Modernized Mainframe Testing
Overview: A significant portion of the software delivery lifecycle is spent on tedious and often repetitive manual testing. Automated testing catches issues earlier, improves code quality, and frees developers to do what they do best—write code.
In the tenth episode of the Building a Better Software Delivery Platform podcast, Compuware Topaz for Total Test Product Manager Steen Brahe tells host Rick Slade that, depending on the circumstances, mainframe developers spend 30-80% of their time on testing. That’s a lot of time that could be spent on coding instead dedicated to repetitive manual tasks.
To offer a little more perspective, a recent Vanson Bourne survey, commissioned by Compuware, shows that 80% of the 400 senior IT leaders surveyed feel that bugs will leak into production unless they can automate more testing. And 90% feel that increasing test automation could be the most important factor to successfully accelerate innovation on the mainframe.
Looking at these numbers, the case for employment of automated testing is a no-brainer. Those surveyed overwhelmingly feel that it would increase code quality, and even more overwhelmingly feel that it would significantly increase innovation on the mainframe. Imagine being able to reclaim 10, 20, or even 60% of a developer’s time and dedicate it to writing new code that provides value to your customers.
But another, more shocking statistic stands out in the survey results. Just 7% of survey respondents currently automate the execution of test cases. Seven percent.
Viewed together, these statistics show why testing is such a crucial component of adopting a DevOps-centric software delivery system (SDS). The point of implementing a modern SDS is to increase the velocity of code delivery without sacrificing quality. Automated testing helps on both fronts.
Developers no longer have to be bogged down manually, and repeatedly, creating and executing test cases. Automated testing not only frees developers to write more code, it reduces the time span of the testing phase, effectively multiplying the efficiency gained. As Steen explains,
Instead of using time on manual testing, they can focus on what gives value to the company, and that is to develop business functionality faster, and that is also what the developers find most fun.
Automation not only saves time, it improves code quality by shifting testing left. Code changes are automatically tested using saved proven test scenarios, reducing the chance of untested or improperly tested code moving through the software delivery pipeline. Bugs can be caught and corrected earlier in the lifecycle instead of being found in later stages or, even worse, in production. Automation thus improves the quality of tests, thereby improving the quality of code.
New advancements also increase this efficiency and shift left. In July, Compuware released its 23rd consecutive quarterly update, which included a new integration between Topaz for Total Test and Topaz for Enterprise Data. The new feature removes time spent waiting for specialists or database administrators to set up data by allowing developers to enter it directly into the test environment.
With these obvious benefits, it’s no wonder that so many survey respondents see automated testing as playing an integral role in the future of innovation on the mainframe. It is possible to make developers more efficient, significantly improve delivery velocity, and improve quality, thanks to automated testing tools like Topaz for Total Test.
Listen to Chapter 10: Automated Testing I to hear Rick’s entire conversation with Steen, including a discussion of virtualized and non-virtualized testing and their roles in the development lifecycle. And don’t miss the second episode on automated testing, Chapter 11, in which Rick talks with Compuware Product Managers Irene Ford and Kevin Corbett about test data management.