Canary Releases Quiz

What you'll be asked

1. What is the primary purpose of a Canary Release in machine learning model deployment?
2. During a Canary deployment, what is the most important indicator that the new model version is performing poorly?
3. When implementing a Canary Release, how is traffic typically routed to the new model?
4. What is the recommended 'rollback' strategy if a Canary release shows signs of failure?
5. In a sophisticated Canary deployment using 'Shadow Traffic' (Dark Launching), what is the key difference compared to a standard Canary release?

Answers and explanations come after you submit.