Unit-3 Cloud Computing Architecture and Management 3.

Marketed by:

Unit Cloud Computing

architecture and

3 management SIA GROUP

Part-A
Short Questions with Solutions
Q1. Write in short about cloud architecture.
Answer : Model Paper-I, Q1(e)

The architecture of any technological model depends on how the model works which is nothing but a hierarchical view
in describing the technology. Similarly, the cloud also has its own architecture which shows its working mechanisms. These
mechanisms operate upon the components and the dependencies present in it. So, it is confirm that the operation of cloud is
absolutely based on the Internet.
Ideally, the architecture of the cloud can be categorized into 4 layers depending upon how the user uses the cloud. They
are,
1. Layer 1 (User/Client Layer)
2. Layer 2 (Network Layer)
3. Layer 3 (Cloud Management Layer)
4. Layer 4 (Hardware Resource Layer).
Q2. Describe layer 1 (user/client layer) of cloud architecture.
Answer : Model Paper-II, Q1(e)

The user/client layer is the first layer in the cloud architecture. This layer holds all users and clients. Here, only both
users and clients starts the connection to the cloud. In particular, the client device could be thin client, thick client, mobile or any
handheld device. These clients provide rudimentary functionalities so as to form access to the web application.
The thin clients can be described as those devices which entirely depends on other system for its operations. In essence,
they hold very low processing capability, thats why thy depends upon other systems. On the other hand, the thick clients are very
general computers which holds adequate processing capability. They can sustain independent operations through their adequate
processing capability.
Q3. Write short notes on the following from anatomy of cloud,
(i) Application
(ii) Platform. Model Paper-III, Q1(e)

Answer :
(i) Application
It is the upper layer and here all the applications are executed.
(ii) Platform
It is the second layer in the structure and lies between the infrastructure and the application. It is meant for executing the
application.

SIA Group
3.2 Cloud Computing [JNTU-hyderabad]
Q4. Write short notes on private cloud access networking.
Answer :
In private cloud deployment option, the technology and approaches stay local to the in-house network structure. This is
due to the fact that cloud is one of the part of organizational network. Ideally, the structure holds Internet VPN or VPN service
taken through network operator.
Here, the performance will not get affected if the application access is executed correctly. This execution should be done
within the organizational network, which is nothing but precloud configuration. In simple terms, the change or transition to private
cloud computing does not compromise the performance.
Q5. What are the new facets in private networks?
Answer :
The traditional private networks are developed for on-premise applications and to acquire atmost internet security. Here,
the applications including e-mail, file sharing and Enterprise Resource Planning (ERP) system are given to on-premise based
servers located at every single corporate data center. In particular software vendors, as an alternative offers Software-as-a-Service
(SaaS) to corporate offices. This alternative serves as an software support to them. However, this becomes more challenging while
accessing and using the software mechanisms from data center servers and the problems also pool in from network architectures.
Q6. What are the drawbacks of web applications?
Answer : Model Paper-I, Q1(f)

The drawbacks of web applications are as follows,

1. The web applications fail to manage heavy loads and varying loads. It is inelastic in nature.
2. They do not support multitenancy.
3. They fail to estimate the quantitative measurement of services despite of the fact that they can easily monitor the user.
4. They only substantiate themselves on one specific platform.
5. They do not operates on pay-as-you go basis. As a consequence any particular service provided to the user is certainly
permanent or trial use. They does not monitor the timings also.
6. They cannot handle highly loaded transaction because of its non-elastic nature.
These issues can be solved through cloud applications.
Q7. Briefly explain the multitenancy of cloud application.
Answer :
This is one of the unique properties of cloud which makes it different from other applications. In these applications, software
is used by multiple users independently. In simple terms, here the independence is concern with the logical independence.
It operates by employing separate applications instance and ensures that any modification made in one application is not
reflected to the other application. But, in physical terms the software is shared among the applications and it is not independent.
And, it is also important to note that the physical separation is very negligible and high emphasis is made on logical independence.
It substantiates any number of application. There is no limitations in the number. However, to some extent logical isolation
providings is highly dependent on physical isolations. In essence, physical closeness of any application entails difficulties in
acquiring multitenancy. It is known that web application and cloud application.
Q8. Briefly describe how to manage the cloud infrastructure.
Answer : Model Paper-II, Q1(f)

The infrastructure of the cloud is the main support system of the cloud. And it is used for gauging the Qos factor. In
particular, failure in managing the infrastructure can lead to the breakdown of the complete cloud. Subsequently, this can present
bad impacts on the QoS. The cloud management mainly operates upon the resource management which consists of multiple internal
tasks such as resource scheduling, provisioning and load balancing. Now, cloud service provider’s core software capabilities are
implemented to manage these tasks. The capabilities include cloud OS that facilitates services to the cloud which in turn controls
the internal structure of the cloud. Technically, the infrastructure of the cloud is very complicated system and compresses many
resources which are mostly shared by multiple users.

L o o k f o r t h e SIA Group logo o n t h e title cover before you buy

Unit-3 Cloud Computing Architecture and Management 3.3
Q9. Define migrating application to cloud.
Answer : Model Paper-III, Q1(f)

The cloud migration can be viewed as the process of transferring the multiple enterprise application and their corresponding
IT environments. These environments are originally placed on traditional hosting type and the transfer is done to the cloud
environment which could be public, private or hybrid.
With cloud migration, the user can get the opportunity to minimize the expenses incurred on applications. In doing so, the
user has to undergo different phases such as, evaluation, migration strategy, prototyping, provisioning and testing.

Q10. List various approaches for cloud migration and explain any one approach.
Answer :
Approaches for Cloud Migration
The vendors adopt following four broad approaches for implementing cloud migration. They are,
1. Migrate existing applications.
2. Initiate from scratch
3. Separate company
4. Buy an existing cloud vendor.
Migrate Existing Applications
In this approach, some of the applications are redeveloped and restructured. Here, advantages of certain virtualization
technologies are atmost utilized to speed-up the work process. However, for new functionalities, the organizations have to hire
top engineers. This task is accomplished after releasing many applications with respect to the timings set by the demand of the
customer.

SIA Group
3.4 Cloud Computing [JNTU-hyderabad]

Part-B
ESSAY Questions with Solutions

3.1 Cloud architecture, layer, anatomy of the cloud, network connectivity

in cloud computing

Q11. Draw and explain the architecture of cloud.

Answer : Model Paper-I, Q6

Cloud Architecture
The architecture of any technological model depends on how the model works which is nothing but a hierarchical view
in describing the technology. Similarly, the cloud also has its own architecture which shows its working mechanisms. These
mechanisms operate upon the components and the dependencies present in it. So, it is confirm that the operation of cloud is
absolutely based on the Internet.
Ideally, the architecture of the cloud can be categorized into 4 layers depending upon how the user uses the cloud. They are,
1. Layer 1 (User/Client Layer)
2. Layer 2 (Network Layer)
3. Layer 3 (Cloud Management Layer)
4. Layer 4 (Hardware Resource Layer).
1. Layer 1 (User/Client Layer)
The user/client layer is the first layer in the cloud architecture. This layer holds all users and clients. Here, only both
users and clients starts the connection to the cloud. In particular, the client device could be thin client, thick client, mobile or any
handheld device. These clients provide rudimentary functionalities so as to form access to the web application.
The thin clients can be described as those devices which entirely depends on other system for its operations. In essence,
they hold very low processing capability, thats why thy depends upon other systems. On the other hand, the thick clients are very
general computers which holds adequate processing capability. They can sustain independent operations through their adequate
processing capability.
It can be said that, the access of cloud application can be done in the same manner as that of web application. However,
the internal properties differ highly in cloud applications. Therefore, offering only client devices in this layer.
2. Layer 2 (Network Layer)
In this layer, the user forms connection to the cloud. The entire infrastructure of the cloud is based on the connection.
Through this connection, the customers acquire the services. This capability serves as Internet in public cloud, which is located
in a specific location. The location is however made abstract and user would be unaware of the location. But, its accessibility can
be done from any where across the world.
On the other hand, the connectivity in private cloud can be offered by local area network (LAN). In a way similar to the
others, here also the cloud is entirely based on the network which is being used. Regardless of the public or private cloud, the users
need least band width which is offered by cloud providers. Therefore, the network layer does not fall under the scope or range of
service-level agreements (SLAs). In simple terms, the SLA dose not considers, the internet connection happening between the
users and cloud to maintain the quality of service (QoS).
3. Layer 3 (Cloud Management Layer)
The cloud management layer comprises of softwares implemented in managing the cloud. In particular, the softwares are
cloud operating system (OS). This software serves as an interface between data center which consists of actual resources and the
user. Additionally, it also contains management software for managing the resources. Basically, these softwares are meant for
resource management i.e., scheduling and provisioning etc., optimization which includes server consolidation, storage workload
consolidation. And then, the software holds internal cloud governance. Unlike layer 2, the network layer falls under the scope of
SLAs, this is mostly set by users and service providers together.
However, it is reported as the violation of SLA if they make any delay while processing the service or leading to inconsistency
in service. In doing so, the service provider has to pay the penalty led by the protocols. The SLAs are meant not just to private
cloud, but also for public clouds. Some of the famous service providers are Amazon Web Services (AWS) and microsoft azure
for public cloud. In case of private cloud creation, deployment and management open stack and eucalyptus are very famous.

L o o k f o r t h e SIA Group logo o n t h e title cover before you buy

Unit-3 Cloud Computing Architecture and Management 3.5
4. Layer 4 (Hardware Resource Layer)
The hardware resource layer comprises provisions with respect to hardware resources. Unlike other layers, here the public
cloud holds data center at the back end. But, incase of private cloud the data center has a huge collection of hardware resources.
They are interconnected to each other situated in a particular location or a high configuration system. Also, this layer falls under
the scope of SLAs and is considered as the most important layer which controls the SLAs. Subsequently, it gives a great impact on
the data centers present in SLAs. The motive here is to grant the service, the moment user accesses the cloud. It is also important
that it should fall within the time anticipated by the SLAs. So, similar to the above layers, any inconsistencies in providing the
service may entail penalty of the service provider. Inorder to avoid this, the layer’s data centers operates on high-speed network
connection and also uses highly efficient algorithms. This supports in the transmission on data from center to the manager. Here
the number of data centers could be unlimited and numerous clouds can share those data centers.
This is the complete architecture of the cloud where in each layer is rigid and it is functional for any cloud applications.

Layer 1 : User/Client Layer

Layer 2 : Network Layer

Management

Layer 3 : Cloud Management Layer

Layer 4 : Hardware Resource Layer

Figure: Cloud Architecture

SIA Group
3.6 Cloud Computing [JNTU-hyderabad]
Q12. Discuss in detail the anatomy of the cloud.
Answer : Model Paper-II, Q6(a)

Anatomy of the Cloud

The anatomy of the cloud is nothing but the structure of the cloud. Unlike cloud architecture, the cloud anatomy also has
unique features which does not match with cloud architecture. The cloud anatomy is independent of any platform and has its
own dedicated technology to carryout operations. On the other hand, the architecture truely specifies the technology upon which
it works.
Ideally, the architecture can be specified as structural view which is largely concern with the technology upon which it is
dependent and with which it operates. It is also concern with those technologies which are dependent on it. It can be said that,
anatomy is one of the part of architecture. The figure given below is a basic structure of cloud. It can be further enhanced and
can be illustrated as the most standard anatomy becoming the foundation for the cloud. The anatomy of cloud comprises five
components. They are,
1. Application
2. Platform
3. Infrastructure
4. Virtualization
5. Physical hardware.
1. Application
It is the upper layer and here all the applications are executed.
2. Platform
It is the second layer in the structure and lies between the infrastrure and the application. It is meant for executing the
application.
3. Infrastructure
It is the third layer in the structure and comprises of resources used by other components. This layer enhances the users
computational capability.
4. Virtualization
It is the fourth layer in the structure. It is basically a procedure of generating logical components of resources across the
existing physical resources. The logical components are responsible for forming infrastructure and they are separated and
does not depends on any other component.
5. Physical Hardware
It is fifth layer in the structure and has server and storage units.

Application

Platform

Virtualized Infrastructure

Virtualization

Server/Storage/Datacenters

Figure: Cloud Structure

L o o k f o r t h e SIA Group logo o n t h e title cover before you buy

Unit-3 Cloud Computing Architecture and Management 3.7
Q13. Explain in detail the network connectivity in cloud computing.
Answer : Model Paper-III, Q6

Cloud computing can be viewed as a technique pertaining to resource sharing. Here, servers, storage and various other
computing infrastructure situated at various locations forms connection through the network. Each time the application is submitted
in the cloud for execution, necessary and suitable resources are taken from the storage of resources. After this, it is allocated to
the applications. Since, these resources are connected through internet the result which are generated are according to the users.
The network performance in cloud computing applications is considered as one of the major concerns effecting entire computing
performance.
Ideally, the cloud computing operates on various deployment options. The significance aspect concern with cloud deployment
models and its respective accessibilities is illustrated in the following network connectivity options.
1. Public cloud access networking
2. Private cloud access networking
3. Intracloud networking for public cloud services
4. Private intracloud networking
5. New facets in private networks
6. Path for Internet traffic.
1. Public Cloud Access Networking
In public cloud access networking, the devices forms connection via internet using certain cloud providers. These providers
substantiates Virtual Private Networks (VPNs) for the customers. The security while accessing the public cloud services is likely
to get compromise which eventually effects the performance. This problem can be addressed by promoting the connectivity using
encrypted tunnels. These tunnels transmits the information through a secure pipes using internet. Subsequently after some use,
this solution failed as it started giving overheads, delays in the connectivity which effected the performance.
So, the software engineers designed a suitable routing method which not just minimizes the delay but also provide security.
It operates by minimizing the delay through reducing transit hops occurring in the end-to-end connectivity between cloud provider
and cloud consumer. In particular, end-to-end connectivity is a complex federation with respect to interconnected providers.
This connectivity gains support through internet. Therefore, due to these reasons, it is important to examine the options prior to
selecting the paths.
2. Private Cloud Access Networking
In private cloud deployment option, the technology and approaches stay local to the in-house network structure. This is
due to the fact that cloud is one of the part of organizational network. Ideally, the structure holds Internet VPN or VPN service
taken through network operator.
Here, the performance will not get affected if the application access is executed correctly. This execution should be done
within the organizational network, which is nothing but precloud configuration. In simple terms, the change or transition to private
cloud computing does not compromise the performance.
3. Intracloud Networking for Public Cloud Services
The intracloud networking consideration in cloud computing is concern with public cloud services. The cloud provider
holds certain resources and the services the cloud to the customer highly depends upon those dements which are geographically
placed far from each other. Although, the distance does not impacts this, they stay connected through internet.
In particular, the visibility of the public cloud computing networks is internet to service provider. The user and customers
can not see them. Despite of such measures, security of connection and access mechanisms of resources are significant issue.
Apart from this, QoS also presents certain issues while connecting the resources across the world. Therefore, the problems related
to performance and violations are solved at commercial level by SLAs.

SIA Group
3.8 Cloud Computing [JNTU-hyderabad]
4. Private Intracloud Networking
The private intracloud networking is one of the most complex option for networking and connectivity with in the cloud
computing. This is due to the fact that, it is entirely based upon the quantity of the intracloud connections. These connections are
associated with those applications that are executed with in the environment. In particular, this type of networking is sustained
by connection between the huge data center sites run by the companies.
In doing so, at least case all the cloud implementations with respect is cloud computing depends upon intracloud networking.
This assist them in connecting the users to their respective resource suitable for their assigned applications. The moment, resources
are allocated the extent of utilization of intracloud networking highly depends upon the segregation application into small portions.
Now, this segregation may or may not be done based upon the service oriented architecture (SOA) between the many systems.
However, if the SOA is run according to its principle then, it can make the traffic travel between components of applications and
also between application can user. In doing so, the performance of cloud computing is highly compromised. This impact is caused
due to the difference occuring among the current application and the relationship of the network with the application.
5. New Facets in Private Networks
The traditional private networks are developed for on-premise applications and to acquire atmost internet security. Here,
the applications including e-mail, file sharing and Enterprise Resource Planning (ERP) system are given to on-premise based
servers located at every single corporate data center. In particular software vendors, as an alternative offers Software-as-a-Service
(SaaS) to corporate offices. This alternative serves as an software support to them. However, this becomes more challenging while
accessing and using the software mechanisms from data center servers and the problems also pool in from network architectures.
The global enterprises architecture could not optimize the performance with respect to cloud applications because the
design was too old. On the other hand, applications such as mission critical applications travel from on-premise base to cloud
based. Here, the availability and accessibility to the network stays as critical and the businesses does not run if applications such
as ERP and e-mails becomes inaccessible.
6. Path for Internet Traffic
The internet traffic has a conventional path setup and consists of less number of internet gateways. As a consequence, the
end users may face performance and availability issues while accessing cloud-based applications. However, the problem can be
alleviated by implementing more widely distributed internet gateway infrastructure and connectivity. This establishes an efficient
gateways for accessing applications and also provides lower latency access to the applications.
Once the traffic to cloud applications increases, the total amount of legacy network’s capacity with respect to the traffic also
increases. This occurs at regional gateways. Some of the applications which generates more bandwidth are video conferencing.
On the other hand, other applications such as mission-critical applications (like ERP) will consume less bandwidth.
This rises the need for netizens to plan accurate connectivity and path between providers and consumers.

3.2 applications on the cloud, managing the cloud, managing the cloud
infrastructure, managing the cloud application

Q14. How the power of computer is realized through applications? Explain.

Answer : Model Paper-II, Q6(b)

Applications on the Cloud

The applications are said to be the power of computer. In simple terms, the power of computer can be estimated through
the application. Among many types of applications, the first type of application is said to be stand-alone application. It is meant to
be run on a single system and it is independent of the network i.e., its functioning does not depend on the network. In particular,
they make use of the same machine upon which it is installed. They entirely depend upon the resources and features persent in
the system and independent of data or processing power of other systems. In short it can be said as self-sustaining. However, with
the increases requirements of users in terms of applications, the web applications came into picture.

L o o k f o r t h e SIA Group logo o n t h e title cover before you buy

Unit-3 Cloud Computing Architecture and Management 3.9
There was a huge difference between web applications and stand-alone application and stand-alone applications in various
aspects. The most important difference was the client-server architecture itself and these web applications were totally dependent
upon the network for the functioning. So consequently, it consists of two main components one is called client and the other one
is called server. The latter one is a high-end machine consisting of installed web applications which can be accessed through other
client systems. On the other hand, the former one is location independent and can be placed anywhere in the network, which helps
them to access the web applications. This turned out to be very useful and has become an extensive part of the day-to-day life.
however, despite of its full utilization there are many drawbacks connected to it. They are,

1. The web applications fail to manage heavy loads and varying loads. It is inelastic in nature.

2. They do not support multitenancy.

3. They fail to estimate the quantitative measurement of services despite of the fact that they can easily monitor the user.

4. They only substantiate themselves on one specific platform.

5. They do not operates on pay-as-you go basis. As a consequence, any particular service provided to the user is certainly
permanent or trial use. They does not monitor the timings also.

6. They can not handle highly loaded transaction because of its non-elastic nature.

These issues can be solved through cloud applications.

Stand-alone
Applications

Web
Applications

Cloud
Applications

Figure: Computer Application Evolution

SIA Group
3.10 Cloud Computing [JNTU-hyderabad]
Q15. What are the various features of cloud applications?
Answer : Model Paper-I, Q7(a)

Features of Cloud Application

The cloud application has unique features and does not share any properties with other applications. It can be accessed as
a web application but with different properties. Based on the studies made by NIST, some of the properties which distinguishes
cloud application from web application includes,
1. Multitenancy
2. Elasticity
3. Heterogeneous cloud platform
4. Quantitative measurement
5. On-demand service.

Multitenancy

Cloud
Application Elasticity
Features

Heterogeneous Cloud Platform

Quantitative Measurement

On-demand Service

Figure: Features of Cloud

1. Multitenancy
This is one of the unique properties of cloud which makes it different from other applications. In these applications, software
is used by multiple users independently. In simple terms, here the independence is concern with the logical independence.
It operates by employing separate applications instance and ensures that any modification made in one application is not
reflected to the other application. But, in physical terms the software is shared among the applications and it is not independent.
And, it is also important to note that the physical separation is very negligible and high emphasis is made on logical independence.
It substantiates any number of application. There is no limitations in the number. However, to some extent logical isolation
providings is highly dependent on physical isolations. In essence, physical closeness of any application entails difficulties in
acquiring multitenancy. It is known that web application and cloud application share similarities because the users implements it
in same manner. In the figure given below, the users share same applications.
User 1/tenant 1

User 2/tenant 2

Application
User 3/tenant 3

User 4/tenant 4

Figure: Multitenancy

L o o k f o r t h e SIA Group logo o n t h e title cover before you buy

Unit-3 Cloud Computing Architecture and Management 3.11
2. Elasticity
It is one of the unique properties of cloud which helps it for efficient functioning. Typically, elasticity can be described as
the ability of a system to stay extendible so as to adapt to variations occurring in the workload. It sustains this ability by allocating
and deallocating the resources automatically. As a result, at each point in time, the resources present meets the current demand
accurately.
This property also allows the efficient management of number of users i.e., from one user to several users at a time. Apart
from this, quick changes in load is also taken care due to this property i.e., increase and decrease of users and their related work
is sustained.
3. Heterogeneous Cloud Platform
This type of feature permits the cloud to support heterogeneity. In simple terms, the user can deploy any type of application
in the cloud. This property leaves the cloud flexible for the users and developers can easily deploy it. So, using web-browsers,
the users can access the deployed applications.
4. Quantitative Measurement
This property measurement is unique to cloud and is not available in web application. This property allows the quantitative
measurement or assessment of services. These services are acquired based on certain charges and the user has to pay the provider.
In simple terms, the application or resources serves as a utility on pay-per-use basis. This implies that the usage is not just
monitored but also measured. Apart from this, other measurable parameters include, other services, usage of link and also many
other parameters supporting cloud applications.
5. On-demand Service
This property provides on-demand service to the user. It is location independent and the user can utilize it regardless of
its location. So, the user can seamlessly access web-application without any limitations of time duration and the device used.
Q16. Discuss in detail two components of managing the cloud.
Answer : Model Paper-II, Q7

Managing the Cloud

The cloud management is crucial because it is implemented so as to maintain the quality of service and it is given top most
priority. The cloud is entirely based upon the manner in which it is managed. So, it is categorized into two components namely.
1. Managing the cloud infrastructure
2. Managing the cloud application.
1. Managing the Cloud Infrastructure
The infrastructure of the cloud is the main support system of the cloud. And it is used for gauging the Qos factor. In
particular, failure in managing the infrastructure can lead to the breakdown of the complete cloud. Subsequently, this can present
bad impacts on the QoS. The cloud management mainly operates upon the resource management which consists of multiple internal
tasks such as resource scheduling, provisioning and load balancing. Now, cloud service provider’s core software capabilities are
implemented to manage these tasks. The capabilities include cloud OS that facilitates services to the cloud which inturn controls
the internal structure of the cloud. Technically, the infrastructure of the cloud is very complicated system and compresses many
resources which are mostly shared by multiple users.
Management of resources in an inefficient manner may entails several adversaries i.e., effecting the performance, functionality
of the system. Also, mismanagement of resources can bring down the complete system. Thus, making the performance as basic
aspect of cloud as every part in the cloud depends upon the SLAs. Ideally, the SLAs is directly related to the performance. In
simple terms, the SLAs can be at optimum rate if and only if performance is good. On the other hand, the cloud functionality is
always provided and taken into account at any cost. But, any incosistency in facilitating the functionality can effect the entire
purpose of cloud management. And subsequently, if the cloud is functioning partially, then also the SLAs won’t give the correct
output.

SIA Group
3.12 Cloud Computing [JNTU-hyderabad]
The cost was one of the important criterias in the development of the cloud which includes the business prospects. If the cost
at service providers end is less in cloud management then the provider will minimize the price, this is necessary to attract strong
user base. It makes the users to use the service for improving the profit margin. On the other hand, when resource management
cost is high then eventually, the accessing cost of resource also becomes high. This never keeps the business in loss and the service
provider never bear the cost which means that the user has to make heavy payments. As a consequence, the service provider bear
adverse impact due to high cost. Then, the user may not use the service and the provider looses the wide user base. Thus leading
to a minimal growth in the industry and this presents big problem while dealing with the competitors. So, it is necessary to employ
efficient management which offers less to cost.

Apart from these issues, when the cloud operates at higher level, it invites more issues related to resource management.
These issues could be power consumption and optimization of multiple objectives for minimizing the cost. However, some
of the approaches such as consolidation of server and storage workloads can efficiently accomplishes these tasks. The role of
consolidation is to minimize the consumption of energy which eventually results in the performance increase of the cloud in
certain cases. Ideally, the consolidation of server is nothing but efficient utilization of computer server resources. The aim is to
minimize the total number of servers or its location required by the organizations.

Moreover, different service delivery models follows different management methods. These management methods are based
on load fluctuation. Ideally, load fluctuation can be defined as a point where in the workload with respect to the system varies
frequently. This is considered as an important criteria for cloud applications.

Load fluctuation can further be categorized into two types,

(a) Predictable

(b) Unpredictable.

(a) Predictable

These fluctuations can be handled in an easy manner. The preconfiguration of the cloud is necessary so as to handle the
fluctuations.

(b) Unpredictable

These fluctuations are difficult to handle. Despite of this, cloud is preferred by many users across the world.

2. Managing the Cloud Application

Many business companies are transferring their corporate applications on cloud platforms. This enhances the quickness
and easiness so as to meet up with huge requirements prevailing in the globalization of business. It also assists in generating
responsiveness towards the market demands. However, such transfer of applications can come up with its own complexities. In
simple terms, the application structure becomes highly composite and complex. They demand maximum utilization of capabilities
such as storage and database given by cloud providers. It also takes into account the capabilities of third party SaaS such as email
and messaging.

Once the companies understand the availability of application, they can easily examine the infrastructure, the required
services it uses and the updation of application. Therefore, the complex structure of cloud application shows the need of visibility
with in the services so as to estimate complete availability and updation.

Now, the role of the cloud application management is to alleviate these issues and generate a potential solution. This
solution should allow the user to have deep examination in the application for executing it in the cloud. Apart from this, it should
also design enterprise policies such as governance, auditing and environment management. This should be done at the time of
application deployment. Such cloud based monitoring as well as management services have certain benefits such as collecting
multiple of events, analysing them and determining the crucial information necessary to perform additional remedial actions.
These actions include adjusting capacity or provisioning new services. Apart from this, application management requires tools
necessary to manage the other environments.

L o o k f o r t h e SIA Group logo o n t h e title cover before you buy

Unit-3 Cloud Computing Architecture and Management 3.13

3.3 migrating application to cloud, phases of cloud migration approaches

for cloud migration

Q17. Write in short about migrating application to cloud. Also, discuss the phases of cloud migrating.

Answer : Model Paper-III, Q7

Migrating Application to Cloud

The cloud migration can be viewed as the process of transferring the multiple enterprise application and their corresponding
IT environments. These environments are originally placed on traditional hosting type and the transfer is done to the cloud
environment which could be public, private or hybrid.

With cloud migration, the user can get the opportunity to minimize the expenses incurred on applications. In doing so, the
user has to undergo different phases such as, evaluation, migration strategy, prototyping, provisioning and testing.

Phases of Cloud Migration

1. Evaluation

In this phase, all the components are evaluated. These components include current infrastructure, application architecture
and environment. They are evaluated with respect to certain factors like compute, storage, monitory and management, SLAs,
operational processes, risk, security, financial considerations compliance. Apart from this, licensing needs are determined
for developing business. Thus, it is required for moving the cloud.

2. Migration Strategy

In this phase, strategy is developed depending upon the reports generated after evaluation. The strategy employed is
hotplug strategy Here, the application and their corresponding data interface dependencies are kept separate. Also, these
applications can be utilized or operationalized all at one time.

Apart from this, another strategy called fusion strategy is used so as to migrate some of the applications. For rest of the
applications, dependencies exist based on existing licenses, specialized server requirements which could be mainframes,
highly sophisticated interconnections with other applications.

3. Prototyping

In this phase, the activities initiates only after the accomplishment of migration strategy. Here, validation is necessary so
as to ensure that small part of the applications are put to test on the cloud environment along with test data setup.

4. Provisioning

In this phase, premigration optimizations are determined and then implemented. To accomplish this, cloud servers are
provided corresponding to all environments which are determined. Apart from this, necessary platform softwares and
applications are put into work. This alerts the necessity of configuration which are tuned so as to match the newly generated
environment sizing.

At this point, files and databases also gets replicated. This step also ensures that internal and external integration points
are configured correctly. This leads to the support of web services, batch jobs operation and management software within
the new environments.

5. Testing

In this phase, tests such as postmigration are performed so as to confirm that migration is accurately done.

In addition to this, performance and load testing, failure and recovery testing, scale out testing are carried out. These testing
do not support certain elements, so it is done against traffic load and resource utilization levels.

SIA Group
3.14 Cloud Computing [JNTU-hyderabad]
Q18. Discuss various approaches for cloud migration.

Answer : Model Paper-I, Q7(b)

Approaches for Cloud Migration

The vendors adopt following four broad approaches for implementing cloud migration. They are,
1. Migrate existing applications.
2. Initiate from scratch
3. Separate company
4. Buy an existing cloud vendor.
1. Migrate Existing Applications
In this approach, some of the applications are redeveloped and restructured. Here, advantages of certain virtualization
technologies are atmost utilized to speed-up the work process. However, for new functionalities, the organizations have
to hire top engineers. This task is accomplished after releasing many applications with respect to the timings set by the
demand of the customer.
2. Initiate from Scratch
In this approach, the engineers rebuild or restructure the existing application instead of using other systems. As a consequence,
the R & D decisions changes in this regards. But if the team is small, then using complex development environments, the
approach can be successful.
3. Separate Company
In this approach, an individual new company is established which has a separate brand, management R & D and sales. Some
of the benefits which include use of internet protocol (IP) and investment from the existing company. Thus, minimizing
many prevailing conflicts with the establishment of new company in cloud.
This separated company acts as a subsidiary to the already existing company. And one of the important factors is that, the
new company can act, function and operate just like a cloud based start-up
4. Buy an Existing Cloud Vendor
In this approach, large vendor buying cloud based competitor so as to attain two things.
1. At first, it eliminates competitor.
2. Secondly, it makes the vendor capable to hit the ground operating in the cloud space.
Since, this is kind of merging the potential risk associated with it is innovation, drive subsequently the operational approach
of cloud-based company is likely to get compromise.

L o o k f o r t h e SIA Group logo o n t h e title cover before you buy