AWS Partner Network (APN) Blog
How to Succeed at Large-Scale Mainframe Replatforming with TmaxSoft OpenFrame on AWS
By Manoj Aerroju, Vice President of Pre-Sales – TmaxSoft
By Bruno Sahinoglu, Mainframe Modernization Solutions Architect – AWS
Mainframes continue to support major applications in many enterprises, running the software that is core to their business.
With age have come fixes, bolt-ons, and spaghetti architecture, but some businesses believe replacing the mainframe with a new one or totally rewriting applications is too expensive for their budgets.
The good news is that replatforming with TmaxSoft OpenFrame, TmaxSoft Tibero, and Amazon Web Services (AWS) is a cost-effective solution to aging mainframes. In this post, we describe the replatforming components, why the combined solution is unique, and best practices for implementation.
TmaxSoft is an AWS Mainframe Modernization Competency Partner and global software innovator focused on infrastructure and data modernization. TmaxSoft’s solutions offer enterprises viable alternatives to support their global IT powerhouses and drive competitive advantage.
Mainframe Replatforming with OpenFrame
Replatforming moves existing mainframe applications to x86 open distributed systems. When that system is on AWS, you can scale resources to match user demand as OpenFrame supports both horizontal and vertical scaling and integrates with AWS services like Amazon Relational Database Service (Amazon RDS), Amazon Simple Storage Service (Amazon S3), AWS Lambda, Amazon Aurora, and Amazon EMR.
OpenFrame facilitates replatforming by migrating mainframe applications into a three-tier architecture without changing the business logic and providing tools for modernizing mainframe data to a standalone database tier that supports industry-standard SQL relational databases.
Figure 1 – Mainframe to OpenFrame migration.
The migration starts with a source code analysis by a tool developed by TmaxSoft called OFMiner that sets the stage. TmaxSoft’s advanced automated migration tools move the code to a distributed environment.
Application source code is recompiled in the AWS environment using OpenFrame compilers. There are minimal to no code changes required to the application code during recompilation. Business logic remains the same while the data is being modernized to ASCII character-set, which is more portable than the EBCDIC character-set.
OpenFrame has a unique value proposition with strong technical features:
- In addition to COBOL, C, and PL/1, OpenFrame replatforming support includes Assembler and Easytrieve Plus, solving the complexity of mixed code and the risks associated with piecemeal solutions.
- OpenFrame supports the replatforming of hierarchical IMS DB and IDMS DB databases to a relational database (such as Tibero or Amazon Aurora) via an abstraction layer that handles all of the current application calls without modification. This migrated database is available natively in the replatformed environment as a relational database. This ability to migrate your data, as well as your mainframe applications, is a complete and unique proposition for replatforming.
- OpenFrame supports the major relational databases within Amazon RDS such as Aurora.
- Mainframe interoperability allows you to migrate application modules while still using the mainframe, limiting downtime. This is accomplished with Inter System Communication (ISC) features including Distributed Program Link (DPL) and Dynamic Transaction Routing, as well as communication via MQ and DB2 connect/client.
- OpenFrame supports high availability architecture and can support multi-node deployments across multiple AWS Availability Zones.
- OpenFrame supports multiple COBOL compilers in the market, including OFCOBOL, the TmaxSoft compiler.
- Although OpenFrame provides an integrated development environment (IDE) for maintaining and deploying your software, TmaxSoft also supports the integration of popular IDEs such as Eclipse.
- RACF compatibility allows quick security migration and administration of the system.
OpenFrame Architecture on AWS
The OpenFrame solution runs existing mainframe applications on an x86-based Amazon Elastic Compute Cloud (Amazon EC2) instance, as shown in Figure 2 below. Typical mainframe technologies are used—3270 screens, web, COBOL, JCL, or DB2. OpenFrame enables you to move difficult-to-maintain databases to Tibero or Aurora on AWS.
Figure 2 – OpenFrame on AWS architecture.
Best Practices for Program and Data Migration
Following these best practices facilitates the migration of programs and data from mainframe to OpenFrame, and allows customers to choose from a wide range of target RDBMS such as Tibero, Aurora, DB2 LUW, Oracle, and Tibero.
Mainframe Discovery and Inventory Assessment
Once it has assessed the current mainframe environment and its applications, TmaxSoft provides a detailed report that includes an estimate of TmaxSoft license pricing and maintenance, a list of any third-party software that may be required, an overview of any identified risks or gaps in functionality, recommended target hardware/cloud specifications, inbound-outbound interfaces analysis.
The report includes expected benefits and an estimate of the project duration that includes details on tasks and roles and responsibilities as well. The report also gives the recommended target hardware/cloud specifications and an AWS design and infrastructure estimate for the services that would be required, such as a computing instance (Amazon EC2), elastic load balancer, and more.
Clean Your Unused Source Code with OpenFrame
The assessment process identifies any unused source code (programs, JCLs, PROCSs) which helps you narrow down migration and reduce the time and effort to replatform.
At a large financial services company, TmaxSoft found that almost 50 million lines of code were not actively used in the application being replatformed, which in turn helped narrow down the scope and reduce the migration and testing time.
Prepare the Open System and Migrate Your Resources
Don’t migrate everything all at once. Instead, choose one mainframe application and data at first conversion as a pilot. Bring the remaining systems and databases over after you have identified any system anomalies.
Test System and Database Performance
After the migration, run tests on a pre-production environment to see how well everything performs by staging use of the new system and databases by departments or functions. Once on the target environment, test one or two production monthly cycles to validate results and measure performance.
The flexibility of AWS allows you to launch, scale-in, scale-out, and shut down instances until you reach the expected performance metrics. You should not turn off the old system right away; instead, wait a for a comfortable period—this will vary—to make sure everything is operating to expectations.
During this time, data is synchronized between the old and new environments on a daily basis, allowing to switch over to the new environment completely.
Your customers and employees expect a fast, personalized experience that resembles what they get on mainframe 3270 terminals. Replatforming delivers this expectation on several levels.
TmaxSoft OpenFrame increases agility and flexibility, enabling you to configure active clustering on your AWS environment. This serves as the foundation for the highly available and scalable solution in this always-on world.
Replatforming creates an environment that can scale with AWS resources based on business demand. It can also unlock the value of your mainframe apps by exposing them to web services for mobile and digital applications, transforming the customer experience.
For more detailed information about OpenFrame visit the TmaxSoft website.
TmaxSoft – AWS Partner Spotlight
TmaxSoft is an AWS Mainframe Modernization Competency Partner and global software innovator focused on infrastructure and data modernization.