Join me for: Managing Test Data in Enterprise Environments

Test Data Management Strategies for Real Enterprise Software

I haven’t done a lot of conference presentations, but do love a captive audience. The management of Test Data, especially in big, complex, real software is something I’m passionate about, and is the topic of my talk at the 2024 StarEast Conference in Orlando, FL.

Here’s a sneak peak at what I’ll be covering.

Effective Test Data Management is Critical

Test data management is the foundation that every successful automation strategy sits on. When enterprise automation efforts fail, it is almost always due to challenges setting up or maintaining data, not failures in how the test was driving that application.

In other words, as software gets more complex, managing test data becomes *the* big, hairy problem automation engineers must deal with.

Effective Test Data Management is HARD

The software underlying today’s businesses is nothing like the clear, simple, well-designed, and understandable diagrams found in your off-the-shelf architecture books or favorite engineering blogs. Real software is complex, polyglot, poorly documented, ever-changing, and rarely built the way you would if just from scratch. That’s life. That’s reality.

A test data management strategy has to deal with all this. It has to deal with the fact that data is shared, that data is interconnected, that it has implicit dependencies, and the person who really understands the data you are most interested in retired to the Bahamas last year and probably won’t take your call.

Every test data management strategy is desperately trying to keep the 100 separate pieces of data critical for this test exactly as they need to be, at all times, while other people or other tests go about their day doing exactly the opposite. If any one piece of data get intentionally or inadvertently changed, tests start failing, and flakey tests are the kryptonite of successful automation strategies.

Strategies for Managing Test Data

There is no one tool that solves the problem, no secret answer I have that you don’t, and unfortunately AI can’t save us. This is a hard problem and the solutions will depend on variables specific to your organization, your technology stack, and your zodiac sign. Maybe not the last one, but you get the idea.

I’ll cover how effective test data management strategies will consider things like:

• Environment decomposition
• Data lifecycle management
• Data sourcing
• Data location & test proximity

… and some more stuff you’ll have to attend to hear about.

This is a huge, important topic and I hope I can do it justice in the 50 min I’ve been given.

If you happen to be attending or speaking at StarEast this year, I’d love to meet up and chat software quality.