1 min lesson
Defense in depth
Choose two examples from the table in "Defense in depth" and explain what each teaches you to do.
Step 1 of 2
The two layers and what each one is forpre-deploy vs post-deploy
- Layer
- Pre-deploy
- Method
- A/A baseline + canary on a traffic slice
- Catches
- Gross regressions before full rollout
- Blind spot
- Rare slices, slow-burn effects, novelty
- Layer
- Post-deploy
- Method
- Change-point detection on the live series
- Catches
- What the canary's sample missed
- Blind spot
- Lag - users may feel it before the point shifts
- Layer
- Continuous
- Method
- Anomaly detection per segment
- Catches
- Drifts with no deploy (a provider degrades)
- Blind spot
- Alert fatigue if thresholds are naive
| Layer | Method | Catches | Blind spot |
|---|---|---|---|
| Pre-deploy | A/A baseline + canary on a traffic slice | Gross regressions before full rollout | Rare slices, slow-burn effects, novelty |
| Post-deploy | Change-point detection on the live series | What the canary's sample missed | Lag - users may feel it before the point shifts |
| Continuous | Anomaly detection per segment | Drifts with no deploy (a provider degrades) | Alert fatigue if thresholds are naive |
Defense in depth: each layer covers a failure the others structurally cannot see.
Start a canary by proving your measurement is trustworthy. An A/A test routes two arms to identical code and confirms the metric shows no difference. If your A/A flags a phantom regression, your detector is miscalibrated and every real alert is suspect.
Learn more
Full explanation
The Canary-to-Live Pipeline
Interactive diagram. Step through it with the Next and Previous controls below, or Tab to a region to read its detail.
diagram: flow
Each step earns the next. The gate is the promotion decision - a failed canary is a non-event for users, which is the whole point.