Harnessing the Twister: Zephyr Data-Driven Testing and Insights
summary
Twister is Zephyr’s Python test runner and manager that actually builds, flashes and runs thousands of tests in parallel across racks of devices and CI, and collects actionable data like coverage, footprint and test-run metadata for trend, anomaly and ML/LLM analysis. It supports hardware classes, CMake-only and build-only flows, reuse via runtime artifact cleanup and aggressive no‑clean, and now emits per-test JSON and coverage per instance so reports show which test covers which lines. Multi-image support is arriving to run client/server scenarios.
AI-written summary · about
chapters (8)
- 0:00Twister and Zephyr intro
- 6:49Twister for developers
- 13:20Twister evolution
- 21:59Optimizing Twister workflows
- 28:46Disk space management
- 33:45Advanced test analysis
- 36:40Twister roadmap
- 40:14Q&A
related talks
49mHarnessing the Twister: Zephyr Data-Driven Testing and Insights
35mTwister: A Test Utility for Contributors
20mIntroduction To Twister and Adding Support for Hardware Runners
23mTwister V2: New Framework for Testing Based on Pytest. Overview & Current Status
43mPytest Tests in Twister
43mPytest Tests in Twister
46mZephyr Application Development Using TDD: Getting Started
46mZephyr Application Development Using TDD: Getting Started