Forum Discussion
Deployment Pipeline
What is the principal, non-obvious risk when using the Selective Deploy feature in a Power BI deployment pipeline to push only a single modified Report from DEV to TEST, assuming the underlying Semantic Model (Dataset) has not been deployed/refreshed and has diverged between the stages?
Amik_singh The risk is Silent Schema Mismatch. The deployment pipeline copies the Report's metadata (including its visuals and field references) but does not force a deployment or schema check on the existing Semantic Model in TEST. If the single Report relies on a field or measure that was deleted or renamed in the DEV Semantic Model since the last full deployment, the Report will be deployed successfully, but will immediately display as broken in the TEST environment, potentially without an easy-to-trace error that links back to the selective report deployment.
please mark it as a solution if this helps.
1 Reply
Amik_singh The risk is Silent Schema Mismatch. The deployment pipeline copies the Report's metadata (including its visuals and field references) but does not force a deployment or schema check on the existing Semantic Model in TEST. If the single Report relies on a field or measure that was deleted or renamed in the DEV Semantic Model since the last full deployment, the Report will be deployed successfully, but will immediately display as broken in the TEST environment, potentially without an easy-to-trace error that links back to the selective report deployment.
please mark it as a solution if this helps.