Introduction to Data Science

and Machine Learning

• 27, 28, 29 December 2017

Unit 6

Future directions in Cloud

Computing
• Data Science Track,
ETA, Infosys Ltd.

PUBLIC Copyright @ 2018, Infosys Limited 1

Copyright Guideline
© 2017-2018 Infosys Limited, Bangalore, India. All Rights Reserved.

Infosys believes the information in this document is accurate as of its publication date;
such information is subject to change without notice. Infosys acknowledges the
proprietary rights of other companies to the trademarks, product names and such other
intellectual property rights mentioned in this document. Except as expressly permitted,
neither this documentation nor any part of it may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical, printing,
photocopying, recording or otherwise, without the prior permission of Infosys Limited
and/ or any named intellectual property rights holders under this document.

PUBLIC Copyright @ 2018, Infosys Limited 2

 • Cloud Migration
• When to and not to migrate on Cloud?
Future directions in Cloud
Computing • Cloud Migration Process
• Cloud Computing – Risks
• Cloud Computing – Future Trends

PUBLIC Copyright @ 2018, Infosys Limited 3

 Cloud Migration

PUBLIC Copyright @ 2018, Infosys Limited

Is Cloud Migration always beneficial?

• Over the past decade, cloud services have rapidly become one of the most defining
and beneficial technologies in IT.
• In some cases, cloud services prove to be extremely beneficial. In others, however,
cloud migration probably doesn't make sense.
• Hence we need to consider whether or not to move to Cloud.

PUBLIC Copyright @ 2018, Infosys Limited 5

Factors to consider before Migration

• Existing Infrastructure:
¾ Consider existing investments in infrastructure
¾ hardware ,network architecture etc.

• Application Performance :
¾ Whether existing application can perform adequately in the cloud.

• Cost:
¾ Consider Budgetary aspects
¾ Whether cloud migration is cost-effective considering current investments in infrastructure.

PUBLIC Copyright @ 2018, Infosys Limited 6

When to and not to migrate on Cloud?

PUBLIC Copyright @ 2018, Infosys Limited

When to migrate to cloud?

• Existing hardware is becoming obsolete

• Starting up a new organization
• Application is receiving high load and it is challenging to scale up fast
• Scale up resources instantly in case of increased demand
• To minimize the IT management costs and focus on business aspects
• To facilitate faster deliveries when there is need to focus more on development
aspects rather than IT infrastructure.
• The data storage need is increasing and is difficult to do in-house
• If there is a need to create a distributed team, cloud is a solution as it allows people
to connect to the resources from anywhere.
• Backups need to be created as disaster recovery solutions. With cloud, setting up of
a backup datacenter is possible with minimal effort and time.
PUBLIC Copyright @ 2018, Infosys Limited 8
Challenges

• The idea of migrating IT assets to the cloud can be intimidating because of the change
management required
• Fundamental process around migration calls for understanding the benefits of the new
system, assessing gaps in the existing system
• Getting to the right design and deciding what goes where and what is to be retained is
challenging
• Obtaining business approvals would need a strong business case
• Workload Migration can be time consuming
• Deciding on the right tool based on the organization’s need would need multiple
reviews and considerations which is a tedious process

PUBLIC Copyright @ 2018, Infosys Limited 9

When not to migrate to Cloud

• In case of critical and sensitive application data, migrating to cloud is not

recommended.
• If you have already invested in on-premise datacenter setup ,management and
scalability is being handled properly , customers are satisfied, better not to migrate
to cloud.
• If we are using licensed technologies in current set up, migrating on to cloud might
involve legal issues.
• Applications might experience response time delays after migrating to cloud.

PUBLIC Copyright @ 2018, Infosys Limited 10

When not to migrate to Cloud

• You may have limited control over the infrastructure/applications after cloud
migration, hence may lead to reduced transparency.
• You may need to modify the existing designs and architectures to fit it in the shoes
of CSP.

PUBLIC Copyright @ 2018, Infosys Limited 11

Cloud Migration Process

PUBLIC Copyright @ 2018, Infosys Limited

Which Cloud to opt for?

Once its decided to migrate to Cloud, we need to choose a cloud service model out of the
below:
•IaaS – Infrastructure as a Service (e.g. Amazon Web Services, Google Cloud Platform,
Windows Azure).
•PaaS – Platform as a Service (e.g. Google Cloud Platform ,AWS Elastic Beanstalk).
•SaaS – Software as a Service (e.g. Salesforce, Google Apps, Office365).

PUBLIC Copyright @ 2018, Infosys Limited 13

Which cloud deployment?

After deciding on cloud service model, we need to choose a cloud type out of the below:
• Public (if you host your IT infrastructure and services through a CSP like Amazon Web
Services.)
• Private (if you need to create an dedicated cloud either on-premise and of-premise for
secured and confidential data handling using a cloud provider like OpenStack)
• Hybrid (when you need a blend of both private and public cloud platform features)

PUBLIC Copyright @ 2018, Infosys Limited 14

Deployment Strategy

Decide where the services should run

• Infrastructure • Operated solely for

available to public an organization
• Owned by Cloud • Can exist either on-
Service Provider premise or off
premise
Public Private

• Two or more clouds Hybrid Community • Infrastructure shared

by several
that interoperate or
organizations
federated through
supporting specific
technology
community that have
common concerns

PUBLIC Copyright @ 2018, Infosys Limited 15

Access your applications before migration

After choosing the apt CSP, we need to make our application ready for migration. The
steps to be followed for the same are:

• Complexity in design
9 Traditional application design might be complex and strongly coupled.
9 Make your application capable of handling scalability and distributed architecture.

• Complexity in Integration
9 Applications dependencies (payment gateways , storage gateways etc.) should be
handled carefully . The order of migrating all dependencies should be maintained.
9 Check if dependencies are getting affected due to cloud migration. Find out ways
to remediate those effects.

PUBLIC Copyright @ 2018, Infosys Limited 16

Access your applications before migration

• Host operating system

¾ Check for the OS compatibility issues for your application. Some applications may
not be supported by all OS.

• Application database
¾ Sensitivity of the data should be checked. In case data is highly critical and
sensitive, cloud migration is not recommended. Alternatively you may opt for
hybrid models.

• Network
¾ Issues like multicasting may occur if we change our network to virtual networks.
Check an alternative to that as well.

PUBLIC Copyright @ 2018, Infosys Limited 17

Cloud Migration Approaches

• While there are several common components found in successful cloud migration, there is
no one-size fits all solution to deciding on the best approach.
• Every organization has its own unique goals, constraints, budgets, politics, culture, and
market pressures that will guide its decision-making process along the way.
• There are couple of approaches adopted towards Cloud Migration:
– Five-Phase Migration Process
1. Migration Preparation and Business Planning
2. Portfolio Discovery and Planning
3. Design and Migrate Applications
4. Validate the Applications Migrated to Cloud.
5. Modern operating Model :Fine tune Apps in new Cloud Environment and turn off
the on premise Model

PUBLIC Copyright @ 2018, Infosys Limited 18

Cloud Migration Approaches

– Six Common Strategies for Migrating to the Cloud (“The 6 R’s”)

• Rehosting
• Replatforming
• Repurchasing
• Refactoring / Re-architecting
• Retire
• Retain

PUBLIC Copyright @ 2018, Infosys Limited 19

Cloud Computing – Future Trends

PUBLIC Copyright @ 2018, Infosys Limited

Future technology trends in Cloud Computing

• Innovation: Cloud is not longer meant only for upgrading and scaling the IT
infrastructure but would give an adverse platform to rapid innovation in variety of latest
technologies like Big Data, Mobile , Analytics , Artificial Intelligence.
• Hybrid Computing: Hybrid computing will continue to be the best cloud choice as it
gives the customers the leverage on what to expose to cloud and which resources to be
managed in house.
• Getting into the comfort zone of cloud : Since cloud is still new in IT world, IT is not
very comfortable with terminologies of cloud. Trusting a CSP when you don’t know
cloud well is difficult. Slowly these trends will change and will lead to people adopting
cloud extensively.

PUBLIC Copyright @ 2018, Infosys Limited 21

Future technology trends in Cloud Computing

• Pricing will not be the only reason: To adopt cloud, what we foresee is pricing as the
major reason. Where-ever we read the benefits of cloud, first thing discussed is pricing.
All major CSPs are fighting the price battles. Soon people will realize the benefits of
cloud and will choose cloud for other reasons as well.
• Cloud- a buzz for start ups: Cloud removes the hindrances which come in the way to
open small home businesses. People will adopt cloud as they will get to know that with
minimal investment also they can launch their startups .

PUBLIC Copyright @ 2018, Infosys Limited 22

Learning Outcomes
After going through this unit, you are able to :
• Identify the scenarios when to migrate or not to migrate to Cloud
• Comprehend the steps involved in Cloud migration
• Understand the latest trends in Cloud Computing

PUBLIC Copyright @ 2018, Infosys Limited 23

Thank You

PUBLIC Copyright @ 2018, Infosys Limited 24