CLOUD COMPUTING - 101

STUDENT STUDY GUIDE

1.01 Cloud Computing: Common Terms and Definitions

1.02 Cloud Computing and Virtualization
1.03 Early Examples of Cloud Computing
1.04 Cloud Computing Deployment Models
1.05 Positive Indicators for Cloud Readiness
1.06 Negative Indicators for Cloud Readiness
1.07 Cloud Service Categories
Two-Minute Drill
WHAT IS CLOUD COMPUTING?

The interesting thing about cloud computing is that we’ve redefined cloud
computing to include everything that we already do. I can’t think of
anything that isn’t cloud computing with all of these announcements. The
computer industry is the only industry that is more fashion-driven than
women’s fashion. Maybe I’m an idiot, but I have no idea what anyone is
talking about. What is it? It’s complete gibberish. It’s insane. When is this
idiocy going to stop?
Larry Ellison, CEO, Oracle, September 2008

O ver the past few years, virtualization and cloud computing have

changed the way information technology (IT) is organized. It has become less
important for companies to maintain their own local servers and more
advantageous for them to share computing resources, which may be owned
by a third-party service provider. Virtualization enables hardware to be
shared among virtual servers. Cloud computing enables these servers to be
shared with multiple users so that their location does not matter anymore,
as shown in Figure 1-1.
FIGURE 1-1 With cloud computing, location is not relevant.

CERTIFICATION OBJECTIVE 1.01

Cloud Computing: Common Terms and Definitions
Figure 1-2 illustrates that cloud computing is a business model using old
technology (virtualization, servers, and disk storage) where ownership of
physical resources rests with one party and where the service users are billed
for their real use. An organization can even use virtualization for internal
customers, or an organization can subscribe to cloud services over the
Internet.
FIGURE 1-2 Cloud computing is a business model using old technology.

Cloud computing means different things to different types of people. An IT

administrator might define it as the sharing of pooled computing resources
over the Internet. A business owner could refer to cloud computing as using
software that is hosted on a cloud provider’s equipment, which could result
in a cost advantage compared to hosting these services internally. A home
user might refer to cloud computing in terms of the free storage of files at a
cloud provider’s site as opposed to storing files locally on a personal device.
All of these definitions share common characteristics, including the use of
offsite hosted computing services accessible over a network, such as the
Internet. However, these definitions differ in terms of what types of
computing services are offered and at what cost. Public cloud computing
refers to IT services being offered to any user with an Internet connection,
whereas private cloud computing refers to IT services being offered within a
single organization on that organization’s equipment.
With cloud computing, the cloud service provider hosts and gives access to
the software and data storage. Data might be stored in one or more data
centers and might even be replicated between data centers for data
redundancy.
Existing technologies such as virtualization allows multiple computer
operating systems (virtual servers, or cloud servers) to run simultaneously
on one physical computer as if they were each running on their own
dedicated computer.
There are distinct categories of cloud computing services. Platform as a
Service (PaaS) allows developers to create and test software without the
investment in expensive in-house hardware.Software as a Service (SaaS)
answers desktop productivity needs for end users. Infrastructure as a
Service (IaaS) allows IT operations personnel to run applications in the
cloud instead of using in-house computing equipment.
As organizations grow or shrink, they can simply request more (or less)
computing resources from their provider and pay the corresponding fee. This
ability to rapidly add and remove computing resources is called elasticity.
For paid cloud services, in addition to a monthly subscription fee, you would
pay usage fees. For example, the more disk space you use to store files in the
cloud, the more you would pay. This means capacity planning is an important
skill since it can save organizations money.
Types of computing resources a client could consume include the following:
Email accounts
Data storage
Data backup
Servers (virtual machines)
Customized websites
Cloud Computing Characteristics
The National Institute of Standards and Technology (NIST) states that cloud
computing services have the following attributes, as shown in Figure 1-3:
FIGURE 1-3 Cloud computing characteristics

Elasticity You can add or remove services/users quickly. This allows an

organization to grow or shrink on demand as there are spikes in business
activity.
On-demand You can access the service from anywhere, anytime, often
simply using a web browser. The physical location of the software and your
data at any given time is not known and may be spread across multiple data
centers; this maximizes the “anywhere, anytime” nature of cloud computing.
Pooled computing resources at the provider’s site The provider
invests in the proper hardware, configuration, maintenance, and auditing of
their physical infrastructure.
 Monitored and measured service usage Paid cloud services often
have a monthly recurring subscription fee as well as usage fees. Time using a
specific resource, percentage of processing capabilities, amount of disk space
used—these are all monitored by the cloud provider that then charges
customers accordingly.
Broad network access Access to cloud services is available from any
type of computing device such as a smartphone, table, laptop, or desktop
computer over the Internet.

Cloud elasticity is similar to scalability. Whereas scalability is defined as the ability

of a system to grow to accommodate need, elasticity allows cloud customers to not
only increase but also decrease the amount of IT services consumed.
CERTIFICATION OBJECTIVE 1.02
Cloud Computing and Virtualization
Virtualization technology makes cloud computing possible, and it has been
around for decades. Providers can host many virtual machines based on
client needs without allocating a physical server to each client. Some cloud
offerings do allow clients to use physical servers, but this is costly and is not
a common approach.
Nonvirtualized computers have one set of hardware and one set of software,
as shown in Figure 1-4. Virtualization is a set of techniques for hiding
hardware resources behind software abstractions to simplify the way in
which other software or end users interact with those resources.
FIGURE 1-4 Multiple virtual machines run on a single virtualization host.
The popular press typically identifies virtualization as a category of vendor
products. However, in an enterprise, virtualization should be understood as
an integrated approach encompassing software, hardware, and a number of
techniques and processes that, when taken together, increase the ease with
which IT assets can be delivered, managed, and shared.
Besides creating cloud virtual servers on demand, you can even migrate a
physical or virtual on-premises server in your organization to the cloud. This
allows for a quicker and smoother transition to cloud-delivered services.
CERTIFICATION OBJECTIVE 1.03
Early Examples of Cloud Computing
The swinging 1960s introduced the concept of what we now know as cloud
computing. Computing resources were so expensive during that era that time
sharing became a common means of allowing different groups of people to
use computing power concurrently by doing their work in isolated sessions.
Government agencies, universities, and some private organizations were
part of this. Cloud computing does not have to be available to public
subscribers; private clouds offer computing access within an organization.
In the 1990s Hotmail was (and still is!) widely used as a public cloud service
to access email from a web browser instead of a specific mail program. This
meant you could access your email from any Internet-connected machine
without having to install and configure specific mail software. Application
service providers (ASPs) at this time offered additional services beyond
email that were available to users through a web browser. ASPs differ from
cloud services in that clients purchased specific software to be hosted by the
ASP. The problem with this approach was that providers did not have
expertise in the wide array of hosted software used by their clients. With
cloud services, providers offer standard software to their clients, such as
Hotmail or Google Docs; these both are examples of Software as a Service.
CERTIFICATION OBJECTIVE 1.04
Cloud Computing Deployment Models
As we have seen, cloud computing can be available to any subscriber or just
within an organization. In addition to the cloud service models, there are also
a number of ways to deliver or roll out these cloud services. In some of these
deployment models, the financial commitment lies with the organizations
that use these clouds. This means they do not completely fit our earlier
definitions of cloud computing.
Private cloud
Community cloud
Public cloud
Hybrid cloud
Figure 1-5 depicts the ways in which these cloud services can be deployed.
FIGURE 1-5 Cloud deployment models

Private Cloud
Companies using their own hardware and software assets to deliver web
services are said to be using aprivate cloud. This solution will usually use
virtualization and may offer a web portal that allows managers to provision
services for their users.
Community Cloud
Community clouds pool computing resources and make them available to
several organizations with common needs, such as very quick access time to
an application or adherence to strict security and audit guidelines. Cloud
customers are often referred to as tenants. Multitenancy often describes the
multiple cloud customers using a shared pool of computing resources. Even
if cloud customers are using the same cloud services, each customer must
have an isolated computing environment from other cloud customers where
their configurations and data are stored separately.
Public Cloud
Public clouds offer computing services to all Internet users. This is what most
people think of when they think of cloud computing. Some cloud offerings
are free, such as Gmail and Google Docs. Others, such as Microsoft Office
365, require that customers pay a subscription and usage fee for services
delivered via their Internet connection. The service provider has made the
investment in computing resources, and this frees up capital resources for
enterprise customers.
An Internet connection is required for public cloud subscribers. The nature
of the cloud service and how many local users will access it concurrently can
sometimes necessitate a better Internet link. The amount of data throughput,
often expressed in bits per second (bps), is referred to as bandwidth. A 100
megabit per second (100Mbps) connection allows more throughput than a
20Mbps connection.
Hybrid Cloud
Hybrid clouds are a combination of both public and private clouds. An
organization can integrate some of their on-premises services with a cloud
provider. For example, a company’s local user accounts can be used to
authenticate against cloud services such as cloud email and cloud storage.
Authenticating users in one environment and authorizing them to use
resources in a different environment is referred to as federation, which is
discussed in further detail in Chapter 3.
A second hybrid cloud example might include a company with an on-
premises email server that uses a cloud antivirus service. Of course, this
would imply a trusted network connection between the mail server and the
cloud provider.
Cloud computing appeals to business leaders because it is a business
concept; in other words, you pay only for what you use (“pay as you go” or
“rent instead of buy”). From a business or financial perspective, this could be
interpreted as meaning computing dollars can be categorized as operating
costs rather than capital expenditures. Accountants and business leaders will
not be interested in the underlying virtualization technology that makes this
all possible. Chapter 3 discusses public and private cloud deployment
models in more detail.
CERTIFICATION OBJECTIVE 1.05
Positive Indicators for Cloud Readiness
Buying hardware, buying software licenses, paying IT staff to install and
support the hardware and software—these all require capital that might be
put to better use elsewhere within an organization. For small and new
businesses, the capital to invest in IT might be nonexistent. Paying only for
what you use can be an affordable solution to meet your computing needs,
and this applies equally to larger enterprises. Virtualization and cloud
computing can overcome the following problems:
Running out of capacity
Costly excess capacity
Tied-up capital
Figure 1-6 shows the problems that can be solved with the proper cloud
solution.
FIGURE 1-6 Problems overcome by using cloud computing
Businesses that grow or shrink rapidly benefit greatly from elasticity.
Sometimes projects undertaken by a business require an IT investment only
until project completion. Why have expensive computer hardware and
software sit idle at the end of a project? Cloud services can grow and shrink
as your business needs do, and you pay only for what you are using. Think of
your water bill; you pay only for the amount of water you’ve used.
Netflix presents a great example of elastic infrastructure needs. Media
content must be converted to many different formats since there are so many
different viewing devices. Even though Netflix is Internet content delivered
to the end user (otherwise known as Software as a Service), Netflix requires
enormous processing and storage capabilities (also known as Infrastructure
as a Service). Once the media conversion is complete, the usage charges for
CPU utilization cease. Compare this to Netflix purchasing all of the hardware
required for periodic media conversions.

One of the most common cloud computing concerns is

data security. Cloud providers must pass periodic third-party
audits, and as such your data is probably safer with them than
with you. Despite this, many decision makers consider this a
showstopper.
CERTIFICATION OBJECTIVE 1.06
Negative Indicators for Cloud Readiness
Not all organizations benefit from virtualization and public cloud computing.
Predictable and fixed workloads can typically make optimal use of their
hardware and do not need the scaling.
Organizations that own large data centers often have enough scale to be as
flexible and efficient as cloud computing providers are. They will, however,
typically benefit from virtualizing their infrastructure.
Legal and security reasons, also termed compliance reasons, can require
an organization to know more details about the location of its data and
servers than a cloud computing provider can provide.
Organizations with systems requiring a high degree of operational
assurance, such as military systems and systems responsible for health,
safety, and emergency response, may be a poor fit for the guarantees offered
by typical public cloud service providers today.
Cloud computing is not a cookie-cutter solution to all financial and
technology challenges an organization may face, as shown in Figure 1-7. Care
must be taken when evaluating potential cloud solutions.
FIGURE 1-7 Cloud computing is not always the best choice.
As mentioned, essential services such as emergency, health, law
enforcement, or the military may not be suitable cloud customers. Cloud
providers provide service level agreements (SLAs) to their clients. SLAs
guarantee clients a certain level of service and uptime, but this may not be
acceptable to the essential services listed earlier.
CERTIFICATION OBJECTIVE 1.07
Cloud Service Categories
Cloud usage models have evolved over time as a result of user needs. What
follows are the three most common cloud service categories, also listed
in Figure 1-8.
FIGURE 1-8 Cloud computing models
 Software as a Service (SaaS) For users. A user account allows access
to software that is accessible anytime from anywhere using a web browser.
Data is stored in the cloud and thus facilitates collaboration. Examples
include Gmail, Salesforce, and LinkedIn.
Platform as a Service (PaaS) For developers. This service model
allows rapid development of new applications and websites. Servers,
networking, storage, databases, and so on, are made available by the cloud
provider. Vendor-specific tools present the issue of vendor lock-in. Examples
include Windows Azure and Google App Engine.
Infrastructure as a Service (IaaS) For IT administrators.
Infrastructure includes the self-provisioning of servers, storage, and so on,
via a web-based portal. This service model includes the creation of cloud
servers, cloud-based storage, and backup. Examples include Rackspace and
Amazon S3 Storage. Companies using some of these services include Shell,
the New York Times, and NASDAQ.
Anything as a Service (XaaS) is a term for any service delivered over the
Internet instead of being hosted locally on your network, so it includes IaaS,
PaaS, SaaS, and many others.

There are many other classifications of cloud services

such as Disaster Recovery as a Service (DRaaS), Desktop as a
Service (DaaS), and many more.
TWELVE ADOPTION SCENARIOS
Businesses vary greatly in size, sector and maturity, and they can have very
different IT requirements.Chapter 4 provides a selection of real-life case
studies for Software, Platform and Infrastructure as a Service (SaaS, PaaS
and IaaS); but Table 1.1 below lists common adoption scenarios where a
choice could be made between cloud computing solutions and non-cloud
solutions. The scenarios are wide-ranging, but they all serve to demonstrate
the relative convenience of IT solutions in public cloudswhere capital
investment in hardware and software by the customer is not necessary.

TABLE 1.1 Non-cloud and public cloud solutions for 12 common IT

requirements

Requirement Non-cloud solution Public cloud

solution
Start-up business needs Set up an e-mail server Use an online office
e-mail, file sharing and and a VPN for file sharing; application software
office applications for install necessary software suite (SaaS)
varying numbers of on employee computers
employees and and ask contractors to
contractors provide their own office
application software
Office PCs or laptop Upgrade computers and Choose SaaS and
computers are a few install new, improved run it in a web
years old and running desktop software browser on existing
desktop software slowly, computer hardware
but the software is old,
too, and needs
replacing

Remote working Set up VPN connections to Use virtual hosted

capability to enable allow remote desktop desktops accessible
employees to use the connections and network from anywhere
same desktop software drive access or use a third- (including office and
and data as they have in party service to connect to home) or switch to
the office PCs in the office from other SaaS solutions
anywhere
Requirement Non-cloud solution Public cloud
solution

Mobile working Modify key IT systems to Use SaaS to get

capability make them accessible ready-made mobile
from mobile devices such working solutions
as ‘smart’ phones

New Customer Set up one or more servers Use a SaaS CRM

Relationship for the CRM database and which can be
Management (CRM) install desktop software on accessed via a web
system for sales staff users’ computers (if browser
required)
Security fears Encrypt key data stored on Store and access
surrounding data stored laptops so that they cannot data in the cloud
on laptop computers be accessed without a using SaaS without
taken out of the office password downloading it onto
laptops

Need to collaborate Install third-party or in- Choose and use

effectively with partner house developed software SaaS collaboration
companies online on a web server software or develop
new software with
PaaS

Software development Set up numerous servers Use a cloud-based

required to manage for development, testing software
enterprise business and production development
processes platform (PaaS) that
supports systems
integration

Distribute new business Create a desktop software Create web-based

software application to application and sell it software using PaaS
customers online or in DVD form that can be sold on
the platform
provider’s online
market place
 Requirement Non-cloud solution Public cloud
solution

New marketing Upgrade server and Move website to an

campaign will bandwidth to cope with a elastic public cloud
temporarily overload large number of hits in a (IaaS) so it can
current web server short space of time respond dynamically
to highs and lows in
web traffic
Automated backups for Set up a secondary data Use IaaS to back up
IT systems and centre to copy backups to data and virtual
business data servers
Run a large and Use all available Run the simulation on
complex computer computing hardware for as a temporary, cloud-
simulation long as it takes based (IaaS)
computer grid

QUICK TECHNOLOGY TIPS

The particular cloud computing technologies you choose (if any) depend on
your working practices, your business size, your current IT systems and the
skills of your internal staff. All these points will be covered in later chapters,
but here are some quick technology tips to bear in mind as you proceed
through the book:
 If you want off-the-shelf software that you can access from anywhere
then choose Software as a Service; but look for solutions with
comprehensive web services interfaces to facilitate integration with
other systems.
 If you want to customize your software choose Platform as a Service
over SaaS; but choose the solution that best matches your in-house
skills, beware vendor lock-in, and consider choosing a platform that
has a wide range of ready-made applications that you can plug in to
yours.
 If you want complete control over your application servers use
Infrastructure as a Service; but consider the portability of your virtual
machines as you may wish to move them between clouds at some
point.
 If you do not want your data to be hosted in a public cloud then
use private cloud technologies; but choose an internal cloud
management system that supports a hybrid cloud configuration in case
you ever need to manually or dynamically increase your available
computing capacity.