Regression Testing for Mainframe: How to Maximize Speed and Security

Maria Homann

Automation Expert

Most mainframe green screen applications are custom built and rely heavily on the institutional knowledge of key engineers in order to grow and integrate with newer systems like SAP and Salesforce.

With growth comes the need for more regression testing, but the existing strategies for accelerating regression testing are not keeping pace and sacrificing security. So what can businesses do to accelerate their regression testing securely?

We’ll start by reviewing the definitions of mainframe, regression testing, and how the two work together. Then we’ll dive into how to optimize your regression testing for mainframe. 

Understanding green screen, mainframes, mid-sized servers, and mainframe emulators

Green screen automation and mainframe automation are two sides of the same coin.

Green screens are what you use to interact with a mainframe, and mainframes are the servers where information is stored. Confusingly, the two terms are often used interchangeably.

The term mainframe is sometimes used as an umbrella term to cover: 

  • Mainframes (a high-end IBM server)
  • IBM AS/400 and iSeries, and DECsystems (mid-sized servers)
  • Mainframe emulator software that runs on a PC

For the sake of simplicity, we’ll continue to use the term mainframe to cover all of the above in this blog post.

A simple way to explain the relation between these terms is that green screens run on employee PCs and connect to either a mainframe, mid-sized server, or mainframe emulator via an encrypted connection.

Green-screen-connects-mainframe

Learn more: Everything You Need to Know About Green Screen Automation

Regression testing: what and why?

Regression testing is a hugely important, yet often down-prioritized, part of quality assurance. It ensures that bugs aren’t released into the product when new features are developed. 

Many QA teams are under immense pressure to release new features fast and frequently, and for this reason, regression testing tends to be rushed or skipped.

And this pressure only increases over time. As the product grows, so does the amount of features and functionalities to test. Resources rarely grow at the same pace, so finding a way to keep up without adding more hours can be a challenge.

TA_2_Test-cases-accumulation

Test automation helps solve this problem. Though test automation tools are also an investment in time and resources, they can help keep the testing burden to a minimum and save a lot of time and resources down the line.

Regression testing is ideal for automation, since it’s predictable and highly repetitive:once you’ve designed a regression test and added it to your regression suite, you can run it at every release without any additional effort from the team.

Related reading: The Difference Between Regression Testing Vs Retesting

Why should you run regression tests on a mainframe?

We’ve established the importance of regression testing in general. But in the context of mainframe testing, there are extra reasons to prioritize regression testing.

Most Banking, Financial Services, and Insurance (BFSI) organizations have mainframes as a part of their IT infrastructure. In addition to mainframes, infrastructure like this often includes other systems, databases, and technologies, such as mobile, desktop, and web applications. In other words, BFSI IT infrastructure tends to be large and complex.

With many complex processes and data security at stake, it’s all the more important that BFSI organizations have the processes in place to secure quality releases and minimize risk of software errors, including:

  • The risk of updates interfering with business-critical processes and business continuity.
  • The risk of data loss or data leaks as a result of faulty integrations.
  • The cost of repetitive, manual testing that cannot be scaled without adding additional human resources.
  • The cost of increasingly rare green screen experts that know legacy programming languages.

So what can businesses do to speed up their regression testing in mainframe while maintaining the security of their IT infrastructure? 

Related reading: How to do regression testing (faster)

How to optimize your regression testing

To determine how to speed up your regression testing, you first need to understand where you are in your testing journey. 

Your strategy can be categorized into four buckets. 

  1. No regression testing. Here, the quality of your software is probably suffering. This pushes people towards step number two, manual regression testing. 
  2. Manual regression testing. The quality has improved, but it’s still not ideal because you can’t get the coverage you need with manual testers. To speed up testing, you adopt test automation.
  3. Automation framework (e.g. Selenium). You’ve sped up testing, and you’re catching bugs but some are still slipping through the cracks. The automation can only be developed by a few highly technical test engineers, and the maintenance is very complex making testing inefficient and difficult to scale. This pushes teams to search for an alternative scalable solution. 
  4. Scalable automation platform. Here, testing is highly efficient. It can scale across teams without requiring technical know-how. It’s easier to implement and develop so the value is realized sooner. It can be used across your entire technology stack, whether in mainframe, web or desktop. 

To avoid staying in a high risk state, businesses move onto the third and fourth bucket. However, the difference between the last two steps are vast. One requires costly development and maintenance work, making it difficult to scale. For this reason, more businesses are moving towards the ideal state of adopting scalable test automation for improved speed and security. 

How can you get there? 

Leapwork is a no-code test automation tool that enables business users and QA managers to build automation. We enable businesses to accelerate their time to market and improve the quality and security of their software, all with the same visual language.

  • Visual language with no-code. With Leapwork’s visual language that anyone can use, both technical and non-technical people can build, maintain and scale test automation enterprise wide. Build any number of functional tests, even the most complex ones. 
  • Extensive enterprise solution. Scale your regression testing across technologies and applications while seamlessly integrating into your DevOps ecosystem. 
  • Easy to build, scale and maintain. Our unique visual drag and drop building blocks, along with our recording approach lets technical and non-technical’ s build test automation. Teams are better equip to isolate error, no matter what application or software is in use. Quickly and efficiently debug and troubleshoot in a single flow using synchronized video, process flows and audit logs. 

Want to learn more about no-code regression testing in mainframe green screen? We’ve create a simple and easy guide to green screen testing where we highlight how to speed up your testing.

New call-to-action