A Technical Seminar Report

on
CLOUD COMPUTING
Submitted in the partial fulfillment of the requirement for the award of the Degree
BACHLOR O! NGINRING
in
LCTRONICS " COMMUNICATION NGINRING
o#
$IS$S$ARA%A TCHNOLOGICAL UNI$RSIT%
BLGAUM
SUBMITTD B%
BALARA& '(S( )*SU+,C++,-
.+//
DPARTMNT O! LCTRONICS " COMMUNICATION
NGINRING(
SDM INSTITUT O! TCHNOLOG%0
U&IR12,*.*+(
1

$IS$S$ARA%A TCHNOLOGICAL UNI$RSIT%
S(D(M( INSTITUT O! TCHNOLOG%0 U&IR12,*.*+
DPTARTMNT O! LCTRONICS " COMMUNICATION

CERTIFICATE
This is to certify that Mr. BALARA& ' S bearing USN:
*SU+,C++, a bonafide student of SDM !nstitute of Technology
U"ire has presented the seminar entitled
CLOUD COMPUTING and submitted the report towards the
partial fulfillment for the award of #achelor of $ngineering in
$lectronics % &ommunication $ngineering during the academic
year '(11.


Signature of the in charge Signature of the HOD
'

Abstract
Clo34 comp3tin5 describes computation software data access and
storage ser)ices that do not require end*user +nowledge of the physical
location and configuration of the system that deli)ers the ser)ices. ,arallels
to this concept can be drawn with the electricity grid where end*users
consume power resources without any necessary understanding of the
component de)ices in the grid required to pro)ide said ser)ice. &loud
computing is a natural e)olution of the widespread adoption of
)irtuali-ation ser)ice*oriented architecture autonomic and utility
computing. Details are abstracted from end*users who no longer ha)e need
for e.pertise in or control o)er the technology infrastructure /in the cloud/
that supports them. &loud computing describes a new supplement
consumption and deli)ery model for !T ser)ices based on !nternet protocols
and it typically in)ol)es pro)isioning of dynamically scalable and often
)irtuali-ed resources. !t is a byproduct and consequence of the ease*of*
access to remote computing sites pro)ided by the !nternet. This frequently
ta+es the form of web*based tools or applications that users can access and
use through a web browser as if it were a program installed locally on their
own computer.
0

CONTNTS

/( Intro43ction to clo34 comp3tin566666666666666(( +2
.( Concept o# clo34 comp3tin5666666666666666((((+7
.(/ Compari8on666666666666666666666 +,
.(. Implementation6666666666666666666( +,
.(9 Characteri8tic866666666666666666666 +,
.(* conomic8666666666666666666666(( +:
.(2 Companie8666666666666666666666( +:
9( Hi8tor;6666666666666666666666666 +<
*( Le5al i883e866666666666666666666666 /+
2( Sec3rit; i883e86666666666666666666666 //
7( Ri8= miti5ation666666666666666666666(( /.
,( 'e; characteri8tic866666666666666666666 /9
:( Component86666666666666666666666(( /*
<( Architect3re6666666666666666666666((( /,
/+( T;pe86666666666666666666666666( /:
//( Role86666666666666666666666666( .+
/.( Stan4ar4866666666666666666666666(./
/9( Ca8e 8t34;66666666666666666666666 .,
/*( Concl38ion6666666666666666666666(( .:
/2( Re#erence866666666666666666666666 .<
1

Chapter 1
Introduction
!magine yourself in the world where the users of the computer of today2s
internet world don2t ha)e to run install or store their application or data on
their own computers imagine the world where e)ery piece of your
information or data would reside on the &loud 3!nternet4.
5s a metaphor for the !nternet /the cloud/ is a familiar clich6 but
when combined with /computing/ the meaning gets bigger and fu--ier.
Some analysts and )endors define cloud computing narrowly as an updated
)ersion of utility computing: basically )irtual ser)ers a)ailable o)er the
!nternet. 7thers go )ery broad arguing anything you consume outside the
firewall is /in the cloud/ including con)entional outsourcing.
&loud computing comes into focus only when you thin+ about what we
always need: a way to increase capacity or add capabilities on the fly
without in)esting in new infrastructure training new personnel or licensing
new software. &loud computing encompasses any subscription*based or
pay*per*use ser)ice that in real time o)er the !nternet e.tends !&T8s e.isting
capabilities.
&loud computing is at an early stage with a motley crew of pro)iders
large and small deli)ering a slew of cloud*based ser)ices from full*blown
applications to storage ser)ices to spam filtering. 9es utility*style
infrastructure pro)iders are part of the mi. but so are SaaS 3software as a
ser)ice4 pro)iders such as Salesforce.com. Today for the most part !T must
plug into cloud*based ser)ices indi)idually but cloud computing aggregators
and integrators are already emerging.
:

Chapter 2
Cloud computing- The Concept
Clo34 comp3tin5 is !nternet 3/cloud/4 based de)elopment and use of computer
technology 3/computing/4. !t is a style of computing in which dynamically scalable and
often )irtuali-ed resources are pro)ided as a ser)ice o)er the !nternet. Users need not
ha)e +nowledge of e.pertise in or control o)er the technology infrastructure /in the
cloud/ that supports them
.
The concept incorporates infrastructure as a ser)ice >IaaS? platform as a ser)ice >PaaS?
and software as a ser)ice >SaaS? as well as ;eb '.( and other recent technology trends
which ha)e the common theme of reliance on the !nternet for satisfying the computing
needs of the users. $.amples of SaaS )endors include Salesforce.com and <oogle 5pps
which pro)ide common business applications online that are accessed from a web
browser while the software and data are stored on the ser)ers.
The term cloud is used as a metaphor for the !nternet based on how the !nternet is
depicted in computer networ+ diagrams and is an abstraction for the comple.
infrastructure it conceals.
=

2.1 Comparison:
&loud computing is often confused with grid computing 3/a form of distributed
computing whereby a 8super and )irtual computer8 is composed of a cluster of networ+ed
loosely*coupled computers acting in concert to perform )ery large tas+s/4 utility
computing 3the /pac+aging of computing resources such as computation and storage as
a metered ser)ice similar to a traditional public utility such as electricity/4 and autonomic
computing 3/computer systems capable of self*management/4.
!ndeed many cloud computing deployments as of '((> depend on grids ha)e autonomic
characteristics and bill li+e utilities ? but cloud computing can be seen as a natural ne.t
step from the grid*utility model. Some successful cloud architectures ha)e little or no
centrali-ed infrastructure or billing systems whatsoe)er including peer*to*peer networ+s
li+e #it Torrent and S+ype and )olunteer computing li+e
2.2 Implementation:
The ma"ority of cloud computing infrastructure as of '((> consists of reliable ser)ices
deli)ered through data centers and built on ser)ers with different le)els of )irtuali-ation
technologies. The ser)ices are accessible anywhere that has access to networ+ing
infrastructure. The Cloud appears as a single point of access for all the computing needs
of consumers. &ommercial offerings need to meet the quality of ser)ice requirements of
customers and typically offer ser)ice le)el agreements. 7pen standards are critical to the
@

growth of cloud computing and open source software has pro)ided the foundation for
many cloud computing implementations.
2.3 Characteristics:
5s customers generally do not own the infrastructure they merely access or rent they
can a)oid capital e.penditure and consume resources as a ser)ice paying instead for
what they use. Many cloud*computing offerings ha)e adopted the utility computing
model which is analogous to how traditional utilities li+e electricity are consumed while
others are billed on a subscription basis. Sharing /perishable and intangible/ computing
power among multiple tenants can impro)e utili-ation rates as ser)ers are not left idle
which can reduce costs significantly while increasing the speed of application
de)elopment. 5 side effect of this approach is that /computer capacity rises dramatically/
as customers do not ha)e to engineer for pea+ loads. 5doption has been enabled by
/increased high*speed bandwidth/ which ma+es it possible to recei)e the same response
times from centrali-ed infrastructure at other sites.
2.4 Economics:
&loud computing users can a)oid capital e.penditure 3&ap$.4 on hardware software and
ser)ices rather paying a pro)ider only for what they use. &onsumption is billed on a
utility 3e.g. resources consumed li+e electricity4 or subscription 3e.g. time based li+e a
newspaper4 basis with little or no upfront cost. 7ther benefits of this time sharing style
approach are low barriers to entry shared infrastructure and costs low management
o)erhead and immediate access to a broad range of applications. Users can generally
terminate the contract at any time 3thereby a)oiding return on in)estment ris+ and
uncertainty4 and the ser)ices are often co)ered by ser)ice le)el agreements with financial
penalties.
5ccording to Nicholas &arr the strategic importance of information technology is
diminishing as it becomes standardi-ed and cheaper. Ae argues that the cloud computing
paradigm shift is similar to the displacement of electricity generators by electricity grids
early in the '(th century.
B

2.5 Companies:
,ro)iders including 5ma-on Microsoft <oogle Sun and 9ahoo e.emplify the use of
cloud computing. !t is being adopted by indi)idual users through large enterprises
including <eneral $lectric C87r6al and ,rocter % <amble.
Chapter 3
istor!
The Cloud is a term with a long history in telephony which has in the past decade been
adopted as a metaphor for internet based ser)ices with a common depiction in networ+
diagrams as a cloud outline.
The underlying concept dates bac+ to 1>=( when Dohn Mc&arthy opined that
/computation may someday be organi-ed as a public utility/E indeed it shares
characteristics with ser)ice bureaus which date bac+ to the 1>=(s. The term cloud had
already come into commercial use in the early 1>>(s to refer to large 5TM networ+s. #y
the turn of the '1st century the term /cloud computing/ had started to appear although
most of the focus at this time was on Software as a ser)ice 3SaaS4.
!n 1>>> Salesforce.com was established by Marc #enioff ,ar+er Aarris and his fellows.
They applied many technologies of consumer web sites li+e <oogle and 9ahooF to
business applications. They also pro)ided the concept of /7n demand/ and /SaaS/ with
>

their real business and successful customers. The +ey for SaaS is being customi-able by
customer alone or with a small amount of help. Gle.ibility and speed for application
de)elopment ha)e been drastically welcomed and accepted by business users.
!#M e.tended these concepts in '((1 as detailed in the 5utonomic &omputing
Manifesto ** which described ad)anced automation techniques such as self*monitoring
self*healing self*configuring and self*optimi-ing in the management of comple. !T
systems with heterogeneous storage ser)ers applications networ+s security
mechanisms and other system elements that can be )irtuali-ed across an enterprise.
5ma-on.com played a +ey role in the de)elopment of cloud computing by moderni-ing
their data centers after the dot*com bubble and ha)ing found that the new cloud
architecture resulted in significant internal efficiency impro)ements pro)iding access to
their systems by way of 5ma-on ;eb Ser)ices in '((: on a utility computing basis.
'((@ saw increased acti)ity with <oogle !#M and a number of uni)ersities embar+ing
on a large scale cloud computing research pro"ect around the time the term started
gaining popularity in the mainstream press. !t was a hot topic by mid*'((B and numerous
cloud computing e)ents had been scheduled.
!n 5ugust '((B <artner Hesearch obser)ed that /organi-ations are switching from
company*owned hardware and software assets to per*use ser)ice*based models/ and that
the /pro"ected shift to cloud computing will result in dramatic growth in !T products in
some areas and in significant reductions in other areas./
Chapter 4
"olitical Issues
The Cloud spans many borders and /may be the ultimate form of globali-ation./ 5s such
it becomes sub"ect to comple. geopolitical issues: pro)iders must satisfy myriad
regulatory en)ironments in order to deli)er ser)ice to a global mar+et. This dates bac+ to
the early days of the !nternet where libertarian thin+ers felt that /cyberspace was a
distinct place calling for laws and legal institutions of its own/E author Neal Stephenson
en)isaged this as a tiny island data ha)en called Iina+uta in his classic science*fiction
no)el &ryptonomicon.
Despite efforts 3such as US*$U Safe Aarbor4 to harmoni-e the legal en)ironment as of
'((> pro)iders such as 5ma-on ;eb Ser)ices cater to the ma"or mar+ets 3typically the
United States and the $uropean Union4 by deploying local infrastructure and allowing
customers to select /a)ailability -ones./ Nonetheless there are still concerns about
1(

security and pri)acy from indi)idual through go)ernmental le)el e.g. the US5
,5TH!7T 5ct and use of national security letters and the $lectronic &ommunication
,ri)acy 5ct8s Stored Communications Act.
Chapter 5
#egal Issues
!n March '((@ Dell applied to trademar+ the term /cloud computing/ 3U.S. Trademar+
@@10>(B'4 in the United States. The /Notice of 5llowance/ it recei)ed in Duly '((B was
canceled on 5ugust = resulting in a formal re"ection of the trademar+ application less
than a wee+ later.
7n 0( September '((B US,T7 issued a /Notice of 5llowance/ to &<acti)e CC& 3U.S.
Trademar+ @@0::'B@4 for /&loud7S/. 5 cloud operating system is a generic operating
system that /manageJsK the relationship between software inside the computer and on the
11

;eb/ such as Microsoft 5-ure. <ood 7S CC& also announced their /&loud/ operating
system on 1 December '((B.
Hichard Stallman founder of the Gree Software Goundation belie)es that cloud
computing endangers liberties because users sacrifice their pri)acy and personal data to a
third party. !n No)ember '((@ the Gree Software Goundation released the 5ffero <eneral
,ublic Cicense a )ersion of <,C)0 designed to close a percei)ed legal loophole
associated with free software designed to be run o)er a networ+ particularly software as
a ser)ice. 5n application ser)ice pro)ider is required to release any changes they ma+e to
5ffero <,C open source code.
Chapter $
%is& 'itigation
&orporations or end*users wishing to a)oid not being able to access their data ? or e)en
losing it ? should research )endors8 policies on data security before using )endor
ser)ices. 7ne technology analyst and consulting firm <artner lists se)en security issues
which one should discuss with a cloud*computing )endor:
/( ,ri)ileged user access?inquire about who has speciali-ed access to data and
about the hiring and management of such administrators.
.( Hegulatory compliance?ma+es sure a )endor is willing to undergo e.ternal
audits andLor security certifications.
1'

9( Data locations?as+ if a pro)ider allows for any control o)er the location of data.
*( Data segregation?ma+e sure that encryption is a)ailable at all stages and that
these /encryption schemes were designed and tested by e.perienced
professionals/.
2( Heco)ery?find out what will happen to data in the case of a disasterE do they
offer complete restoration and if so how long that would ta+e.
7( !n)estigati)e Support?inquire whether a )endor has the ability to in)estigate any
inappropriate or illegal acti)ity.
,( Cong*term )iability?as+ what will happen to data if the company goes out of
businessE how will data be returned and in what format.
!n practice one can best determine data*reco)ery capabilities by e.periment: as+ing to
get bac+ old data seeing how long it ta+es and )erifying that the chec+sums match the
original data. Determining data security is harder. 5 tactic not co)ered by <artner is to
encrypt the data yourself. !f you encrypt the data using a trusted algorithm then
regardless of the ser)ice pro)ider8s security and encryption policies the data will only be
accessible with the decryption +eys. This leads to a follow*on problem: managing pri)ate
+eys in a pay*on*demand computing infrastructure.
Chapter 7
Key characteristics
Cost is greatly reduced and capital e.penditure is con)erted to operational
e.penditure. This lowers barriers to entry as infrastructure is typically pro)ided
by a third*party and does not need to be purchased for one*time or infrequent
intensi)e computing tas+s. ,ricing on a utility computing basis is fine*grained
with usage*based options and minimal or no !T s+ills are required for
implementation.
Device and location independence enable users to access systems
using a web browser regardless of their location or what de)ice they are using
e.g. ,& mobile. 5s infrastructure is off*site 3typically pro)ided by a third*party4
and accessed )ia the !nternet the users can connect from anywhere.
Multi-tenancy enables sharing of resources and costs among a large pool of
users allowing for:
o Centrali@ation of infrastructure in areas with lower costs 3such as real
estate electricity etc.4
10

o Pea=1loa4 capacit; increases 3users need not engineer for highest
possible load*le)els4
o Utili@ation an4 e##icienc; impro)ements for systems that are often only
1(*'(M utili-ed.
Reliability impro)es through the use of multiple redundant sites which ma+es
it suitable for business continuity and disaster reco)ery. Nonetheless most ma"or
cloud computing ser)ices ha)e suffered outages and !T and business managers are
able to do little when they are affected.
Scalability )ia dynamic 3/on*demand/4 pro)isioning of resources on a fine*
grained self*ser)ice basis near real*time without users ha)ing to engineer for
pea+ loads. ,erformance is monitored and consistent and loosely*coupled
architectures are constructed using web ser)ices as the system interface.
Security typically impro)es due to centrali-ation of data increased security*
focused resources etc. but raises concerns about loss of control o)er certain
sensiti)e data. Security is often as good as or better than traditional systems in
part because pro)iders are able to de)ote resources to sol)ing security issues that
many customers cannot afford. ,ro)iders typically log accesses but accessing the
audit logs themsel)es can be difficult or impossible.
Sustainability comes about through impro)ed resource utili-ation more
efficient systems and carbon neutrality. Nonetheless computers and associated
infrastructure are ma"or consumers of energy.
Chapter (
Components
Clo34 comp3tin5 Component8
5pplications Gaceboo+ A <oogle 5pps A SalesGorce A Microsoft 7nline
&lient
#rowser3&hrome4 A Girefo. A &loud A Mobile 35ndroid A i,hone4 A Netboo+
3$ee,& A MS! ;ind4 A Nettop 3&herry,al A Nonbu4
!nfrastructur
e
#itTorrent A $&' A <o<rid A Sun <rid A 0tera
,latforms 5pp $ngine A 5-ure A Mosso A SalesGorce
Ser)ices 5le.a A G,S A MTur+ A SOS
Storage S0 A SimpleD# A SOC Ser)ices
11

Standards 5"a. A 5tom A ATMC : A H$ST
(.1 A pplication
5 cloud application le)erages the &loud in software architecture often eliminating the
need to install and run the application on the customer8s own computer thus alle)iating
the burden of software maintenance ongoing operation and support. !or eBample:
,eer*to*peer L )olunteer computing 3#ittorrent #7!N& ,ro"ects S+ype4
;eb application 3Gaceboo+4
Software as a ser)ice 3<oogle 5pps S5, and Salesforce4
Software plus services (Microsoft Online Services)
(.2 Client
5 cloud client consists of computer hardware andLor computer software which relies on
clo34 comp3tin5 for application deli)ery or which is specifically designed for deli)ery
of cloud ser)ices and which in either case is essentially useless without it. !or eBample:
1:

Mobile 35ndroid i,hone ;indows Mobile4
Thin client 3&herry,al Nonbu g7S*based systems4
Thic+ client L ;eb browser 3<oogle &hrome Mo-illa Girefo.4
(.3 In)rastructure
Cloud infrastructure such as !nfrastructure as a ser)ice is the deli)ery of computer
infrastructure typically a platform )irtuali-ation en)ironment as a ser)ice. !or eBample:
Gull )irtuali-ation 3<o<rid S+ytap4
Management 3HightScale4
&ompute 35ma-on $lastic &ompute &loud4
,latform 3Gorce.com4
(.4 "lat)orm
5 cloud platform such as ,latform as a ser)ice the deli)ery of a computing platform
andLor solution stac+ as a ser)ice facilitates deployment of applications without the cost
and comple.ity of buying and managing the underlying hardware and software layers.
!or eBample:
;eb application framewor+s
o ,ython D"ango 3<oogle 5pp $ngine4
o Huby on Hails 3Aero+u4
o .N$T 35-ure Ser)ices ,latform4
;eb hosting 3Mosso4
,roprietary 3Gorce.com4
(.5 *er+ice
5 cloud service includes /products ser)ices and solutions that are deli)ered and
consumed in real*time o)er the !nternet/. Gor e.ample ;eb Ser)ices 3/software
systemJsK designed to support interoperable machine*to*machine interaction o)er a
1=

networ+/4 which may be accessed by other cloud computing components software e.g.
Software plus ser)ice or end users directly. Speci#ic eBample8 incl34e:
!dentity 375uth 7pen!D4
!ntegration 35ma-on Simple Oueue Ser)ice4
,ayments 35ma-on Gle.ible ,ayments Ser)ice <oogle &hec+out ,ay,al4
Mapping 3<oogle Maps 9ahooF Maps4
Search 35le.a <oogle &ustom Search 9ahooF #7SS4
7thers 35ma-on Mechanical Tur+4
(.$ *torage
Cloud storage in)ol)es the deli)ery of data storage as a ser)ice including database*li+e
ser)ices often billed on a utility computing basis e.g. per gigabyte per month( !or
eBampleC
Database 35ma-on SimpleD# <oogle 5pp $ngine8s #igTable datastore4
Networ+ attached storage 3MobileMe iDis+ Nir)ani. &loudN5S4
Synchroni-ation 3Ci)e Mesh Live Desktop component MobileMe push functions4
;eb ser)ice 35ma-on Simple Storage Ser)ice Nir)ani. SDN4
Chapter ,
Architecture
Cloud architecture the systems architecture of the software systems in)ol)ed in the
deli)ery of cloud computing comprises hardware and software designed by a cloud
architect who typically wor+s for a cloud integrator. !t typically in)ol)es multiple cloud
components communicating with each other o)er application programming interfaces
usually web ser)ices.
This closely resembles the UN!P philosophy of ha)ing multiple programs doing one
thing well and wor+ing together o)er uni)ersal interfaces. &omple.ity is controlled and
the resulting systems are more manageable than their monolithic counterparts.
Cloud architecture e.tends to the client where web browsers andLor software
applications access cloud applications.
1@

Cloud storage architecture is loosely coupled where metadata operations are centrali-ed
enabling the data nodes to scale into the hundreds each independently deli)ering data to
applications or user.
Chapter 1-
T!pes
1B

1-.1 "ublic cloud
Public cloud or external cloud describes cloud computing in the traditional mainstream
sense whereby resources are dynamically pro)isioned on a fine*grained self*ser)ice
basis o)er the !nternet )ia web applicationsLweb ser)ices from an off*site third*party
pro)ider who shares resources and bills on a fine*grained utility computing basis.
1-.2 "ri+ate cloud
Private cloud and internal cloud are neologisms that some )endors ha)e recently used to
describe offerings that emulate cloud computing on pri)ate networ+s. These products
claim to /deli)er some benefits of cloud computing without the pitfalls/ capitali-ing on
data security corporate go)ernance and reliability concerns.
;hile an analyst predicted in '((B that pri)ate cloud networ+s would be the future of
corporate !T there is some uncertainty whether they are a reality e)en within the same
firm. 5nalysts also claim that within fi)e years a /huge percentage/ of small and medium
enterprises will get most of their computing resources from e.ternal cloud computing
pro)iders as they /will not ha)e economies of scale to ma+e it worth staying in the !T
business/ or be able to afford pri)ate clouds.
1>

The term has also been used in the logical rather than physical sense for e.ample in
reference to platform as ser)ice offerings though such offerings including Microsoft8s
5-ure Ser)ices ,latform are not a)ailable for on*premises deployment.
1-.3 !brid cloud
5 hybrid cloud en)ironment consisting of multiple internal andLor e.ternal pro)iders

/will
be typical for most enterprises/.
Chapter 11
%oles
11.1 "ro+ider
'(

5 cloud computing provider or cloud computing service provider owns and operates li)e
cloud computing systems to deli)er ser)ice to third parties. The barrier to entry is also
significantly higher with capital e.penditure required and billing and management creates
some o)erhead. Nonetheless significant operational efficiency and agility ad)antages can
be reali-ed e)en by small organi-ations and ser)er consolidation and )irtuali-ation
rollouts are already well underway. 5ma-on.com was the first such pro)ider moderni-ing
its data centers which li+e most computer networ+s were using as little as 1(M of its
capacity at any one time "ust to lea)e room for occasional spi+es. This allowed small
fast*mo)ing groups to add new features faster and easier and they went on to open it up
to outsiders as 5ma-on ;eb Ser)ices in '((' on a utility computing basis.
11.2 .ser
5 user is a consumer of cloud computing. The pri)acy of users in cloud computing has
become of increasing concern. The rights of users are also an issue which is being
addressed )ia a community effort to create a bill of rights.
11.3 /endor
5 )endor sells products and ser)ices that facilitate the deli)ery adoption and use of cloud
computing.

Gor e.ample:
&omputer hardware 3Dell A, !#M Sun Microsystems4
o Storage 3Sun Microsystems $M& !#M4
o !nfrastructure 3&isco Systems4
&omputer software 30tera Aadoop !#M HightScale4
o 7perating systems 3Solaris 5!P Cinu. including Hed Aat4
o ,latform )irtuali-ation 3&itri. Microsoft QMware Sun .QM !#M4
Chapter 12
*tandards
Cloud standards a number of e.isting typically lightweight open standards ha)e
facilitated the growth of cloud computing including:
'1

Application
o &ommunications 3ATT, PM,,4
o Security 375uth 7pen!D SSCLTCS4
o Syndication 35tom4
Client
o #rowsers 35D5P4
o 7ffline 3ATMC :4
Implementation8
o Qirtuali-ation 37QG4
Plat#orm
o Solution stac+s 3C5M,4
SerDice
o Data 3PMC DS7N4
o ;eb Ser)ices 3H$ST4
Stora5e
o Database35ma-on Simple D# <oogle 5pp $ngine #igTable Datastore4
o Networ+ attached storage 3MobileMe iDis+ Nir)ani. &loudN5S4
o Synchroni-ation 3Ci)e Mesh Live Desktop component MobileMe push
functions4
o ;eb ser)ice 35ma-on Simple Storage Ser)ice Nir)ani. SDN4
Chapter 13
Case *tud!
''

13.1 Ama0on EC2
Ama0on Elastic Compute Cloud 3also +nown as /$&'/4 is a commercial
web ser)ice that allows customers to rent computers on which to run their own
computer applications. $&' allows scalable deployment of applications by pro)iding
a web ser)ices interface through which a customer can create )irtual machines i.e.
ser)er instances on which the customer can load any software of his choice. 5
customer can create launch and terminate ser)er instances as needed paying by the
hour for acti)e ser)ers hence the term /elastic/. 5 customer can set up ser)er
instances in -ones insulated from each other for most failure causes so that one may
be a bac+up for the other and minimi-e down time. 5ma-on.com pro)ides $&' as one
of se)eral web ser)ices mar+eted under the blan+et term 5ma-on ;eb Ser)ices
35;S4.
History
5ma-on announced a limited public beta of $&' on 5ugust ': '((=. 5ccess to $&'
was granted on a first come first ser)ed basis. $&' became generally a)ailable on
7ctober '0 '((B along with support for Microsoft ;indows Ser)er.
Virtual machines
$&' uses Pen )irtuali-ation. $ach )irtual machine called an /instance/ functions as
a )irtual pri)ate ser)er in one of three si-esE small large or e.tra large. 5ma-on.com
si-es instances based on /$&' &ompute Units/ ? the equi)alent &,U capacity of
physical hardware. 7ne $&' &ompute Unit equals 1.(*1.' <A- '((@ 7pteron or
'((@ Peon processor. The system offers the following instance types:
Small !nstance
The small instance 3default4 equates to /a system with 1.@ <# of memory 1
$&' &ompute Unit 31 )irtual core with 1 $&' &ompute Unit4 1=( <# of instance
storage 0'*bit platform/
Carge !nstance
The large instance represents /a system with @.: <# of memory 1 $&'
&ompute Units 3' )irtual cores with ' $&' &ompute Units each4 B:( <# of
instance storage =1*bit platform/.
$.tra Carge !nstance
The e.tra large instance offers the /equi)alent of a system with 1: <# of
memory B $&' &ompute Units 31 )irtual cores with ' $&' &ompute Units each4
1=>( <# of instance storage =1*bit platform./
Aigh*&,U !nstance
'0

!nstances of this family ha)e proportionally more &,U resources than memory
3H5M4 and address compute*intensi)e applications.
Aigh*&,U Medium !nstance
!nstances of this family ha)e the following configuration:
1.@ <# of memory
: $&' &ompute Units 3' )irtual cores with '.: $&' &ompute Units each4
0:( <# of instance storage
0'*bit platform
!L7 ,erformance: Moderate
Aigh*&,U $.tra Carge !nstance
!nstances of this family ha)e the following configuration:
@ <# of memory
'( $&' &ompute Units 3B )irtual cores with '.: $&' &ompute Units each4
1=>( <# of instance storage
=1*bit platform
!L7 ,erformance: Aigh
Pricing
5ma-on charges customers in two primary ways:
Aourly charge per )irtual machine
Data transfer charge
The hourly )irtual machine rate is fi.ed based on the capacity and features of the )irtual
machine. 5ma-on ad)ertising describes the pricing scheme as /you pay for resources you
consume/ but defines resources such that an idle )irtual machine is consuming
resources as opposed to other pricing schemes where one would pay for basic resources
such as &,U time.
&ustomers can easily start and stop )irtual machines to control charges with 5ma-on
measuring with one hour granularity. Some are thus able to +eep each )irtual machine
running near capacity and effecti)ely pay only for &,U time actually used.
5s of March '((> 5ma-on8s time charge is about R@0Lmonth for the smallest )irtual
machine without ;indows and twel)e times that for the largest one running ;indows.
The data transfer charge ranges from R.1( to R.1@ per gigabyte depending on the
direction and monthly )olume.5ma-on does not ha)e monthly minimums or account
maintenance charges.
perating systems
;hen it launched in 5ugust '((= the $&' ser)ice offered Cinu. and later Sun
Microsystems8 7penSolaris and Solaris $.press &ommunity $dition. !n 7ctober '((B
'1

$&' added the ;indows Ser)er '((0 operating system to the list of a)ailable operating
systems.
,lans are in place for the $ucalyptus interface for the 5ma-on 5,! to be pac+aged into
the standard Ubuntu distribution.
Persistent Storage
5ma-on.com pro)ides persistent storage in the form of $lastic #loc+ Storage 3$#S4.
Users can set up and manage )olumes of si-es from 1<# to 1T#. The ser)ers can attach
these instances of $#S to one ser)er at a time in order to maintain data storage by the
ser)ers
13.2 *ales)orce.com
Sale8#orce(com 3N9S$: &HM4 is a )endor of &ustomer Helationship Management
3&HM4 solutions which it deli)ers to businesses o)er the internet using the software as a
ser)ice model
1rigins
Salesforce.com was founded in 1>>> by former 7racle e.ecuti)e Marc #enioff. !n Dune
'((1 the company went public on the New 9or+ Stoc+ $.change under the stoc+ symbol
&HM. !nitial in)estors in salesforce.com were Marc #enioff Carry $llison Aalsey
Minor Magdalena 9esil and !gor Sill <ene)a Qenture ,artners.
Current status
Salesforce.com is headquartered in San Grancisco &alifornia with regional headquarters
in Dublin 3co)ering $urope Middle $ast and 5frica4 Singapore 3co)ering 5sia ,acific
':

less Dapan4 and To+yo 3co)ering Dapan4. 7ther ma"or offices are in Toronto New 9or+
Condon Sydney and San Mateo &alifornia. Salesforce.com has its ser)ices translated
into 1: different languages and currently has 10=(( customers and o)er 1((((((
subscribers. !n '((B Salesforce.com ran+ed 10rd on the list of largest software
companies in the world.
Gollowing the Gederal ta+eo)er of Greddie Mac and Gannie Mae in September '((B the
S%, :(( remo)ed the two mortgage giants after ;ednesday September 1( '((B and
added Gastenal and Salesforce.com to the inde. effecti)e after Griday September 1'
'((B.
Products and Services
Customer %elationship 'anagement
Salesforce.com8s &HM solution is bro+en down into se)eral applications: Sales Ser)ice
% Support ,artner Helationship Management Mar+eting &ontent !deas and 5nalytics.
2orce.com "lat)orm
Salesforce.com8s ,latform*as*a*Ser)ice product is +nown as the Gorce.com ,latform. The
platform allows e.ternal de)elopers to create add*on applications that integrate into the
main Salesforce application and are hosted on salesforce.com8s infrastructure.
These applications are built using 5pe. 3a proprietary Da)a*li+e programming language
for the Gorce.com ,latform4 and Qisualforce 3an PMC*li+e synta. for building user
interfaces in ATMC 5D5P or Gle.4.
AppE3change
Caunched in '((: 5pp$.change is a directory of applications built for Salesforce by
third*party de)elopers which users can purchase and add to their Salesforce en)ironment.
5s of September '((B there are o)er B(( applications a)ailable from o)er 1:( !SQs.

Customi0ation
Salesforce users can customi-e their &HM application. !n the system there are tabs such
as /&ontacts/ /Heports/ and /5ccounts/. $ach tab contains associated information. Gor
e.ample /&ontacts/ has fields li+e Girst Name Cast Name $mail etc.
'=

&ustomi-ation can be done on each tab by adding user*defined custom fields.
&ustomi-ation can also be done at the /platform/ le)el by adding customi-ed applications
to a Salesforce.com instance that is adding sets of customi-ed L no)el tabs for specific
)ertical* or function*le)el 3Ginance Auman Hesources etc4 features.
4eb *er+ices
!n addition to the web interface Salesforce offers a ;eb Ser)ices 5,! that enables
integration with other systems
13.3 C#1.5 61perating *!stem7
Clo34 is a EFroG8er Fa8e4 Operatin5 8;8tem/ created by HGoo4 OS LLCI a Cos
5ngeles*based corporation. The company initially launched a Cinu. distribution called
g7S which is based on Ubuntu now in its third incarnation
!ro"ser and perating System
&loud is a combination of a simplified operating system that runs "ust a web browser
pro)iding access to a )ariety of web*based applications that allow the user to perform
many simple tas+s without booting a full*scale operating system. #ecause of its
simplicity &loud can boot in "ust a few seconds. The operating system is designed for
Netboo+s Mobile !nternet De)ices and ,&s that are mainly used to browse the !nternet.
Grom &loud the user can quic+ly boot into the main 7S because &loud continues booting
the main 7S in the bac+ground.
&ombining a browser with a basic operating system also allows the use of cloud
computing in which applications and data /li)e and run/ on the !nternet instead of on the
hard dri)e.
&loud can be installed and used together with other operating systems or can act as a
standalone operating system. ;hen used as a standalone operating system hardware
requirements are relati)ely low.
5t the moment &loud is only officially a)ailable built into the <!<5#9T$ M>1' Touch
Screen Netboo+ but a ,ri)ate #eta test is currently 3early Gebruary '((>4 running.
'@

Reception
$arly re)iews compared the operating system8s user interface to Mac 7S P and noted the
similarity of its browser to <oogle &hrome although it is actually based on a modified
Mo-illa Girefo. browser.
Chapter 14
'B

Conclusion
&loud &omputing is a )ast topic and the abo)e report does not gi)e a high le)el
introduction to it. !t is certainly not possible in the limited space of a report to do "ustice
to these technologies. ;hat is in store for this technology in the near futureS ;ell &loud
&omputing is leading the industry2s endea)or to ban+ on this re)olutionary technology.
&loud &omputing #rings ,ossibilitiesTT..
!ncreases business responsi)eness
5ccelerates creation of new ser)ices )ia rapid prototyping capabilities
Heduces acquisition comple.ity )ia ser)ice oriented approach
Uses !T resources efficiently )ia sharing and higher system utili-ation
Heduces energy consumption
Aandles new and emerging wor+loads
Scales to e.treme wor+loads quic+ly and easily
Simplifies !T management
,latform for collaboration and inno)ation
&ulti)ates s+ills for ne.t generation wor+force
Today with such cloud*based interconnection seldom in e)idence cloud computing
might be more accurately described as /s+y computing/ with many isolated clouds of
ser)ices which !T customers must plug into indi)idually. 7n the other hand as
)irtuali-ation and S75 permeate the enterprise the idea of loosely coupled ser)ices
running on an agile scalable infrastructure should e)entually ma+e e)ery enterprise a
node in the cloud. !t8s a long*running trend with a far*out hori-on. #ut among big
metatrends cloud computing is the hardest one to argue with in the long term.
&loud &omputing is a technology which too+ the software and business world by storm.
The much deser)ed hype o)er it will continue for years to come.
'>

Chapter 15
%e)erences
J1K. www.wi+ipedia.com
J'K. www.infoworld.comLarticleL(BL(1L(@L1:G$*cloud*computing*realityU1.html
J0K. www.wi+i.cloudcommunity.orgLwi+iL&loud&omputing:#illUofUHights
J1K. www.da)idchappell.comL&loud,latforms**&happell. ,DG
J:K. www.ama-on.com
J=K. www.thin+gos.comLcloudLinde..html
J@K. www.salesforce.com
JBK. www.google.com
J>K. &hip &omputer Maga-ine December '((B * Geb '((> $dition
0(

01