Get Your CertDatabricks Certified Machine Learning ProfessionalPlayCompete
‹ All Databricks Certified Machine Learning Professional questions

Databricks ML Professional Practice Question 5: Cross Catalog Model Promotion

Databricks Certified Machine Learning Professional · Test 1 · question 5 of 10

Model Lifecycle Managementhard

An organisation uses three Unity Catalog catalogs: dev, staging, and prod. Version 7 of staging.ml.churn has passed all validation. Both the production batch job and the production serving endpoint load models:/prod.ml.churn@champion. Promotion must meet these requirements:

- Production promotes exactly the artifact that was validated, with no retraining. - Production workloads never read from the staging catalog. - No production code or job configuration changes. - Rollback is a single metadata change.

Which approach meets all of these requirements?

Select one answer
Show answer and explanation

Answer:

  • D. Call copy_model_version to copy staging.ml.churn version 7 into prod.ml.churn, then set the champion alias on the new prod version

MlflowClient.copy_model_version() copies an existing Unity Catalog model version into another registered model, even one in a different catalog. The validated artifact becomes a new version of prod.ml.churn without retraining. Its version number in prod may not be 7. Next, set_registered_model_alias("prod.ml.churn", "champion", <new version>) moves the alias. Both production workloads resolve @champion the next time they load the model, so no code changes. To roll back, point the alias at the previous version, which is a single metadata change. The copy needs read access on the source model and ownership or CREATE MODEL VERSION on the target.

Option A is not possible. Models in Unity Catalog do not support stages. Environments come from the three-level namespace and deployment status comes from aliases.

Option B points production at the staging catalog and changes production configuration, which breaks two requirements.

Option C sets the alias on the wrong model. The production URI refers to prod.ml.churn@champion. An alias on staging.ml.churn has no effect on it, and granting EXECUTE does not change that.

Reference: Manage model lifecycle in Unity Catalog

Practice Test 1 ›

Study guides for Databricks Certified Machine Learning Professional

Best MLOps certifications 2026: which test real MLOpsThe best MLOps certifications in 2026 from Microsoft, Databricks, Google and AWS, ranked by how much each exam tests MLOps, with cost and order.
Best AI certifications 2026: AWS, Google, Microsoft, ClaudeThe best AI certifications in 2026 from AWS, Google, Microsoft, Claude and Databricks, by level, with price, status and which one to take first.