Keep the platform moving

Keep data products running through breaking schema changes.

Schema Dam reads DataHub to find the real blast radius, contains only affected products, and writes verified recovery back for the next operator or agent.

READY · 6 STEPS NORMAL

All eight products are current.

Start the incident to see how Schema Dam handles an unmanaged breaking change without stopping the whole platform.

Affected0

Still running8

Backlog0

Checkpoint3

DataHub contextREAD

{ }Schema fieldOrders.category_code changed

↗Column lineageWhich products actually use it?

◎Identity & ownershipExact assets and responsible teams

Containment frontier Waiting for the breaking event…

Deterministic controllerACT

Runtime decision No intervention · 8 current

Gate stateOPEN

Serving modePRIMARY

CanaryNOT RUN

ApprovalNOT NEEDED

Canary-tested artifact none

Ordered replay0 / 10

Product boundariesLIVE

Category AnalyticsOrders input · last-good availableCURRENT

Inventory ForecastOrders + Products companion inputCURRENT

RecommendationOrders + Products · fallback readyCURRENT

5 unaffected productsCURRENT

Order Count · Settlement · Search Index · Fraud Detection · Revenue Dashboard

Native DataHub incidentNOT CREATED

Impact & owners3 affected · 5 uninterrupted

Approval evidenceExact SHA · canary proof

Verified resolutionReplay · health · resume

What this is: a deterministic browser replay of the checked-in reference scenario. Runnable proof: SQLite + HTTP boundary by default; live DataHub E2E is documented in the repository.