Insufficient IT performance, lack of connection between the mainframe and the cloud, high costs for licenses and IT operations, and lack of personnel and skills in “old” programming languages are all reasons for modernizing the mainframe. zFuture is the tool-based modernization suite for modernizing applications on the mainframe, integrating the mainframe into new IT environments, or replacing the mainframe.
T-Systems approaches modernization projects according to the DECIDE.PLAN.ACT. scheme: the starting point of the project is the analysis of the current mainframe technology stack. This is carried out as part of an assessment based on the Gartner 7R approach, the “standard” for IT legacy modernization projects.
This results in T-Systems recommending to customers which applications should be modernized on the mainframe or by migration, and which should be removed or replaced by standard solutions (DECIDE). A detailed examination of the code of applications, databases, data management, schedulers, batch processes, archiving, file transfer, security, access, encryption and interfaces is carried out in order to plan the concrete implementation. A specific time and cost plan for the modernization (PLAN) is derived from this. A validated business case, risk and project management with realistic timeframes for the transformation and modernization of the mainframe components, and statements on test scenarios, tests and target architectures underpin the modernization project (ACT).
A mainframe-based central time recording system needed to be modernized. Increasing license costs for the database, both to date and forecast for the future, were an enormous cost driver for mainframe operation. The time management system used the historical integrated database management system (IDMS) for its database. The system saves and processes the data in batch mode.
T-Systems presented several modernization options. The decision was made in favor of transforming the IDMS database to DB2. In contrast to IDMS, DB2, which is established in the mainframe environment, offers a wide range of modernization options. In addition, the relational DB2 databases not only support COBOL, but also Java, C#, .NET, C++ and SQL. This cross-database migration is the most complex migration path for databases. Data models and formats must be transformed. The greatest effort lies in the conversion of the source code for the database management system (DBMS). The DBMS manages the database. For example, it offers interfaces for creating, editing, and querying. Over the years, the code for the IDMS database had grown to 4.5 million lines. This had to be adapted and corresponding dependencies – a common source of errors in mainframe refactoring projects – had to be taken into account. In addition, the code was subject to frequent maintenance and further development cycles, for example, to comply with new legal requirements or to reflect developments within the company. For cost and efficiency reasons, the code transformation was automated with T-Systems partner Astadia. Astadia tools are very well suited to transforming code from legacy languages into state-of-the-art programming languages such as Java. The code was transformed and tested in several cycles, and errors were identified and corrected. The final solution was developed within 18 months. It allows a degree of automation of 99.5 percent and ensures zero-error transformation for the code transfer.
A cloud migration approach – here the customer’s goal is to implement its “off-mainframe” strategy towards the cloud in order to achieve significant cost savings in mainframe hardware, software licenses, and operating costs. The initial focus is on the modernization of a business-critical, COBOL-based application in the Production division. It is a system with many instances and interfaces requiring migration, with worldwide distribution and use in Europe, Asia, and America and 30 separate factory installations.
A detailed assessment serves as the basis for understanding the application, dependencies, the role of the application in the business model, the associated infrastructure, and the identification of risks. As part of the cloud migration project, the customer application is being migrated to the software defined mainframe (SDM) of T-Systems’ technology partner LzLabs. Other measures include the modernization of the AJM scheduler and the switch to Control-M as well as the transformation of the macro programming language to native COBOL. In addition, the desired Azure cloud environment is prepared for the SDM and a new DevOps development process is designed and implemented. As part of systematic risk management, regression tests are designed and implemented for parallel operation to avoid risks when migrating application elements from the mainframe to the cloud. The decision-making basis for the implementation of the modernization project comprises a convincing business case, measures to prevent data loss, and the elimination of potential performance problems (5.5 GHz mainframe vs. 3.6 GHz Cloud).
The COBOL-based, central HR system used at the German production sites of a leading German car manufacturer needs to be migrated to a hybrid cloud environment. Data and the avoidance of data loss during migration are of great importance to the customer.
The starting point for the project is the automated source code analysis via the T-Systems migration partner Astadia. The Astadia transformation tools are used extensively and the AJM job networks are analyzed automatically by the technology partner apsware.
As part of the transformation, the COBOL source code is migrated to JAVA, the JCL code to Bash, the transformation of REXX and CLIST scripts and assembler programs to another scripting language. For the database migration, Astadia DataTurn is used for the DB2 migration in Postgre SQL. New REST services are also being implemented. Testing is a key customer requirement. Test scenarios exist on the customer side and can be used. Test automation is an essential task within the migration project in order to minimize effort and implementation time and eliminate manual sources of error. Functional and regression tests are carried out as part of the project in order to minimize the risk during migration to the target platform. Load and performance tests were carried out to verify the integration of all components on the target platform.
The use of modern cloud migration tools, proven mainframe expertise, many years of experience in integration and migration projects, and competence in testing enabled T-Systems to both gain the trust of the customer and implement the legacy modernization projects.
Our customer consistently pursued a cloud-first strategy and therefore also outsourced the mainframe. The strategy implementation process began with a thorough planning and assessment using the AWS Migration Acceleration Program (MAP) tools, which help organizations assess their mainframe workloads, identify the best migration strategies, create a clear roadmap for the replatforming journey, and select the appropriate replatforming partner.
The solution involved the migration of customer applications from the mainframe to an x86 Software Defined Mainframe (SDM) provided by technology partner LzLabs. This included supporting the design, construction, and testing of the SDM environment where all application environments (ENG, DEV, INT, PROD) were migrated without having to rewrite the COBOL application code, because all modules can be executed on x86. The cloud environment was prepared for AWS using a Cloud Migration Framework and by setting up the AWS Landing Zone. T-Systems took over the SDM on AWS and the management of the AWS cloud services. Together, these AWS services – EC2 for scalable data processing, IAM for secure access management, S3 for persistent storage, and AWS backup for automated data protection – formed a cohesive support system that enabled efficient and secure replatforming of mainframe applications to the cloud.
The solution drastically reduced the customer’s annual license costs and hardware, software, and operating costs by around 70% by replacing the mainframe. The customer was able to use the freed-up capital for other activities. Know-how bottlenecks in mainframe technology were closed. As a mainframe and AWS Cloud Technology Partner, T-Systems offers the customer expertise from a single source and facilitates a mainframe offload strategy and cloud migration. What’s more, the application operation is managed by a proven cloud service provider.