When systems stop behaving: controlled chaos in practice

Real systems get messy, slowing down, misbehaving, drifting out of sync and exposing every assumption we make, and this session shows how bringing controlled chaos into a safe environment can change how testers and developers think, learn and solve problems. You’ll see how shifting states, strange responses and unpredictable load reveal gaps in testing and drive better design choices, as we explore practical ways to create these environments without complex tooling and why they consistently lead to stronger, clearer thinking across a team.

The session includes a short live demo where the audience becomes part of a distributed system, experiencing instability and inconsistency in real time and giving a simple, memorable way to see how systems behave when the world stops being clean and predictable. You’ll leave with practical ideas you can apply immediately, including simple ways to introduce controlled chaos and help teams think more clearly about how systems really behave.

What you’ll learn

• Understand how real systems behave under non-ideal conditions, including drift, partial failure and inconsistency
• Recognise where happy-path thinking breaks down in both testing and system design
• Use simple techniques to simulate failure conditions such as latency, partial responses and inconsistent state in a safe environment
• Apply practical techniques to observe system behaviour and use those insights to improve design, debugging and testing decisions