Containerization Can Take Mainframe Software Testing to the Next Level
While the sheer processing power and operating efficiency of the mainframe have led to an incredible track record over the years, the platform’s future depends on its ability to keep pace with cloud-based systems that have mastered the intricacies of modern development methodologies, specifically DevOps.
The mainframe world has indeed embraced Agile and DevOps development, but some aspects of the continuous delivery (CD) cycle pose difficulties. In particular, the need for rapid testing continues to be a thorn in the side of enterprises that aim to put incredibly complex applications into production faster and with fewer experienced personnel resources. In some cases, these delays can lead to four- to six-week lead times for minor application changes, and 45 percent of customers deal with delays in mainframe application delivery. For mainframe business-service delivery to thrive another half-century, these hurdles must be overcome.
Bringing Distributed Containerized Testing Methods to the Mainframe A survey conducted by Vanson Bourne for Compuware found that 90% of IT leaders in mainframe shops believe test automation is the key ingredient in accelerating the pace of innovation. Thanks to the increasing complexity of application environments, 92% of teams are spending more time testing code than ever. As a result of the intensive demands of manual testing methodologies, these teams are forced to cut corners in an effort to keep up with the application delivery cycle which poses availability concerns, not to mention some serious security risks. When defects are caught during coding and unit testing, Deloitte estimates they cost about $80 per fix [see Figure 6 below]. Those that make it into production come with a much heftier price tag of $7,600 each in lawsuits, lost customers, and brand damage.
Overall, testing accounts for 51% of the time required to release a new application, feature, or functionality on the mainframe compared with 39% on other platforms. The holdup is largely due to complex data testing requirements, unavailable systems, and an inability to accurately model production environments. Fortunately, application containerization can help better model the main business logic on the mainframe and the platform’s integration with other systems, whether internal or third-party. As a result, teams can rapidly create mainframe testing repositories, enabling faster and more accurate test cycles and at a reduced cost.
With the containerized method automating the creation and execution of functional tests, developers can save immeasurable hours of setup and orchestration of Unit Tests, User Acceptance Tests (UATs), and Integrated Tests. That means product owners can easily test code against simulated requirements, ensuring a greater degree of accuracy and performance in production. Containerization has enabled parallel testing in other development environments for years, and it’s time for mainframe organizations to start reaping the benefits. By automating testing cases, mainframe teams can meet business demands for speed, innovation, and customer experience, while drastically reducing the frequency at which bad code makes it into a production environment. For more information or a demonstration of our z/OS Application Environment for Parallel Testing (APT) solution.
