals benefit by moving to a more technology-based patient
record system, society as a whole would benefit by integrating those
systems together in a national database that relies on secure cloud
computing technologies.
Besides the advantage of better patient care, cost savings would be
enormous. In this day and age, when the health care debate is often
framed in terms of dollars and cents, a cloud-based national patient
record system is an obvious element that should be included.
RAND Corporation claims savings that could be achieved would
reach $80 billion per year, assuming a 90 percent adoption rate by
hospitals and physicians. To put that figure in perspective, it's a full
four percent of the $2 trillion spent annually on health care in the
United States.

8RAND Corporation. "The potential benefits and costs of
increased adoption of health information technology." Richard
Hillestad, July, 2008. Testimony presented before the Senate
Finance Committee on July 17, 2008.

63

The benefits can also be seen just looking narrowly at adverse drug
events. Every year, there are errors in medication that result from
lack of allergy or drug interaction warnings, handwriting errors, and
poor dosage monitoring. The RAND study further estimates that
the safety benefits would be enormous, avoiding as many as 2.2
million such adverse drug events per year, saving nearly $4 billion
per year.
Could these savings be achieved without a cloud computing
infrastructure? Not likely. The great benefits illustrated by RAND
cannot be achieved if those electronic records are stove piped,
retained only by each individual provider. The cloud-based
infrastructure suggested by the DHHS brings the scope of the
proposal into greater perspective, delivering the benefit on a much
wider scale, and allowing for the greater level of benefits that result
only from data sharing to exist.
Government: NASA and Nebula
As much as we all like to complain about government inefficiency,
hulking bureaucracies, and outdated procedures, there are a few areas
where government really does excel in setting the standard for the rest of
the country. In the area of public access to documents over the Internet
for example, the Feds have done quite well. It's no longer necessary to
drive downtown to a government office, or make a phone call and wait a
week for a clerk to mail a form to us—we can just download it over the
Internet. Even state Departments of Motor Vehicles—well known and
maligned for hour-long waits and grumpy employees—have gotten onto
the bandwagon of technology, and in most states it's now possible to
renew your license plate online or at an automated machine in the office
lobby. And once again, it may well be the government that sets the pace
for embracing the cloud computing model.
It's not surprising that the biggest cloud project in government comes
from NASA, an agency that always tends to be out in front of the pack
with new technology. A cloud computing pilot called Nebula, being
developed at the NASA Ames Research Center, "integrates a set of open-
source components into a seamless, self-service platform, providing high-
capacity computing, storage and network connectivity using a virtualized,
scalable approach to achieve cost and energy efficiencies." NASA says
that Nebula provides for rapid development of applications that are both
policy-compliant and secure, promotes collaboration, and encourages
reuse of code.

64

Nebula is a wonderful example of cloud computing done right. It is open
source, which means it is transparent and highly interoperable. It is a full,
true cloud system that incorporates infrastructure, platform, and software,
all three of the main components of cloud computing. Nebula is already
in use for educational and public outreach uses, collaboration, and
mission support. Amateur astronomers use it to upload high-resolution
photographs, and the LCROSS participation site, where amateur
astronomers work with NASA scientists to get a better view of the moon,
is built on the Nebula platform.
Two useful elements of Nebula elegantly illustrate the benefits of cloud
computing. It automatically increases computing power and storage as the
web application needs it. This is a central benefit of cloud computing in
general—the user need not worry about compute power and storage,
since that's all automatically and transparently taken care of on the back
end. When more storage space is needed, it's allocated. If more compute
power is needed, you get it transparently. Second, Nebula addresses the
security worry. It was built to be secure, as well as compliant with
government policies (of which there are many).
The Federal CIO Vivek Kundra, who was formerly the District of
Columbia technology chief, has been a strong proponent of the use of
cloud computing in government as a way to gain efficiency and save
taxpayer dollars. In Washington DC, he was able to eliminate a $4 million
initiative to create an intranet for the DC government, and instead, shift
the district government to Google Apps—accomplishing the same goal,
and saving a huge amount of money. Kundra's move to the cloud enabled
DC to save money and improve efficiency. For example, now the district's
training information can be obtained through online videos on Google
Apps; the same Google Apps is also used to add more transparency to
government by making procurement data available to the public. Kundra
claims, "The cloud will do for government what the Internet did in the
'90s. It's a fundamental change to the way our government operates by
moving to the cloud. Rather than owning the infrastructure, we can save
millions."9

9Gautham Nagesh. "Local technology czar could be headed to
Obama administration." Nextgov, November 26, 2008.

65

XIII. The Open Cloud
Open source software has in general been on the rise, and there's no
doubt that it delivers many benefits to developers and end users alike.
There are numerous cloud computing services that are either written
entirely in open source code, or at least incorporating open source into
the final application. Two of the biggest advantages of open source are
lower cost, and greater flexibility. These benefits fit well into the entire
cloud computing paradigm, which delivers the same. Open source in
short, enhances cloud computing's promise to deliver greater cost savings
and flexibility to those who use it.
It does this through two means: First, by streamlining the development
end by allowing developers the use of existing open source code rather
than "reinventing the wheel." This model correlates closely with the use of
a cloud platform, which also allows cloud applications developers to build
applications on top of an existing application infrastructure, so that
routine functions need not be built from scratch. Doing so not only
provides an advantage in terms of reduced development cost, it also has
the advantage of allowing the developer to access code that has already
been proven. In this respect, cloud computing applications, particularly
cloud computing applications that have been built using open source
components, are more likely to be robust and possess fewer flaws than
applications built entirely from scratch from the ground up. Second, the
open source paradigm answers the question, "What happens to my cloud
application if the provider goes out of business?" If cloud-based
applications are based on open software models, then if and when a cloud
provider goes out of business, an individual client could easily take over
their own applications if necessary, or transfer them to another provider.
Is open software free software?
Here's a quick answer to that question: No. Casual observers often
confuse these two very separate software movements. The "free software"
movement is an ideological platform that suggests that all software should
be free, and it is only practical in a very limited sense. The "open source"
movement is a technological platform that espouses open development
because it allows for advantages such as continuous improvement of the
code base, and easier customization for individual users. The latter is
practical in almost all cases of software development.
Here's a useful description from GNU:
"The two terms describe almost the same category of software, but they
stand for views based on fundamentally different values. Open source is a

66

development methodology; free software is a social movement. For the
free software movement, free software is an ethical imperative because
only free software respects the users' freedom. By contrast, the
philosophy of open source considers issues in terms of how to make
software 'better'—in a practical sense only. It says that non-free software
is an inferior solution to the practical problem at hand. For the free
software movement, however; non-free software is a social problem, and
the solution is to stop using it and move to free software."10
This is an interesting concept and fascinating fodder for ivory tower
discussions in university seminar rooms and coffee houses, but ultimately
an impractical one. Ultimately, it is true that open software is indeed a
methodology used to make software technologically superior, but the
above argument of course eliminates incentive to make the software in the
first place.
Open source delivers the advantages of:
 Flexibility to adapt and customize software to suit individual
needs
 Lower cost of development
 More robust development due to continuous revision
Free software, on the other hand, constrains development by limiting that
development only to academics, hobbyists, and people with too much
time on their hands. This would eliminate an entire class of development
professionals who create software for a living.
We advocate the use of open source in cloud-based platforms, as well as
cloud-based applications for the above advantages, and most importantly,
to overcome the potential drawback of cloud providers going out of
business and leaving proprietary applications, which companies may have
come to depend on, inaccessible.

XIV. Security and risks
Security is incredibly important in today's environment. Cyber-attackers
and other types of black hat folk want to infiltrate your network, often for
personal gain, and the losses every year due to cyber-attacks are

10Richard Stallman. "Why open source misses the point of free
software." Free Software Foundation.

67

enormous. We take great measures to protect our data and our networks
with firewalls, anti-virus and anti-malware software, physical protections
such as locked data centers, and sophisticated authentication and
authorization techniques.
Any good IT security manager is paranoid, and the belief that "everybody
is out to get me" is one that serves the IT security mission well. "Trust no
one" is the watchword. The poor IT security manager is as a result often
resented by end users, who must comply with regular password changes,
policy items that may be annoying or inconvenient, and procedures that
may make access more difficult. The payoff isn't always obvious; since the
most ideal outcome for the security manager is that "nothing happens."
It is only by looking at what happens to other people and statistics related
to loss and frequency of attack, that we realize that the security
investment is a good one. The 2008 CSI Computer Crime and Security
Survey shows that there is an average reported annual cost of nearly half a
million dollars for financial fraud, $350,000 for dealing with "bot"
computers in the network; and an overall average annual loss of just under
$300,000. Twenty-seven percent of respondents said they had detected at
least one targeted attack.11
It's interesting to note though, that the security issue has its own cloud-
based solution that is growing in popularity. Security is increasingly
delivered as a managed service by a third party provider, a factor that
gives weight to the relevance of cloud computing and "as a service"
offerings in respect to the security question.
There are several obvious reasons why security is being delivered, quite
successfully, on an outsourced basis through the cloud. Like many other
types of services that are delivered over the cloud, security is a specialized
field. Many smaller companies especially lack the high-end expertise
required to run security in-house, and having access to the best security
experts in the business from a third-party provider will afford those
companies better security, more expertise and knowledge, and access to
higher-end security applications and equipment than they could provide
on their own.
What happens when data and applications are put into the cloud? Do we
lose control over the security precautions? What happens to security?
These are fair questions that must be addressed. The word "cloud"
implies, by its very nature, that the exact physical location of data and

11 2008 CSI Computer Crime and Security Survey.

68

applications may not even be known. The abstraction provided by the
virtualization technology used by cloud providers makes physical location
even harder to pin down.
Security improvement through common security models
in the cloud platform
One unnecessary limitation to cloud computing is that at present, cloud
application providers tend to implement their own proprietary security
approaches. This gives rise to a number of concerns and questions
concerning international privacy laws, exposure of data to foreign entities,
stovepipe approaches to authentication and role-based access, and "leaks"
in multi-tenant architectures. These security concerns have slowed the
adoption of cloud computing technology, although it need not pose a
problem.
The very nature of a cloud platform is that it imposes an instance of
common software elements that can be used by developers to "bolt on"
to their applications without having to write them from scratch. This
advantage is especially useful in the area of security. The cloud "platform
as a service" brings an elegant solution to the security problem by
implementing a standard security model to manage user authentication
and authorization, role-based access, secure storage, multi-tenancy, and
privacy policies. Consequently, any SaaS application that runs on the
common platform would immediately benefit from the platform's
standardized and robust security model.
Understanding how operating systems work can provide a good point of
reference. An operating system, such as Windows, OS X, or UNIX, has
security features built in. The operating system vendor constantly makes
refinements to their own security models, which are issued in regular
updates and patches. Individual applications therefore, need not worry
about addressing those security issues. Now of course, relying solely on
the operating system's security features is poor practice; most users will
add in firewalls, anti-virus software, encryption, or one or more methods
of authentication and authorization, but, the basics are already taken care
of.
Cloud platforms work the same way. When a developer builds a cloud
application on top of a cloud platform, they benefit from the platform's
existing common security model.
Cloud computing provides superior physical security
This may seem to be a counter-intuitive argument. Yet lack of physical
security is the cause of an enormous amount of loss. In a GAO report on

69

NASA cyber-security, it was noted that a stolen laptop had data on it that
was subject to arms traffic regulations.12 This is by no means an isolated
incident, and stolen laptops are a surprisingly common source of data loss.
It may surprise the reader to know that loss is often the result of an
internal attack. Yes, insiders account for much of the losses that occur.
Industrial espionage is a reality. And while the specter of black hats
hacking into your network from a third world country is very much real,
very often, the "black hat" is in reality a trusted employee. It's the guy
from the Accounting department who you have lunch with. It's the lady
who brings you coffee in the morning and always remembers that you like
two sugars. It's the recent college grad with so much potential, who did
such a great job on that last report.
A survey by The Strategic Counsel sponsored by CA actually showed that
internal threats are a bigger threat than attacks from outside.13 The survey
showed that internal security breaches are increasing, even while other
threats are decreasing. Forty-four percent of respondents indicated that
internal breaches were a major challenge.
Of course, insiders can attack your network and data regardless of where
it is located, given enough incentive and information, but physical
proximity of the actual hardware and data makes it much easier to gain
access.
There are several security advantages to the cloud. NIST defines these
advantages as follows:
 Shifting public data to an external cloud reduces the exposure of
the internal sensitive data
 Cloud homogeneity makes security auditing/testing simpler
 Clouds enable automated security management
 Redundancy / Disaster Recovery
All four points are well taken. Cloud providers naturally tend to include
rigorous security as part of their business models, often more than an

Michael Cooney. "NASA network security torched."
12

NetworkWorld, October 15, 2009.

13"Survey: Internal security threats outpace attacks from external
sources." Contingency Planning, July 17, 2008.

70

individual user would do. Most IT directors and CIOs understand that
disaster recovery should be a part of their environment, but still, other
things get in the way and disaster planning gets put aside. Or worse, a
company implements a minimal disaster plan policy, which is then shelved
and forgotten until it is out of date.
According to a recent British survey, 52 percent of organizations had
specific business continuity plans. Despite increasing awareness,
companies remain complacent about it, and only 64 percent of managers
say that business continuity is seen as important to their organizations.14
While many businesses neglect business continuity and disaster planning
on their own, they surely expect it to be provided when a third party is
handling their data, and cloud providers as a rule will incorporate disaster
planning far more than an individual company will, also including
specifics of it in their service level agreements.
In this respect, it's not just a matter of cloud computing providers
deploying better security, the point is rather, that they deploy the
precautions that individual companies should, but often don't.
The fallacy of direct control
In looking at the security of cloud computing, it is necessary to look at the
alternative, and the inherent risks of non-cloud computing. It's a natural
human tendency to want to have control over everything. If somebody
else is controlling something; we want to look over their shoulders while
they're doing it. When the plumber comes to fix our pipes, do we sit in
the living room and watch television and let the poor guy just do his job?
No, of course not. We follow him to the bathroom and watch.
It doesn't make the pipes get fixed any faster or any better, but by
watching, we have regained that sense of control that we lost when we
called the plumber in the first place. It's almost as if by watching the
plumber, we're doing the work ourselves. We are no longer in control, but
we feel as though we are.
In the case of the busted water pipe for example, is it better to maintain
direct control over the situation? Probably not. Most of us who are not
plumbers have better things to do with our time. While it's true that we

Chartered Management Institute and the Cabinet Office. "A
14

decade of living dangerously: The business continuity
management report of 2009." By Patrick Woodman and Dr.
Vidal Kumar, March, 2009.

71

could probably fix that broken pipe ourselves if we had enough time,
how-to manuals, and pipe dope, the plumber (who's done it many times
before) could probably do it faster and better.
We can see in this simple scenario that there are circumstances where it is
better to not have direct control.
For those who resist using the cloud, the alternative is to remain in control
by running your own data center, your own servers, your own storage
farm, and your own applications. Doing so leaves you in complete
control, but there is an opportunity cost involved.
The guy who fixes his own plumbing suffers an opportunity cost because
he must make multiple trips to the hardware store, invest time and money,
purchase hundreds of dollars' worth of tools, and miss his favorite
television show. Just like the guy who fixes his own plumbing, the
company that insists on hosting absolutely everything in-house is losing
out.

CLOUD IN-HOUSE MODEL
COMPUTING
MODEL
Little or no capital Large up-front capital
investment investment
IT staff free to Requires IT staff to
attend to other attend to servers,
concerns applications, etc.
Service level Nobody to blame but
guarantee yourself
Physical security Extra physical
included security required
Security built into Additional security
cloud platform tools must be
deployed and
maintained
Backup and Must deploy backup
disaster recovery and disaster recovery
included protocols

72

The most immediate advantages of the cloud is the lack of an up-front
capital investment and freeing the internal IT staff to attend to more
pressing concerns. Beyond that, there are advantages that relate directly to
security. A cloud computing service provider will typically offer a service
level guarantee to protect against data loss, outage, failure, and cyber-
attack. Typically, this SLA is backed up by specific terms that lay out
performance levels, as well as penalties that the provider may be liable for
if those levels are not met.
The physical security element is important and often overlooked. All the
firewalls and passwords in the world are useless if somebody in a lab coat
carrying a clipboard can bluff his way into your office and walk out with a
pocketful of thumb drives and the CEO's laptop under his arms. Make
no mistake, this does happen, and often. Industrial espionage is alive and
well. If you are hosting your own data center, is your data safe? Sure, it's
firewalled; but is it in a locked room? Is access to that room regulated,
with entry by keycard only? Probably not. Service providers offering cloud
services, collocation centers, and hosting providers typically adhere to
rigorous physical security protocols to protect against physical theft or
tampering.
Besides physical security, the technical security is of the utmost
importance. Hosting your own servers and applications requires extra
measures. A larger organization may need to deploy dedicated IT staff to
security only. Cloud computing, on the other hand, builds security directly
into the cloud platform. While the company still must maintain in-house
security in any case, the provider ensures that the applications and data are
safe from attack.
Lastly, the issue of disaster recovery is vital, and one that is often ignored.
We may tend to think that simple backup is equivalent to disaster
recovery, but it is not. Disaster recovery calls for redundant, off-site
backup, as well as procedures and technology for recovering data and
applications at a moment's notice in case of disaster. It can be costly—but
a cloud provider will already have these measures in place.
Of course, when considering cloud providers, these considerations are
always a factor, and it should be determined ahead of time that the cloud
provider:
 Offers a detailed, specific SLA
 Offers physical security at their data center
 Offers superior technical security to protect data and applications

73

 Offers a detailed backup and disaster recovery plan
With these things taken into account, it becomes very evident that
maintaining direct control over everything comes at a high cost, and in
most cases, those necessary elements are not met.
Alternative Delivery Models
Many customers remain concerned about control for good reasons.
Consider national defense data, which must be carefully guarded with the
absolute greatest security measures. You may think that national defense
data isn't a good candidate for cloud computing, but it could be used
effectively here too. It’s actually possible to employ cloud computing
technologies, such as virtualization and platforms on internal networks
within a firewall. In doing so, customers can enjoy many of the benefits of
cloud computing while minimizing the risks. Along those lines, NIST has
suggested the term “public cloud” for a traditional Internet cloud, and
suggests three alternative deployment models described below:
 Private cloud (enterprise owned or managed) - Private cloud
(also called internal cloud or corporate cloud) is a term for an in-
house cloud computing architecture that provides services to a
limited number of people behind a corporate firewall. Private
clouds are marketed to organizations that want more control over
their data than they can get by using a third-party hosted service
such as Amazon's Elastic Compute Cloud (EC2) or Simple
Storage Service (S3).
 Community cloud (shared infrastructure for specific
community) – Similar to a private cloud, except shared by
several organizations with overlapping concerns. For example,
NASA’s Nebula offering may be considered as a federal
community cloud.
 Hybrid cloud (composition of two or more clouds) – Refers
to a cloud infrastructure that consists of two or more clouds
(private, community, or public) that remain unique entities but
are bound together by standardized or proprietary technology
that enables data and application portability (e.g., cloud bursting
for load-balancing between clouds).
Accessing the cloud
Making the move from on-premises applications and data storage to
cloud-based applications and data storage is a big step, and one of the first
things that come to mind is access. "Can I get to my data and apps?" "Will

74

there be latency and delay?" These are important questions. In the old
days of centralized mainframe computing, and in the early days of
networking, latency was common. Data entry people suffered productivity
loss because they had to wait for processing on the back end before they
could enter more data; applications were less responsive; and web surfing
when the web was new could be a frustrating experience. Remember the
first time you tried to access a graphical web site over a dial-up line? We
think of those experiences and imagine cloud computing to be more of
the same.
The fact is, broadband access and gigabit-speed networking has changed
all that, and the widespread access of broadband is the natural precursor
of widespread cloud computing.
The details: How to buy it
We've painted a broad picture now of cloud computing, what it is, what it
does, and how it helps us. Once you've made the decision to deploy
cloud technology, then all of a sudden you're faced with the details. RFPs,
needs analyses, sorting through the different vendors, watching their
presentations and making a decision.
The process is the same as with any other type of technology. Determine
what you need, put out an RFP, create a short list, and look at the offers.
One of the most important elements of purchasing cloud services is the
service level agreement (SLA). An SLA has long been an important aspect
of any sort of IT service, but in the era of the cloud, it becomes even
more vital since so much more depends on the service.
The first thing to evaluate is whether the cloud service provider offers an
SLA, and what the details of it may be. There are performance promises
that are nothing more than marketing fluff: "Great service or your money
back!" Then there are performance promises that have teeth. The SLA
should include in specific language, at least the following:
 Expected performance levels should be laid out in specific terms.
 Uptime percentage must also be stated in specific numeric values.
 Response time should also be stated in specific terms.
 The provider should also provide for consequences if the terms
are not met. For example, a penalty, free service for a period of
time, etc.
 The SLA may also set out specific tasks and deliverables, such as
reports or other functions.

75

Most cloud providers will offer a boilerplate SLA, and this may well be
adequate—but if you are a large customer, tweaking the terms of the SLA
is common. In some cases, the terms are negotiable.
Cloud providers, and other types of providers as well, have embraced the
value of the SLA, not just because it delivers value to the customer, but
for their own uses as well. From the perspective of the provider, it
protects against escalating client expectations. For example, some services
may deliver varying levels of guarantees, each one with a different rate.
There may be for example, one price for a 98 percent guarantee, and
another for a 99 percent guarantee.

XV. The Future of Cloud
If you’d like to see where cloud computing is going, you simply look at
the evolution of earlier computing platforms.
We will see the adoption of cloud computing technologies “move up the
stack” so to speak. If you think about desktop computing, it started with
hardware for several years, then some basic software, then operating
systems and development tools emerged, followed by an explosion of
special-purpose software. The cloud is evolving in the same ways:
infrastructure, commodity software, basic PaaS for deploying, ‘App Dev’
PaaS (development) tools, followed by an explosion of special-purpose
software on the cloud. The emergence of mature PaaS tools will fuel the
mass adoption of cloud computing in the long run.
Cloud Adoption
Resistance to new technology concepts is inevitable, and cloud computing
is no exception. Today, cloud computing has matured to the level where it
is a viable technology, ready to embrace and bring benefit to your
company.
The reasons why cloud computing's time is now include:
 Economic necessity
 Support from major mainstream software vendors
 Demand from small business for high-end features
 Demand from enterprise users for more cost-effective solutions
 Need for collaborative tools

76

 Cloud technology has already passed the proving ground stage
Economically, the market today is not only ready for cloud computing, it
demands cloud computing. On a macro level, the world is facing a huge
recession from which it will be slow to recover, and businesses of all sizes
need an edge just to stay competitive. Increasing revenues is always a key
strategy of any business, but the reality of the situation is that many
companies are not able to do so in a struggling economy. This leaves only
cost cutting as a way to stay flat or increase the bottom line.
When a business needs to cut expenses, it's not prudent to cut those areas
that contribute to the company's overall mission. Staff reductions may
provide a short-term shot in the arm, but in the long run, this may be
detrimental. The better strategy in cost cutting is to re-evaluate the
company's technological underpinnings, and implement new technology
that allows them to do more with less. Cloud computing is such a
technology.
Furthermore, mainstream software vendors have all staked a claim in the
cloud computing market. Enterprise software vendors, many with
reputations for massively expensive implementations that take months or
years to install successfully, have already rolled out cloud-based versions
of their otherwise bulky systems. The results have been astounding.
Enterprises have been using cloud-based versions of ERP software, for
example, to get up and running on individual modules immediately
instead of having to wait months for a custom rollout. Even if an on-
premises solution is ultimately desired, the cloud-based system allows
them to make an easier transition—and one of the hardest things about a
major ERP installation is the transition. On the lower end, midsized
businesses are taking advantage of these cloud-based enterprise systems to
get functionality that they couldn't afford before they became available on
the cloud. On the low end of the market, small businesses and SOHO
companies also have the cloud at their disposal as well, with offerings
from mainstream vendors like Microsoft, Google, and even Apple
delivering a wide range of cloud-based applications and services that
promote not only productivity, but increased collaboration as well.
The consumer market is especially important for the acceptance of cloud
computing, as this is where the technology initially filters into the business
mainstream. Consumers that have become accustomed to using Google
Apps, Microsoft Live, and Apple MobileMe will demand the same
functionality in the workplace.
The collaborative potential is just as important as the functionality of the
applications themselves. The fact that the cloud promotes collaboration

77

fits in well with today's ways of doing business. The decentralization of
the workplace, the growth of outsourcing, and the desire for
telecommuting and work-at-home solutions all demand collaborative
technologies to work, and this is now possible only through cloud
computing technology.
Finally, we have to look at the adopter stage of any technology. Early
adopters jump in when a technology is new and unproven, and serve the
purpose of providing a testing ground for the rest of us. Today, we see
that cloud computing has a rich collection of providers, both well
established and startup; and that users come from all segments and all
business size classifications. Furthermore, cloud computing providers
have expanded to encompass the entire range of cloud computing
technology (infrastructure, platform, and application), with prominent
vendors already offering robust deliverables in all three categories. Cloud
computing has passed that early adopter stage and is now entering the
mainstream.
To technologists, the future of cloud computing is easy to understand
because we have the advantage of history. To truly understand the future
of cloud computing technology, we merely need to examine the historical
evolution of earlier computing platforms. The cloud is evolving in many
of the same ways, with its infrastructure, platforms, and software.
More important is the effect of the cloud on the people who use it. We
may even say that cloud computing is reaching "critical mass." That is, it
has come too far to put it back in the bottle. It's here, the technology is
ready, and it is already making dramatic changes to the way people do
business, the way we work, and even the way we think. It is creating a new
class of entrepreneurs and ushering a second dotcom boom.
What are the implications of this technology achieving critical mass? For
one, IT buyers will not need to defend why they are buying cloud
computing services—the argument instead will focus on "why are you
using antiquated technologies?" and "Why are you spending ten times too
much on this project when you could be using cloud computing instead?"
Reactive Experimentation
More than any other business-driven technology, cloud has changed the
basic business model by which startups succeed. Traditionally, success in
an entrepreneurial venture has been painstaking, time-consuming, and
largely a function of attempting to copy the success of others.
The very detailed strategic roadmap that every startup is advised to have
may well be on its way out. Of course, strategy and planning will always

78

have a part in any business, but today's businesses—especially tech
startups—have to be able to respond quickly to changes along the way.
They have to be able to experiment. Because the cloud delivers these
businesses the ability to more readily experiment with business and
revenue models, as well as the application development process itself,
strategic planning, and the strict adherence to a roadmap, is being
replaced by experimentation.
Ultimately, this is a positive for the success of business. Try as one might
to create a perfect business plan, the fact is almost all business plans are
flawed on some level. Those flaws are seldom seen until they are
implemented, and by then, it's too late. Changing direction can be
difficult, both logistically and emotionally.
That is, unless one adopts the mindset from the beginning, that the
process of creating, building, and succeeding in a business is a wholly
experimental proposition.
The cloud, and its underlying technologies, promotes this approach by
making it far easier to experiment. A proof-of-concept can be developed
quickly on a platform-as-a-service, and deployed cheaply on a virtual
private server. Especially in the tech realm where user feedback is vital,
this early beta approach allows the developer to get a product in front of
consumers quickly, receive valuable feedback, and adjust the product
along the way—often in ways that benefit the product and the consumer
that would never have been considered with a more traditional long
planning stage.
The Cloud as an Operating System
In the 1996 documentary “Triumph of the Nerds”, Steve Jobs described
his early vision to take the desktop to the masses, "It was very clear to me
that while there were a bunch of
hardware hobbyists that could
assemble their own computers, or
at least take our board and add the
transformers for the power supply
and the case the keyboard and go
get, you know, et cetera, go get the
rest of the stuff. For every one of
those, there were a thousand
people that couldn't do that, but
wanted to mess around with programming - software hobbyists." Steve
Jobs was a visionary. He knew that his computer needed an operating

79

system so that millions of people could use it. Tools like VisiCalc brought
personal computers to the masses.
In many respects, today’s cloud infrastructure is similar to the desktops of
the 1980s. Although it ultimately benefits the ordinary end user, it's mostly
the techies that get excited about it and that continue to refine it. The
result will be the same. Just as PCs were once seen as something "with
potential" but nonetheless only used by a handful of "hobbyists" as Jobs
puts it—or "geeks", to not put so fine a point on it; cloud computing is
seeing the same evolution. Just like prior computing platforms, the cloud
needs an operating system to make it more useable.
The cloud is evolving in many of the same ways as desktop computing.
Cloud platforms are helping to bring cloud software to the masses. In
essence, platforms are evolving into operating systems for the cloud.
1980 2015
PC Cloud
Software in a Box Software as a Service
Operating Systems Platform as a Service
Hard drives Infrastructure as a
Service

An operating system manages computing resources, shields developers
and other users from under-the-hood complexities. That’s what cloud
platforms and brokers do for the cloud. The platform implements a
virtual instance of a core set of functionality, with common features such
as user signup, security, reporting, and so forth. The platform will then
allow developers to build on top of that instance to customize it for their
specific needs, and to additionally build features on top of the platform
without programming, such as creation of forms, data entry collection,
report writing, etc. As a result, users can log into one place and experience
an integrated solution, where they do whatever they do.

The Top Transformational Impacts and Trends
This era of cloud computing is behind some of the biggest shifts in
business since the Industrial Revolution. Today, the genie is out of the
bottle, and change is imminent and inevitable. We're not alone in
predicting a major role for cloud computing. Virtually every research

80

organization and think tank has declared it to be a technology that is
changing the way we think about computing.
The disruptive impact of cloud is uprooting old industries and making
way for new. Whereas change is not new, the rate of change may be
accelerating faster than many imagined. In the coming year, the cloud's
impact on business and government strategies will continue to accelerate,
and it will be the biggest driver behind major IT decisions. This, in turn, is
causing the biggest disruption to the IT industry in the past 25 years.
1. Cloud technologies will converge.
The cloud will continue to forge a massive convergence of technologies—
similar to the evolution of the cell phone to the smart phone. “The lines
between platform as a service (PaaS) and cloud services brokerages will
blur into a conceptual operating system for the ‘Web as a platform’—
providing tools to allow users to take advantage of multiple cloud
solutions at once, and bringing the cloud closer to the end user in more
meaningful ways," according to Landis.
2. Rise of the ‘Cloud Developer’
Non-commodity custom software is beginning to move to the cloud in a
meaningful manner. PaaS and related cloud technologies have reached a
maturity level that allows developers and integrators to build highly
customized, complex offerings on the cloud. For years, everyone has been
talking about the “low-hanging fruit” of commodity email and
infrastructure. The first wave of cloud was about IT. The next wave will
be about a new kind of software developers… ‘cloud developers’.
3. Integration will become the new “killer app.”
The term “killer app” generally refers to the technology that’s so
necessary it drives adoption of a computing paradigm. “Complexity is the
problem of the cloud era,” Jackson said. “The cloud is evolving into a
hodgepodge of disparate cloud services from vendors that are scattered all
over the world.” According to Jackson and Landis, IT professionals will
turn to cloud services brokers to manage the growing complexity problem
by integrating heterogeneous infrastructure services; whereas software
developers will turn to PaaS for integrating disparate Web services to
deliver seamless user experiences to their customers.
4. India and outsourcing countries will drive industry
adoption of PaaS worldwide.
“The software development outsourcing industry thrives on the value
proposition of more for less, as it is what it does well,” said Landis. In

81

2013, PaaS will be adopted by companies in India and in other major
outsourcing countries in a rapid and notable fashion. It will cause a ripple
effect throughout industry because these outsourcing companies are so
integral to modern business operations. The cloud makes geographic
boundaries irrelevant.
5. Major data centers will go undergo a “survival-of-the-
fittest” scenario.
Winners will emerge in the data center shakeout, as many large data
centers will close and sell assets, or become acquired and consolidate.
"Ironically, the cloud movement spawned a ‘gold rush’ to build new data
centers at a time when the stated goal of the cloud was to reduce the
number of data centers," said Landis. "The market is maturing, and the
ultimate result will be a ‘survival-of-the-fittest’ scenario as many legacy
data centers will shut their doors and as many data center customers
decide to move some of their data to the cloud." Cloud services
brokerages will play a larger role for data center service providers to help
their customers sort out the confusion and effectively manage an
increasing number of cloud service providers, according to Jackson.
6. Health IT will adopt PaaS to replace niche “dinosaur” apps.
Health IT will adopt PaaS to replace niche “dinosaur apps.” “The
problem with health IT is that the enterprise systems have grown too
large to merely replace, but there are limits to what they can do,” said
Landis. “For the healthcare industry to move forward and achieve the
goals set out in the Health Information Technology for Economic and
Clinical Health Act, the cloud must play a major role in the next
generation of healthcare IT. Healthcare CIOs are looking at how much it
will cost to upgrade their HITECH-compliant systems—and they are
finding that proprietary models will leave them with unfavorable lock-in,
while others in the healthcare industry embracing the cloud march
forward.”
7. Organizations will rapidly adopt cloud services
brokerages.
The need to use multiple cloud services providers to manage multiple
functions will create a fast adoption of the cloud services brokerage
model: either via a new internal role or external source, like the NJVC
Cloudcuity Management Portal. “The new role of cloud services
brokerages will be further defined and evolve over the next five years to
provide niche services to organizations moving to the cloud, but also

82

realizing that their specific IT needs will require the use of more than one
cloud services provider,” said Jackson.
8. The U.S. government will re-think major IT contracts.
Large system integrators, whose success long depended on very
expensive, highly complex and customized, on-premise solutions, will
redefine their practices and their overall operating and profit models,
especially as more businesses go beyond basic, "out-of-the-box" cloud
services and turn to the cloud for more customized implementations. U.S.
government agencies will begin to add new requirements to several major
IT contracts. According to Jackson: “The federal government is in many
ways leading the path toward the cloud. In 2013 there will be a shakeup in
government contracting. Incumbent system integrators will no longer be
able to rest on past successes, as the government continues to transform
itself and move toward the cloud, and as new cloud service providers
offer innovative and cost-effective solutions geared specifically to the
federal marketplace.”
9. Innovation and entrepreneurship will hit overdrive.
Entrepreneurship will go into overdrive, especially as full-featured, "idea-
to-revenue" platforms, like NJVC Cloudcuity AppDeployer, take
developers from concept, to development and to deployment and sales.
This will trigger a new wave of innovation, entrepreneurship and
disruptive startups that will make things interesting for system integrators.
“We’ve already seen an incredible wave of high-tech innovation, and the
emergence of flourishing incubators and accelerators," said Jackson.
“Cloud platforms, for the first time, provide these innovators with all the
tools they need to succeed, without requiring a multi-million dollar
investment.”
10. Cloud adoption will move from an option to a "must have.”
Cloud adoption will move away from something buyers purchase with
surplus budget money to a “must have” that replaces the traditional IT
enterprise business model. “Until recently, managers have viewed cloud
computing as a proof-of-concept project or something that can be done
or piloted with extra budget money," said Landis. "The reality is the
cloud's value can only be fully realized when traditional and more costly
ways of storing, using and securing data are replaced with new business
models that take advantage of ‘fast-and-lean’ cloud services. In the
coming year, companies will accomplish this by halting old projects, re-
thinking old contracts, and shifting funds to affordable and innovative
cloud services that can transform the IT enterprise."

83

11. Prices fall on Cloud Infrastructure.
Cloud computing already provides a pricing advantage to end users, who
gain access to high-end applications at entry-level prices. The
infrastructure, upon which the rest of the cloud lives, will also decrease in
price as more major players enter into the market to provide commodity
infrastructures to hold the increasing number of cloud applications.
Meanwhile, the competition is steepening. Together, this will make it even
cheaper for application providers to enter into the market.
12. Cloud platforms go open.
It’s “déjà vu all over again” as the proprietary shakeout gives way to open
platform interfaces. These open systems not only simplify development
and provide for more robust applications, they allow for a greater level of
customization, and they answer the vexing question of what happens to
an application if a provider goes out of business. Some platform vendors
(SaaS Maker and App Deployer) already provide open interfaces for
integrating heterogeneous web services from favorite vendors.
13. Home sourcing becomes mainstream.
Cloud computing will drag us kicking and screaming out of our cubicles
and into our homes. There will be resistance on several fronts, but the
move is inevitable due to the incredible efficiency gains and cost savings
to companies. Because applications and data no longer need to reside on
the computer in front of us, the physical office is quickly becoming
redundant.
14. Corporate processes become decentralized.
Larger companies take advantage of the decentralization made possible by
cloud computing. This leads to a greater level of outsourcing, which in
turn triggers the need for smaller companies to fill the need for those
outsourced services.
15. A new wave of entrepreneurship emerges.
Cloud computing ushers in the next great dotcom boom, only this time
things are different. Cloud computing has lowered the barriers to entry so
that anyone can be a dotcom superstar. Entrepreneurs won’t need to be
programming wizards or venture backed. They only need an idea,
ambition, and a credit card.
16. Smart phones with cloud become the norm
Smart phones like the iPhone and BlackBerry continue to gain
functionality and power, and their reach extends further with easier access

84

to wireless broadband. This makes smart phones more attractive as an
actual working machine, and a tool for accessing productivity apps over
the cloud for corporate use.
17. The days of mega-million dollar enterprise software
projects come to an end.
Those multi-year deployments, high failure rates, and big price tags are
already pushing their limits, and enterprise customers are demanding
something better. Enterprise-level cloud computing apps will gradually
replace those huge on-premises implementations with a modular
approach; and the existence of cloud platforms will encourage new
entrants into the enterprise market. The days of mega-million IT projects
will eventually fall by the wayside along with the fall of ground-up Web
2.0 engineering. Think about it – who, these days, would want to write an
e-commerce website from the ground up when you can rent an e-
commerce server? Yesterday’s million dollar systems only cost a few
dollars today. Likewise, cloud platforms will become the norm rather than
the exception. The same thing is happening with other types of platforms,
from social platforms to enterprise business systems.
18. Cloud computing penetrates all areas of business
management.
The earliest applications delivered more consumer-oriented applications
and services, although cloud computing is by no means a consumer-only
technology. Already in widespread use by SOHO and small businesses, it
is expanding into larger enterprises. The result will be that cloud
applications will evolve to accommodate more mission-critical needs,
delivering full-fledged management systems to the largest government
agencies and corporations in the world.
19. Big-name companies struggle for new identities.
Something fascinating happens when computing platforms change: the
big IT boats get rocked. Hello IBM, do you remember that little company
named Microsoft? Hello Microsoft, do you remember that little company
named Google? Hello Facebook, do you remember that little company
named??? The emergence of new cloud offerings from names like
Rackspace will drive competition in the cloud infrastructure arena. Cloud
platforms are enabling tens of thousands of software newcomers. Cloud
platforms will gain attention from infrastructure providers looking for
new competitive advantages. In the end, several new brands will emerge,
both from established players and newcomers to the market. The space
will become more cluttered before eventually shaking out. In the process,

85

some companies will change, some will emerge, and yes, some companies
will die to make room for the new.
20. Social networking systems will evolve into collaborative
management systems.
Today’s managers need to get things done despite growing challenges.
Their teams are more scattered and complex... more difficult to motivate,
coordinate, and hold accountable. An honest manager will tell you that
real work is still being done with spreadsheets and emails. For these
reasons and more, the future of collaboration will be more focused on the
emerging needs of mangers who are coping with more complexity and
demands. They need more than social networking. They need interactive
management systems with real reports.

What’s Next?
The future is unfolding quickly. It has been said that 1 year in computer
technology is like 10 years in the automobile industry. Who then knows
what next year holds?
In the 1980s, PC computing showed us just how fast new computing
technologies could reach the world. Cloud computing will move much
faster because it has several advantages, including today’s Internet.
In a few years, we will go to our cloud desktop. It will probably look a lot
like today’s PC desktop. The underlying technologies will be different, but
we’ll leave those details to the techies. Soon, the hype will subside, but the
cloud will be here to stay. We will use it without thinking about it.
We’ll simply log on to do whatever we do.

86

Acknowledgements

The authors wish to thank Kevin Jackson, from NJVC, for his
contributions and insights into trends in cloud brokerage; Sharon Emeigh,
for her tireless efforts in copyediting and proofreading; and Brianna
Watts, for lending her talent in layout and in designing the book cover.

87

About the Authors

Cary Landis owns and operates Virtual Global, Inc., where he currently serves as
lead architect for the SaaS Maker™ platform as a service. He serves as Executive
Director of FedPlatform.org, a nonprofit that helps Federal organizations adopt
cloud technologies. Landis has served in leadership roles at KeyLogic Systems
and TMC Technologies. He currently serves as Practice Lead for NJVC’s
Cloudcuity™ AppDeployer Platform as a Service. He helped The National
Institute of Standards and Technology (NIST) create its Reference Architecture
for Cloud Computing. As part of that effort, he led a NIST team to clarify the
cloud service models (SaaS, PaaS and IaaS), and led the Federated Community
Cloud committee. Cary resides in Morgantown, WV.
www.virtualglobal.com

88

Dan Blacharski is the founder and editor-in-chief of techie.com, a destination for
disruptive technology and emerging trends. The innovative web destination
offering insight into technology and how it relates to how people work, live, and
play. He has published numerous books and articles for technology trade and
popular press. While in Silicon Valley, Dan created content for, and helped, many
dotcom start-ups fine-tune their messaging strategies, and had an opportunity to
see first-hand what works, what doesn’t, and what still gets supported even
though it defies all common sense.
www.techie.com

89

Index
Amazon, 12, 19, 31, 80 NIST, 12, 18, 24, 76, 80

as-a-service, 11, 12, 13, 23, 29, offshoring, 55
31, 32, 49, 63
on-demand, 12, 18
Benefits, 23
Open source, 71, 72
bloatware, 38
operating systems, 9, 10, 40, 54,
Cloud Infrastructure, 10, 18 75

Cloud Platforms, 10, 22, 23, 25 PaaS, 11, 22

Cloud Software, 10, 29 PC, 9, 10, 14, 20, 47, 53

collaboration, 8, 12, 33, 36, 38, 55, Platform-as-a-Service, 10, 22
59, 63, 64, 70, 84, 94
Private cloud, 80
Collaboration, 36, 57
risks, 22, 24, 25, 50, 73, 77, 80
Community cloud, 80
SaaS, 11, 23, 29, 30, 36, 37, 38,
data center, 17, 20, 21, 47, 78, 79, 50, 53, 75
80
Scalability, 36
elasticity, 12, 23
Security, 73, 74, 78
enterprise software, 17, 25, 26, 48,
Simplicity, 36
49, 50, 93
SLA, 79, 81, 82
Facebook, 12, 25, 49
SOAP, 31
grid computing, 31, 52
SOHO, 38, 43, 50, 84, 93
Health-IT, 66
supercomputing, 31
HIT, 67
teleworking, 57
HPC, 31, 32
total cost of ownership, 28
Hybrid cloud, 80
Virtualization, 17, 21, 22, 37
IaaS, 11, 18
Web 2.0, 36, 37, 55, 59, 64, 93
Infrastructure-as-a-Service, 11, 18
web services, 23, 30
mainframes, 47
Windows, 9, 15, 36, 75
Middleware, 25, 64
XML, 31
NASA, 69, 70, 75, 80

Nebula, 69, 70, 80

90

Cloud computing-made-easy

Recommended

More Related Content

What's hot (20)

Similar to Cloud computing-made-easy (20)

More from GovCloud Network (20)

Cloud computing-made-easy