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

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 [email protected].

For more information, contact [email protected]

© 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