Performance testing is often treated like a checkbox, to see if it could handle the load. In reality, each test answers a different question about your system’s behavior. In this 15-minute talk, we’ll cover the main types of performance testing and why they matter:
To make it stick, we’ll use everyday analogies like a coffee shop during a morning rush, a highway during a holiday weekend, a gym workout gone wrong, and a car left idle overnight. You’ll learn what each test is good (and bad) at revealing, common mistakes teams make, and how to choose the right test instead of just throwing more virtual users at it. Whether you’re a developer, tester, or engineering manager, you’ll leave with a better mental model for performance testing.
- Load testing to see how your system behaves under expected traffic
- Stress testing to find breaking points
- Spike testing for sudden surges
- Soak testing for long-running stability
- Capacity testing to plan for growth
To make it stick, we’ll use everyday analogies like a coffee shop during a morning rush, a highway during a holiday weekend, a gym workout gone wrong, and a car left idle overnight. You’ll learn what each test is good (and bad) at revealing, common mistakes teams make, and how to choose the right test instead of just throwing more virtual users at it. Whether you’re a developer, tester, or engineering manager, you’ll leave with a better mental model for performance testing.
Heather Thacker
Gatling
Heather is a Developer Advocate at Gatling where she enjoys building content and community around making performance testing approachable and essential to teams. Her perspective is shaped by four years as a software developer followed by two years advocating for DevOps best practices. She’s an international speaker that enjoys comparing theory with reality of software development lifecycles, using practical examples and analogies. Her goal is to give engineers different perspectives that will build the confidence they need to make performance a core part of their workflow.