IBM Mainframe to AWS

Reference Architecture

Astadia Mainframe-to-Cloud Modernization Series

 IBM Mainframe to AWS Reference Architecture

Abstract
In businesses today, across all market segments, cloud computing has become the focus of current and
future technology needs for the enterprise. The cloud offers compelling economics, the latest
technologies and platforms, and the agility to adapt your information systems quickly and efficiently.
However, many large organizations are burdened by much older, previous generation platforms,
typically in the form of a mainframe computing environment.

Although old and very expensive to maintain, the mainframe platform continues to run the most
important information systems of an organization. The purpose of this reference architecture is to assist
business and IT professionals as they prepare plans and project teams to start the process of moving
mainframe-based application portfolios to Amazon Web Services (AWS). We will also share various
techniques and methodologies that may be used in forming a complete and effective Legacy
Modernization plan.

In this document, we will explore:

 Why modernize a mainframe
 The challenges associated with mainframe modernization
 An overview of the IBM mainframe
 The IBM mainframe to AWS Reference Architecture
 An overview of AWS services
 A look at the Astadia Success Methodology

This document is part of the Astadia Mainframe to Cloud Modernization Series that leverages Astadia’s
25+ years of mainframe platform modernization expertise.

© 2017 Astadia. Inc. - All rights reserved.

12724 Gran Bay Parkway, Suite 300
Jacksonville, FL 32258

All other copyrights and trademarks the property of their respective owners.
 IBM Mainframe to AWS Reference Architecture

Contents
Introduction ............................................................................................................................ 1
Why Should We Migrate Our Mainframe Apps to AWS? ................................................................. 2
Benefits of Mainframe Modernization ..................................................................................................... 2
Approaches to Mainframe Modernization ............................................................................................... 2
Challenges of Mainframe Modernization ................................................................................................. 3
Why AWS?................................................................................................................................................. 4
Achieving the Positive Impact of Change.................................................................................................. 5
Understanding Typical IBM Mainframe Architecture ...................................................................... 6
IBM Mainframe Heritage .......................................................................................................................... 6
IBM Mainframe Components ................................................................................................................... 6
IBM Mainframe to AWS Reference Architecture ............................................................................ 8
Understanding Amazon Web Services (AWS) ..............................................................................12
Ensuring Project Success .........................................................................................................15
Astadia Mainframe-to-AWS Success Methodology ................................................................................ 15
Conclusion .............................................................................................................................17
IBM Mainframe to AWS Reference Architecture

This page intentionally left blank.

 IBM Mainframe to AWS Reference Architecture

Introduction
The Amazon Web Services (AWS) cloud computing platform is an excellent target environment for
transitioning from an IBM mainframe workload to a cloud implementation. With the security features of
AWS and the ability to scale based on demand for the services, AWS offers a complete operational
environment in support of mainframe workloads that have been migrated to the cloud. In addition, AWS
supports new innovation of the application portfolio, previously held captive by the inflexible nature of
a mainframe computing model, improving the productivity of application developers and support
personnel.

Even more than a typical IT project, planning to modernize mainframe applications is the most important
phase of the total project effort. A good place to begin is with a thorough assessment of the existing
overall mainframe application portfolio. Through the assessment process, all aspects of the existing
portfolio will be inventoried and examined in detail, resulting in a catalog of each application, database,
technology platform and business user profile currently in use. Once completed, the results of this
application rationalization will then guide the sequence of application migration, as well as the different
modernization strategies and techniques that may be called upon over the course of the entire project.
We’ve included an overview of how Astadia tackles Legacy Modernization projects with our Success
Methodology to give you an idea of what’s involved.

Don’t let the enormity and importance of a mainframe modernization project deter you from getting
started. The skilled individuals needed to continue to maintain mainframes are increasingly leaving the
workforce through retirement and are not being replaced. Hardware and software maintenance costs
continue to escalate and the demands of customers, employees and partners require greater innovation
than mainframe platforms can support.

How to Use This Reference Architecture

Begin by reading the “Why Should I Migrate” section first. From there:
Mainframe Experts: Skip to the IBM Mainframe to AWS Reference Architecture and the
Understanding AWS sections
AWS Experts: Start with the Understanding IBM Mainframes section, followed by the Reference
Architecture.
Business Leaders: Spend time with the “Why Should I Migrate…” section and the Ensuring
Project Success section at the end.

About Astadia
Astadia has been in the legacy modernization business since 1994 and has successfully completed more
than 200 mainframe modernization projects. Our repeated success has allowed us to develop a
comprehensive methodology, proprietary software tools and techniques, as well as the “know how” that
comes with more than 25 years of experience handling mission critical applications and data. We’re
pleased to share some of that experience with you through our Mainframe to Cloud Modernization Series
of reference architectures, webinars, whitepapers and more. Visit our website at www.astadia.com for
additional information.

© 2017 Astadia, Inc. 1

 IBM Mainframe to AWS Reference Architecture

Why Should We Migrate Our Mainframe Apps to AWS?

Over the past 10 years, public cloud computing has emerged as the foundation of future enterprise
technology. In terms of technology generations, mainframes are at least two generations old, perhaps
three. Yet, they still survive today and are responsible for running key financial, healthcare and other
vital and sensitive systems around the world.

So, why should you migrate your mainframe workloads, why migrate them to AWS and why is now the
right time?

Benefits of Mainframe Modernization which ones will be determined during the critical
application portfolio rationalization step of the
The specific benefits in moving any mainframe
project’s assessment phase. Here are three of the
workload will vary between organizations and
most common approaches:
even at the application and database level. In
general, here are three of the top reasons driving
Reuse – Often called “lift and shift”, this is a
legacy modernization projects today:
process that reuses the existing
code/program/applications, typically written in
Cost – The economics of cloud computing are
COBOL, by moving them off the mainframe, and
compelling when compared with the status quo of
recompiling the code to run in a mainframe
maintaining a mainframe environment. A total
emulator hosted in a cloud instance. This approach
cost of ownership (TCO) evaluation of the
minimizes the upfront risks and the length of the
subscription-based, consumption driven cost
project, realizing hardware and software cost
model of the cloud versus the exorbitant hardware
savings soonest.
& software maintenance costs of mainframes will
show a very appealing and short-term achievable
Running mainframe applications in an AWS-hosted
ROI (potentially less than 12 months from project
emulator also opens the possibility of new
completion).
innovation leveraging APIs to previously
inaccessible programs and data.
People – Mainframe-specific technical skills are
not being replaced by today’s college or
Rewrite – It may be tempting to say, “Let’s just
technology trade school graduates. The pool of
write new programs from scratch,” to modernize
available talent with relevant knowledge and
the mainframe applications. This approach is
experience is shrinking exponentially each year.
extremely risky and fails a vast majority of the
The cloud leverages modern technology and its
time. It is complex, costly, and time consuming.
use is ingrained into young software engineers
The resources and investment required tends to
worldwide.
greatly exceed the forecast.
Flexibility – The cloud offers an Open Systems
A new, modern codebase may still be the correct
environment in which high productivity and rapid
end objective, but a better approach would be to
innovation happen at a tremendous rate. A
first move the applications to a cloud-based
properly designed implementation of a cloud
emulator, migrate the database to a cloud-based
infrastructure scales easily and quickly, both
database, then focus on replacing modules/code
expanding and collapsing to synchronize with
over a deliberate, multi-phased approach. When
business demand. Backup, redundancy and
it is time to rewrite, there are several code
disaster recovery is seamless. Support for multiple
transformation engines you can choose from to
end-user platforms and devices is inherent.
reduce the effort and minimize the risk.
Database sharing across the enterprise with high
performance is achievable.
Replace – Another mainframe modernization
approach is to completely replace the mainframe
Approaches to Mainframe Modernization functionality with a program or suite of programs,
You may notice throughout this document that we typically a Software-as-a-Service (SaaS)
use the terms “Mainframe Modernization” and application. You typically see this with purpose-
“Mainframe Migration”. Migration is a type of built solutions for finance, human resources,
modernization, whereas modernization manufacturing, enterprise resource planning, etc.
encompasses a broader set of strategies or There are also industry specific apps that may
options. In many cases, you will employ a solve the problem that a custom mainframe
combination of these strategies, the right mix of solution was needed for decades ago.

2 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

The upside of using SaaS is that your organization Application-specific Challenges

no longer worries about maintaining code. There are a couple of general points about the
However, you will find that while you can configure application portfolio that should be noted. As
a SaaS application with various options provided mentioned above, the lack of documentation on
by the vendor, you will not be able to customize these aging systems makes the migration effort
your instance, as the shared codebase runs all more difficult. The project team that drives a
tenants (customers/organizations) using the migration project must then resort to “mining” the
“service”. actual application source code to determine
exactly the behavior of the application.
There are additional variations on these three
modernization approaches and you’ll likely use Another important application-specific issue for
several strategies in achieving your goal to consideration is discovering the integration
completely migrate from the mainframe. It is requirements and dependencies of the application
commonly accepted best practice among legacy with other systems and databases. These
modernization practitioners to primarily use the integrations and dependencies must be clearly
lower-risk, lower-cost Reuse approach first to identified and, if still needed, they must be re-
capture the gains and benefits in the shortest time connected (possibly rebuilt) and made operational
possible, followed by a deliberate and phased along with the migrated system.
approach to Rewrite or Replace the applications.
Running Parallel Systems
Challenges of Mainframe Modernization For a short while, there may need to be some
Mainframe migration projects are complex and parallel processing between the mainframe
require close management of the process, budgets application, while it is still being used in
and timelines that have been set as project goals. production, and the newly migrated system, on
A Reuse approach will involve rehosting (from the new platform. Planning and executing this
mainframe to AWS) and likely some re- parallel processing will be a challenge, and will
engineering and refactoring to complete an entire require extra time and attention to make it
mainframe migration. It will also involve data and successful.
file conversions for transitioning the database to
the cloud. Another example of when you may choose to run
parallel systems is if you want to achieve quick
As we’ve been emphasizing, the first challenge of reductions in mainframe processing consumed by
any mainframe modernization project is to moving the development and test environments to
develop a rock-solid plan built upon a thorough an AWS-based emulator while keeping the
application portfolio assessment and production system on the mainframe for the
rationalization. As you put your plan together and interim.
begin to execute, here are additional factors you’ll
need to watch out for:
Data Integrity
Moving the contents of large databases is very
Documentation challenging on a number of levels. Typically, a
Many mainframe environments with large and database “cleanup” will be necessary to ensure
complex application portfolios do not have that the contents of the new target database is as
documentation that details what these mainframe accurate, and as complete, as possible.
applications do, and how they do it. Many
applications are decades old, so the original A mainframe modernization project is a good time
system, with changes likely every year, has to transform, correct and validate the
become a maintenance nightmare. The external organization’s data.
interaction with these systems, the Input/Output,
is how these systems get defined to the business,
Speed to Completion
and the rest of the system is just a black box.
In almost every project, speed will be a top
priority. The costs and complexities of extended
Migrating a minimally-documented system of this
project cycles can have an enormous negative
nature is tricky and the testing prior to the “go
impact in tangible and intangible terms. As project
live” deployment is critical to mitigating this issue.
cycles get extended, staff attrition can become a
(And, of course, copious documentation should be
big issue and staff fatigue may also become a
captured for the resulting system.)
factor.

© 2017 Astadia, Inc. 3

 IBM Mainframe to AWS Reference Architecture

Paying for a continuation of the primary production range of use cases. They also have the largest
system and funding the development efforts of the compute capacity in use by paying customers. This
new system at the same time will have a attracts open source and application developers as
temporary financial impact for as long as that well as service providers to make their applications
duality continues. Getting to a “go live” status compatible or add their services to AWS.
quickly and efficiently with the new system, and
retiring the old system, will keep unexpected costs The benefits of migrating your mainframe to
to a minimum. AWS are:

Project Funding Easy to Use – AWS is designed with simplicity in

It is very important for any modernization project mind. You can request new services and host your
to be properly funded and supported by the applications using their simple to use web-based
business management team and the executives. AWS Management Console. All their services are
This support is essential to maintain project well-documented and there is a wealth of forums,
continuity and funding throughout the project white papers, and discussion boards.
cycle. Since we stated earlier that speed will be a
factor in the project execution, funding must be in Flexible - You can select from a wide variety of
place to sustain that speed. virtual environments where you choose the
software and services your application requires. If
you find that the environment selections are not
Expertise adequate, you can simply provision different types
Mainframe migration projects come in many of instances or add compute and/or storage on
forms. In every case, a variety of specialist skills demand.
will be needed on the project team. These
specialists may include business analysts working Cost-Effective - Amazon Web Services are billed
to mine and understand the business rules in a consumption model, where you only pay for
embedded in the legacy applications. the compute and storage resources you use with
no upfront commitments and contracts.
It will also include experts in specific programming Alternatively, if you know the minimum compute
languages, databases, networks, terminal devices capacity required during an extended period of
and many other components of the total time, you can reserve capacity at a significant
application portfolio that will need to be addressed discount.
over the course of the migration to the target
platform. Staff must also be available to address Reliable - With AWS, you are taking advantage of
any specific functionality or use case of the its highly redundant, worldwide computing
mainframe application environment. infrastructure that is designed with strong high
availability capabilities, which satisfy some of the
All this technology must be transferred to the most stringent enterprise requirements.
equivalent functionality on the target cloud
platform and work as it did in the original Scalable - AWS includes features such as Auto
mainframe environment. Thorough testing by the Scaling and Elastic Load Balancing which allow
project team, followed by testing amongst the your application to scale in or out, if you
business users of the original mainframe design/architect this into your solution. AWS’s
application system, is an absolute requirement. massive compute and storage infrastructure is
Once testing is completed, a final performance and designed to make resources available when they
tuning (P&T) exercise will ensure that the new are needed.
cloud deployment is performing at optimal levels.
High Performance - AWS offers a wide selection
Why AWS? of compute and storage options to satisfy the
Organizations keep discovering new and improved performance needs of your formerly mainframe-
benefits for moving their mainframe (as well as based applications. Compute and storage can be
other) workloads to the public cloud such as provisioned as they are needed, so if your
flexibility, scalability, automatic backups, application is CPU intensive you can have a larger
automatic software upgrades, cost model CPU/IO ratio and vice versa.
optimization, versioning control and adding
multiple security layers, just to name a few. Secure - AWS provides several security
capabilities and services to improve privacy and
AWS is a recognized leader in the public cloud network response. These include network
space with a diverse customer base and a broad firewalls, AWS Virtual Private Cloud (VPC),

4 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

encryption in transit, and private or dedicated the road, which are likely not available to
network connections. mainframe users today.

Achieving the Positive Impact of Change The overall impact of a successful mainframe
migration project is a positive one for the entire
In any mainframe migration project, the results of
organization. A new and better application
a cloud-based application set may be daunting.
portfolio, a cloud platform to enable innovation,
The change will impact the technical staff, as they
and a large cost savings in the operational and
will likely need to learn new skills.
systems software maintenance categories will be
realized. It’s not unusual to repurpose IT staff
The end-user community may not notice too many
after redeploying the mainframe portfolio to the
changes using a new system if the interfaces are
cloud. The cloud platform has many other benefits,
preserved. In fact, the move to the cloud could
but flexibility and cost takeout are at the top of the
fuel innovation resulting in new capabilities down
list.

© 2017 Astadia, Inc. 5

 IBM Mainframe to AWS Reference Architecture

Understanding Typical IBM Mainframe Architecture

Since their development in the late 1940s, general-purpose mainframes were the computing workhorses
for more than 50 years. Over that time, each mainframe manufacturer continuously enhanced their
unique architectures to outperform competitors and meet evolving business demands. IBM and Unisys
eventually dominated the market and became the gold standards of mainframe computing. This IBM
Mainframe to AWS Reference Architecture is part of the Astadia Mainframe-to-Cloud Modernization
Series of architectures, whitepapers and webinars.

IBM Mainframe Heritage

IBM began producing mainframes in 1952. These operators loaded jobs manually, programs could
early models, known as the IBM 700 Series, were be executed using a queuing mechanism to
based on vacuum tube technology. The next improve efficiency and ROI. IBM-provided
generation, IBM 7000 Series, saw the introduction software quickly grew in its complexity and
of transistor technology and became the mainstay became an important piece of its mainframe
of their mainframe business for many years, with computing solutions.
some models remaining in service up to the 1980s.
The System/360 also consolidated support for
In 1964, IBM announced availability of the processing features, like decimal arithmetic,
System/360, with the “360” representing the idea floating-point arithmetic, and byte addressing.
of 360 degrees – as in, an all-around computing These features were previously available only in
system. Previous IBM mainframes were delivered models built for specific purposes like business or
with no software since it was expected that the scientific calculations.
customer would write all programs. These
programs were loaded and execute manually, one As technologies and software advanced, newer
at a time. models were released as System/370 and
System/390, culminating in the 64-bit eServer
With the release of the System/360, IBM delivered zSeries, z Systems, and IBM’s current line of
software such as compilers for programming zEnterprise mainframes. The “z” refers to “zero
languages and early operating systems. Instead downtime” since the models are built with
of these very expensive machines sitting idle while components that provide hot failover capabilities.

IBM Mainframe Components

User Interfaces interface that allows a more user-friendly interface
Users access the mainframe application through a to mainframe applications. One of such user
variety of means. They could use green screen interfaces could be a web-based or mobile
terminal emulators that provide character mode application serving as a front end to the
interface (TN3270). mainframe.

Alternatively, a variety of custom user interfaces

could be built on top of the character mode

6 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

Batch and they provide high levels of availability,

Mainframes provide batch environments that integrity, consistency, reliability, security, and
handle bulk data processing workloads. Jobs are auditing.
submitted to the system using JCL and processed
with minimal operator interaction. Output from the Database software makes intensive use of the
batch jobs is spooled, printed and distributed to computing and input/output capabilities of the
users. mainframe to provide optimal response times. IBM
offers specialized processors for database
workloads to reduce the burden on general
Transaction Processing processors.
Transaction processing is at the core of most
mission-critical applications with thousands or
millions of transactions being processed daily. IBM
Environmental software
mainframes provide online (real-time, on- Mainframes require software to support the
demand) processing environments (most management, operation, application development,
commonly, CICS and IMS) that make this possible. and security of the system.
Security, transaction integrity, and predictable
response times are of particular importance for Software tools like Time Sharing Option (TSO) and
this type of workload. Interactive System Productivity Facility (ISPF) are
used by administrators and programmers to
create, store, print and delete datasets as well as
Programming Languages submit batch jobs.
Mainframes provide an assortment of
programming languages to suit customer needs. Job scheduling software is used to automate and
Most applications are written in COBOL but other manage batch job execution and workflow
languages are also used: Assembler, PL/I, Natural, streams. Output management systems handle the
Fortran, REXX, etc. 4GL development products collection, storage and distribution of reports to
like Mantis, Informix 4GL, and APS (AppMaster) users. Source management systems are used to
are also used to develop mainframe applications. maintain application source code by tracking
version as well as release lifecycles.
Data Files
Mainframes store data in files with different record Terminals and terminal emulation software allow
organizations and media types. Data files can be users to interact with mainframe software and
sequential, direct access, fixed and variable applications. The Telnet 3250 (TN3270)
lengths, blocked or unblocked, etc. Data files can communication protocol is used to communicate
be stored on disks, magnetic tapes, CDROMs, etc. between the mainframe and a terminal session.
Some examples include VSAM, GSAM, and ISAM.
For the most part, data in these files are stored in Security is tightly controlled at all levels of the
EBCDIC (Extended Binary Coded Decimal mainframe software. Software provided by IBM,
Interchange Code), an eight-bit character such as Resource Access Control Facility (RACF)
encoding system used primarily on mainframes. and Access Control Facility 2 (ACF2) by Computer
Associates, cooperate with system components to
Databases provide a robust yet secure environment for
Mainframes provide high performance database applications and data. Security software is
management systems to support online mission designed to minimize the risk of data exposure and
critical applications. In general, these databases provide regulatory compliance.
can be hierarchical (IMS DB) or relational (DB2),

© 2017 Astadia, Inc. 7

 IBM Mainframe to AWS Reference Architecture

IBM Mainframe to AWS Reference Architecture

As a Reference Architecture, the following diagrams and discussion address a typical use case. However,
each implementation is sure to have its own customizations and variations, which is why a thorough
application portfolio inventory, assessment and rationalization is critical to a successful outcome.

Below, you will see a design that includes details such as AWS components, batch requirements,
programming language conversions and replacements, integration with external systems, 3rd-party
software requirements, and planning for future needs.

In an actual project, you would also consider any unique features that would necessitate custom-made
solutions. We would recommend proof-of-concept conversions on application subsets to test the model
selected, discover any weaknesses, and prove the viability of the design.

Migrating IBM mainframe applications to AWS:

Code Modification
As part of any mainframe modernization project,
there will be a need for partial or, in some cases, Astadia employs an iterative, hybrid process of
extensive code modification. Leveraging our past automated code conversion and human
experience in migrating IBM mainframes to on- intervention. The technology behind the
premises open-systems platforms, Astadia has automation is our Rules-Based Transformation
developed an extensive set of code transformation Engine.
technologies and processes to successfully re-
platform mainframe workloads. We use these in This tool preserves the business logic and rules of
concert with trusted third-party products to modify legacy applications while removing proprietary
source code for deployment to AWS. What follows code that can only execute in the source
is a description of our approach. environment and not in AWS. Its code migration

8 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

filters ensure the preservation of mission-critical intervention ensures that legacy applications will
applications and back-end components such as work in AWS without sacrificing their original
trancodes, security policies, and message routing. functionality.

Though the Rules-Based Transformation Engine Although every mainframe migration is unique,
(RBTE) is a proven technology, Astadia augments there are general source-to-target mappings for
our technology with years of hands-on migration application components that apply to most
experience and collaboration with partners. The projects, as shown in the following table.
combination of automation and human

Code Modification Mapping Table

Source Target

z/OS, MVS, VSE AWS (Windows, Linux, or UNIX)

CICS Micro Focus, Oracle, GT Software (Fujitsu)

IMS Micro Focus, Oracle

Assembler COBOL, C#, Java, or mapped to OS functions

JCL JCL, PowerShell, other scripting

COBOL COBOL, C#, Java

Natural Natural, COBOL, C#, Java

FORTRAN, PL/I FORTRAN, PL/I, COBOL, C#, Java

REXX, PL/I REXX, PowerShell, other scripting

These mappings are only a guideline for the most common mainframe technologies. Other technologies are addressed on
an as-needed basis.

Database Migration
In parallel with code modification, data specialists address unique technical requirements or
will need to perform a thorough analysis of the preferences.
legacy databases and files, and develop a detailed
data migration strategy. After the target database and file structures have
been created and validated, static data can be
We recommend an iterative extract, transform and migrated to the AWS production environment. For
load (ETL) process to identify potential data-typing dynamic or other data that is created and/or
issues, develop fixes, and collaborate with the modified frequently, a data migration strategy
application subject matter experts to validate their must be implemented as part of production
efficacy. This iterative process continues until cutover process.
every issue is eliminated.
Like the application component mapping above,
In most cases, hierarchical and flat file data there are general source-to-target data mappings
structures will be replaced with RDBMS solutions, employed by most mainframe-to-AWS migrations:
but other solutions may be implemented to

© 2017 Astadia, Inc. 9

 IBM Mainframe to AWS Reference Architecture

Database Migration Mapping Table

Source Target

DB2 Aurora, SQL Server, DB2 LUW, Oracle, etc.

IMS DB Aurora, SQL Server, DB2 LUW, Oracle, etc.

VSAM, other flat files ISAM, flat files, Aurora, SQL Server, DB2 LUW, Oracle, etc.

GDGs GDGs

These mappings are only a guideline for the most common mainframe database and file technologies. Other technologies
are addressed on an as-needed basis.

10 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

AWS/IBM Mainframe Integration and Parallel Operations

For some organizations, a one-time deployment applications to AWS as a means of reducing their
and cutover of all mainframe applications to AWS costs or delaying upgrades to existing mainframe
simply isn’t feasible. hardware.

There may also be organizations who prefer to There’s good news here - it is possible to do a
move applications one at a time or in smaller phased rollout of migrated applications and still
groups as a way of gradually embracing AWS as a have ongoing communication and integration with
solution for their big-iron applications. applications residing on the IBM mainframe. This
kind of mixed environment can be achieved with
Still others may intend to keep their mainframes the proper planning, and will generally resemble
indefinitely for a subset of strategic, mission the image below.
critical applications while migrating less critical

IBM mainframe integration with applications & data already migrated to AWS

© 2017 Astadia, Inc. 11

 IBM Mainframe to AWS Reference Architecture

Understanding Amazon Web Services (AWS)

AWS has provided the technology and services that make running mainframe applications in the cloud
a safe, secure, reliable way to achieve high performance results and fuel future innovation for the
organization.

There are specific elements of AWS that are relevant to a mainframe modernization project. Below, we
address some of these – this is not the extent of all AWS services nor is it meant to exclude the use of
other AWS service offerings.

Let’s begin with the following description from AWS’ “Overview of Amazon Web Services,” April 2017:

The AWS Cloud provides a broad set of ability to operate production applications and
infrastructure services, such as computing databases that are more highly available, fault
power, storage options, networking and tolerant, and scalable than would be possible
databases that are delivered as a utility: on- from a single data center. The AWS Cloud
demand, available in seconds, with pay-as- operates 42 AZs within 16 geographic Regions
you-go pricing. From data warehousing to around the world, with five more Availability
deployment tools, directories to content Zones and two more Regions coming online in
delivery, over 90 AWS services are available. 2017.

New services can be provisioned quickly, Each Amazon Region is designed to be

without upfront capital expense. This allows completely isolated from the other Amazon
enterprises, start-ups, small and medium- Regions. This achieves the greatest possible
sized businesses, and customers in the public fault tolerance and stability. Each AZ is
sector to access the building blocks they need isolated, but the AZs in a Region are
to respond quickly to changing business connected through low-latency links.
requirements.
AWS provides you with the flexibility to place
AWS serves over a million active customers in instances and store data within multiple
more than 190 countries. We are steadily geographic Regions as well as across multiple
expanding global infrastructure to help our Availability Zones within each Region. Each
customers achieve lower latency and higher Availability Zone is designed as an
throughput, to ensure that their data resides independent failure zone. This means that
only in the region they specify. As our Availability Zones are physically separated
customers grow their businesses, AWS will within a typical metropolitan region and are
continue to provide infrastructure that meets located in lower risk flood plains (specific flood
their global requirements. zone categorization varies by Region).

The AWS Cloud infrastructure is built around In addition to discrete uninterruptable power
Regions and Availability Zones (AZs). A Region supply (UPS) and onsite backup generation
is a physical location in the world where we facilities, they are each fed via different grids
have multiple AZs. AZs consist of one or more from independent utilities to further reduce
discrete data centers, each with redundant single points of failure. AZs are all redundantly
power, networking, and connectivity, housed connected to multiple tier-1 transit providers.
in separate facilities. These AZs offer you the

In the following section, we’ll take a deeper look at the AWS portion of the Reference Architecture and
the components identified.

12 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

AWS Architecture Supporting the Migration of IBM Mainframe Applications

Your Cloud Environment Another storage option is Simple Storage Service

(S3). EC2 instances connect to S3 through APIs to
Amazon Virtual Private Cloud (VPC) lets you
access and store object data. S3 can be used for
provision a logically isolated section of AWS where
bulk data repositories or "data lakes" for analytics.
you launch and manage AWS resources in a virtual
network that you define. It’s your private area
AWS also offers Amazon Glacier (not shown
within AWS.
above) as a low-cost, reliable service for backup
and archiving of all types of data.
You can think of this as the fence around systems
you have in AWS. You have control over your
S3 is designed to deliver 99.999999999%
virtual networking environment, including
durability and scale past trillions of objects
selection of your own IP address range, creation
worldwide. These services are combined to meet
of subnets, and configuration of route tables and
the storage requirements of your mainframe
network gateways. You can use both IPv4 and
applications.
IPv6 in your VPC for secure and easy access to
resources and applications.
Databases
Computing Resources Amazon’s Relational Database Service (RDS) is
where all your legacy relational data will reside.
Elastic Compute Cloud (EC2) provides secure,
This includes any flat file data that’s been
scalable compute capacity in AWS. It serves as the
converted to relational. For example, all your flat
foundation upon which your application sits. It’s
files and IMS DB data that's been converted to
the container that holds the operating systems,
relational and migrated to RDS. DB2 data would
mainframe emulators, application executables,
also be migrated here.
and other supporting software that make up your
application.
This service is optimized for database
performance. It’s cost-efficient, has resizable
Depending on your specific circumstances, you
capacity, and is designed to reduce time-
may separate some pieces into their own EC2
consuming database admin tasks.
instances, or you may run everything into one
instance. For example, maybe you’ll have an EC2
RDS is available in several familiar database
dedicated to batch COBOL and another dedicated
engines, including Microsoft SQL Server, Oracle,
to Online. You may even segregate EC2s by
PostgreSQL, MySQL and MariaDB. You could also
applications.
set up a DB2 LUW server instance. However, you
may want to consider migrating your relational
Storage data to Amazon Aurora, a MySQL-compatible
Elastic Block Storage (EBS) can be thought of as a database that has been optimized for AWS and can
hard drive for storing data. Lots and lots of data. perform up to 5 times faster than MySQL.
EBS serves as the primary storage “device” for
EC2 instances running migrated applications. An analysis of your existing legacy databases and
applications will reveal all the changes required to

© 2017 Astadia, Inc. 13

 IBM Mainframe to AWS Reference Architecture

migrate your data to Aurora or any other RDBMS infrastructure security rather than application-
running in AWS. level security.

Load Balancing Monitoring

Applications with a high volume of transactions Every IT system needs to be monitored. AWS
require something to balance the workload. CloudWatch is a monitoring service for AWS cloud
Amazon Elastic Load Balancing (ELB) does just resources now running the legacy applications you
that. It automatically distributes incoming deployed to AWS.
application traffic across multiple EC2 instances to
achieve scalability, high-performance, and fault You use this tool to collect and track metrics,
tolerance in your migrated applications. It monitor log files, set alarms, and automatically
provides the load balancing capability needed to react to changes in your AWS resources. This data
route traffic evenly among your applications and is used to resolve problems quickly and keep your
keep them performing efficiently. migrated applications running smoothly – much
like you do on the mainframe today. Other cloud-
Security ready monitoring tools are available from 3rd
parties as well.
In the AWS environment, you’ll be using
Lightweight Directory Access Protocol (LDAP) for
accessing and maintaining distributed directory Source Control
information services. While there are other Just as you have products and processes to control
possibilities, this is most likely where you’ll map your application sources and manage application
your legacy application user IDs, passwords, releases on your mainframe today, you need to
permissions, etc. have a similar set of tools in AWS.

Hosting LDAP services on a smaller separate EC2 AWS CodeCommit is a fully-managed source
instance often makes it easier to maintain control service providing secure and private Git
independently of applications. However, a full repositories. It eliminates the need to operate your
analysis of your legacy security environment is own source control system or worry about scaling
required to determine how to best architect and its infrastructure.
configure security in the migrated system.
CodeCommit is where you’ll store your migrated
AWS Identity and Access Management (IAM) application source code and binaries, new source
enables you to create and manage AWS users and and binaries, an anything else you want to archive.
groups, and use permissions to allow and deny
their access to AWS resources. This is for AWS

14 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

Ensuring Project Success

Astadia’s Legacy Modernization practice has more than 25 years of experience migrating legacy
applications to modern platforms. Since mainframe applications are the mission-critical systems of the
enterprise, Astadia goes to great lengths to ensure a thorough and complete project plan is developed
for each legacy modernization project we undertake.

Astadia’s methodology recognizes the organizational impact that any project of this nature will have on
day-to-day operations, as well as the financial and business implications for organizations in both the
short and long term. Return on Investment (ROI), and Total Cost of Ownership (TCO), are carefully
calculated during this process, and are closely managed throughout the project lifecycle.

Astadia’s Mainframe-to-AWS Success Methodology has been refined over the course of 200+ successful
legacy migration projects, and has become an industry leading approach for our medium and large-
scale mainframe clients.

Astadia Mainframe-to-AWS Success Methodology

Discover
code, data, software,
processes, requirements

Design
architecture, databases,
interfaces, customizations

Modernize
code, static data, application
components

Test Mainframe
functions, data, interfaces, Application
performance, load-balance Deployment

Implement
deploy, migrate dynamic data,
system cutover, monitor

Manage
manage AWS environment,
maintain & enhance apps

Discover Design
Catalog and analyze all applications, databases, Astadia’s project team analyzes source code, data
networks, platforms, and processes in the client’s structures, end-state requirements, and AWS
portfolio. Document the interrelationships cloud components to design and architect the
between applications, and all external integration solution. The design includes details such as types
points in the client’s configuration. This is a key and instances of AWS components, transaction
input to Application Portfolio Management and loads, batch requirements, programming
Application Rationalization. language conversions and replacements,
integration with external systems, 3rd-party

© 2017 Astadia, Inc. 15

 IBM Mainframe to AWS Reference Architecture

software requirements, and planning for future

requirements. Implement
When migrated applications have been tested,
Modernize verified, and optimized, the process of deploying
Astadia employs an iterative, hybrid process of those applications may begin. In reality, many
automated code conversion and human deployment activities are initiated in parallel with
intervention to perform the necessary application earlier phases – things like creating and
changes. The technology behind the automation is configuring AWS component instances, installing
Astadia’s Rules-Based Transformation Engine. and configuring mainframe emulation software,
This tool preserves the business logic and rules of migrating static data, and other infrastructure or
the client’s legacy applications while removing framework activities. In some cases,
proprietary code that can only execute in the environments may be replicated to achieve this, or
source environment and not in AWS. While a least- existing environments may be re-purposed. The
change, least-risk approach is employed, some specifics of this may depend upon application and
source code or supporting components may be data characteristics and client preferences. After
converted to new languages for technical reasons dynamic data is migrated and validated, cutover
or to comply with client preferences. to Production mode can be completed.

Test Manage
Since most mainframe-to-AWS projects are Astadia offers a full range of managed services
typically an as-is migration of applications, testing solutions. Having gained significant application
can focus primarily on two areas: the components knowledge from the architecting and
that have been changed or replaced, and implementing the AWS solution, Astadia is well
application performance. For the most part, suited to take on the burden of managing and
“parallel” testing is sufficient; that is, the results maintaining the migrated applications and their
of operations in the source environment must AWS environments, or the dual environment in the
produce the same values in the target case of a partial migration. This offers clients an
environment, except where differences are opportunity to focus their development efforts on
expected due to platform changes. For strategic initiatives as well as address concerns of
performance testing, the focus is on ensuring the finding programmers skilled in maintaining the
user experience is as good or better than the legacy components still in use.
legacy environment and batch processes complete
within acceptable timeframes.

16 © 2017 Astadia, Inc.

 IBM Mainframe to AWS Reference Architecture

Conclusion
Amazon Web Services and Astadia combine to create a perfect next generation platform for your
mainframe applications portfolio.

Once the mainframe application set has been fully deployed on AWS, you will have the freedom to re-
engineer traditional applications in to a more contemporary computing style, modernize legacy
interfaces and integrate with other applications. In addition, many new services, like mobile and
wireless, can be easily connected to the cloud platform, thus enhancing the overall power of your new
cloud computing environment. Your investment will serve to support all the needs of your enterprise
and the future requirements of your business.

You don’t have to tackle this alone. Astadia has the experience, skilled experts and the technology to
successfully help you complete Legacy Migration projects of all scopes and sizes.

Astadia would be happy to hear from you about your specific Legacy Migration needs and how we may
be of service to you as you prepare to leverage AWS.

To connect with an Astadia expert, please contact us at:

Email: [email protected]
Web: www.astadia.com
Phone: (877) 727-8234

© 2017 Astadia, Inc. 17