Demo: a canary that catches a regression
On This Lesson 3 topics
This lesson is planned. The outline below lists the topics it will cover. The full lesson has not been published yet.
In this lesson
run a canary where automated analysis halts a seeded regression; prepares SRC-011-LAB1.
Draft outline details
Step by step
- State the change on the order subsystem and the guard metrics with thresholds.
- Deploy the new version at 5% traffic; show baseline and canary dashboards side by side.
- Seed the regression (a slow query that raises p99 under load) and apply realistic traffic.
- Watch the analysis fail at the first stage and traffic revert automatically; read the verdict.
- Planned mistake: use a canary too small and too short to see the difference, so the bad version passes. Fix: raise volume and bake time to match the detection need.
- Record what the run proved and what it could not.
Check your understanding
What makes a canary stage trustworthy?
Show answer
Enough traffic and time to detect the difference you care about.
Practice
Apply this lesson with the module's lab, exercises and quiz.
Marking a lesson complete is optional and you can undo it any time. Your progress stays in this browser.