VIEW POINT

Strategy is Key to Migrate from Mainframe to Cloud

Abstract
As a majority of organizations on-board cloud solutions to improve IT operations, the
debate on the merit – and demerits – of cloud rages on. Despite its advantages of
elasticity, scalability and on-demand provisioning, most enterprises struggle with the
decision to migrate mainframes to cloud. This paper compares the similarities between
mainframes and cloud and highlights the unique advantages of cloud with the aim of
helping enterprises make informed decisions on their cloud adoption journey.
 Introduction
Nowadays, the terms ‘cloud’ and ‘mainframe’
represent a polarity on the infrastructure
deployment scale. While mainframes are
considered as legacy systems, cloud is
synonymous with virtualized and sophisticated
infrastructure.
A quick look into the history of mainframes
reveals that they were probably the first
attempt at a pay-per-use model. Consider
the IBM Mainframe charging model whereby
customers are only charged for the MIPS
capacity provisioned. Mainframes began as a
way to leverage integrated infrastructure with
compatible services that hosted enterprise
applications.
Cloud, digital and mobile solutions are
helping companies reduce cost and improve
efficiencies. In 2015, according to a KPMG
survey, global tech leaders predicted that
cloud computing will be the most disruptive
technology (11%) over the next three years
followed closely by mobile apps, data analytics
and the Internet-of-Things (IoT)1.
With cloud computing slated to be the key
driver of enterprise business transformation
(13%)1, it is not surprising that many
enterprises are looking for innovative ways
to migrate applications from mainframes to
the cloud. An IDG report states that 72% of
enterprises have at least one application in
the cloud and 56% of IT decision makers are
strategizing on how to continue leveraging
cloud2.
Enterprises today want to be agile – and cloud
is the obvious solution.
Technology innovations are focused for cloud
and cloud compatibility. IBM, with the zLinux
partition of Mainframes, is trying to bring some
of those innovations to the mainframes. The
focus to bring the services and innovations to
cloud, is not present on providing the support
on Mainframe.
However, for cloud migration to be successful,
enterprises must evaluate what their current
systems can – and cannot – do. For instance,
are existing systems unable to scale to meet
rising business demands? Will re-facing
legacy with a digital layer improve customer
satisfaction? Which cloud – private, public
or hybrid – will support their security
requirements? While creating an enterprise
cloud strategy, it is important to consider how
mainframes and cloud differ – and how they
are similar.

External Document © 2016 Infosys Limited

Stay elastic with powerful
and intelligent computing
Mainframes can quickly, securely and
reliably process large amounts of data.
For workloads that need even higher
computing power, batch scheduling runs
workloads during lean periods – such as
after office-hours between 12 am to 3 am
– or when an event is triggered – such as
on the last day of the month post IPL.
Cloud offers similar capabilities with an
advantage. On cloud, enterprises can
choose resources to reliably manage non-
functional requirements of workloads
such as performance and memory. In
cases where different workloads are
triggered simultaneously, cloud provides
features to preemptively auto-scale
provisioning instances to execute
those workloads for a preset duration.
On-demand elasticity gives enterprises
the benefit of being charged only for
the time during which computing
instances were used, truly leveraging the
characteristic pay-as-you-go feature of
cloud.
It is up to the enterprises to carefully
identify workloads and provision the
right computing capacity. With this
intelligence, enterprises can ensure that
cloud helps them meet non-functional
needs similar to mainframes and, more
importantly, reduce cost.
Shared risk. Shared security.
Mainframe applications are highly secure.
There are several mature, comprehensive
and market-ready security products
for mainframes that offer robust data
security with user identity and resource
level access management.
On cloud, however, security is a different
issue. In shared environments, security
too becomes a shared responsibility.
For most organizations, security
continues to be the foremost barrier to
cloud adoption3. To combat this, cloud
providers allow enterprises to set up
a virtual private cloud as an isolated
environment with restricted access to
applications migrated from mainframes.
Cloud infrastructure providers also have
native services that customize security
solutions for the unique needs of the
enterprise. Enterprises can migrate users
and user permissions from Mainframe
security profiles into a directory service
solution complying with lightweight
directory access protocol (LDAP). To stay
secure, enterprises must understand
how the security solution offered by their
cloud provider handles all their security
demands.
Robust design offers
scalable and highly-available
applications
Typically, mainframes use highly-available
applications to handle workloads
where downtime is unacceptable. On
mainframes, the zOS workload manager
(WLM) supports scalability and high
availability of resources for these critical
workloads. It also balances workloads
by provisioning data and resources
accordingly for various logical partitions
(LPARs).
Similarly, on cloud, enterprises can scale
their IT operations through highly-
available applications. The key difference
is how cloud deals with unexpected
surges. During spikes in demand,
applications can be deployed on
multiple server instances, ensuring zero
downtime or failures and fault tolerant
design. Similar to mainframes, a workload
balancer (WLB) service component is
designed to distribute, balance and
scale workloads in real-time. Due to truly
virtualized nature of the infrastructure,
one could scale to support higher
workloads without any significant time
delay. Batch workloads can be distributed
on different instances according to the
required data and resource needs and
available capacity.
On cloud, there is an additional
advantage of flexibility, for distributed
deployment across regions / data center’s
with minimal effort and thus can help
scale the application portfolio as the user
base scales globally
External Document © 2016 Infosys Limited
External Document © 2016 Infosys Limited
Connect seamlessly to your mainframe and on
premise datacenter
Most cloud providers can provision connectivity between cloud and
on-premises data centers through various methods including virtual
private network (VPN), direct or network connection over multi-
protocol label switching (MPLS). This connectivity is available as a
service on cloud.

Further, cloud-based data and applications can be integrated with the

mainframe using integration patterns such as file transfer protocol
(FTP) for batch workloads, MQ WebSphere for asynchronous event-
driven workloads or web services over the internet or intranets for
synchronous and real-time workloads.

External Document © 2016 Infosys Limited

 Evaluate the cloud migration business case
Gartner already predicts that the global public cloud services market is set to reach US $204 billion in 2016 with the highest
growth stemming from cloud infrastructure services (38.4%)4. A risk-proofed and well-assessed strategy helps companies make the
most from these investments. When creating an enterprise cloud migration strategy from mainframe to on-premises, enterprises
evaluate servers, target technologies, technology compatibility, tool stacks, etc. These considerations are equally important when
assessing the value of cloud migration.
Definitely, cloud offers significant advantages over mainframes. Besides cost savings, organizations access synergistic platforms,
native cloud capabilities and ‘as-a-service’ delivery models. Even if enterprises want to bring their own components and services to
cloud, they can do so provided these are compatible with the cloud instances.
To maximize existing IT investments, it is recommended that enterprises strategically evaluate the business case for migrating
workloads from mainframes to cloud. With the right migration strategy, organizations will achieve tangible cost benefits and
advanced infrastructure capabilities to enable next generation business needs and sustain current and future business growth.

External Document © 2016 Infosys Limited

Conclusion
As the first iteration of the pay-as-
you-go model, mainframes share
several similarities with cloud.
Like mainframes, cloud offers high
computing power, elastic workload
provisioning, customizable network
security solutions, and connectivity
to on premise datacenter. However,
the advantage of cloud is agility
to development and deployment,
leveraging technology disruptors
and innovation faster, the ability to
truly offer pay-per-use functionalities
through intelligent workload
management and the design of
applications that scale in real-time,
enabling next generation business
capabilities and user experience, with
significant improved time to market.

Enterprises with existing legacy

systems looking to leverage the
elasticity, scalability and high-
availability of applications on cloud
must carefully assess their priorities
for doing so if they are in need of
improved time to market, agility
in development and deployment,
minimized cost and leverage next-
generation capabilities.

External Document © 2016 Infosys Limited

References

1. 5 insights and predictions on disruptive tech from KPMG’s 2015 Innovation Survey, Nov 2014
https://softwarestrategiesblog.com/tag/idc-saas-forecasts/

2. IDG Cloud Computing Survey 2015

http://www.idgenterprise.com/resource/research/2015-cloud-computing-study/

3. Cloud Security Concerns Rise as Investment in Cloud Grows, Businesswire, May 2016
http://www.businesswire.com/news/home/20160517005583/en/Report-Reveals-Cloud-Security-Concerns-Rise-Investment

4. Gartner Says Worldwide Public Cloud Services Market Is Forecast to Reach $204 Billion in 2016, January 2016
http://www.gartner.com/newsroom/id/3188817

About the Author

Dhaval Jagani
Senior Technology Architect, Infosys

Dhaval has over 13 years of experience as an Application Architect and in Mainframe Modernization
space. He anchors the technology modernization for Health care and Insurance clients at Infosys.
He has worked closely with various enterprise business and architecture stakeholders to solve
complex business problems and led many enterprise initiatives from an architecture stand point.
Dhaval has extensive techno-functional knowledge of Healthcare and Insurance ecosystem and has
extensive experience in the Mainframe Integration and Modernization solutions.

He can be reached at Dhaval_Jagani@infosys.com.

For more information, contact askus@infosys.com

© 2016 Infosys Limited, Bengaluru, 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.

Stay Connected