Azure ML Model Version Lifecycle: Archiving vs. Unregistering
Short answer
C Archive the model version to prevent deployment while keeping it registered for compliance and rollback.
Question
An older model version must no longer be deployable, but it must remain available for compliance review and potential rollback.
You need to change the state of the model version to meet the requirements.
What should you do?
Key Takeaways
The main trap is confusing 'archive' with 'delete/unregister'; archiving keeps the model registered but non-deployable, whereas deleting removes it entirely.
Learn how to manage Azure ML model versions by archiving them to prevent deployment while retaining them for compliance and rollback purposes.
Common mistake: Choosing 'Delete' or 'Unregister' because they sound more definitive than 'Archive', not realizing that these actions permanently remove the model from the workspace, violating the compliance requirement.
Expert Analysis
The core concept tested here is the lifecycle management of Azure Machine Learning model versions, specifically distinguishing between retention states and deletion actions. The scenario requires a state change that prevents deployment while preserving data for compliance and rollback.
Archiving the model version (C) is the correct action. In Azure ML, archiving a model marks it as non-deployable in the portal or via SDK, but it remains registered in the workspace registry. This ensures the artifacts are retained indefinitely for audit trails, compliance reviews, or potential future rollbacks, exactly matching the requirements.
Why other options fail:
- Deleting (B) or Unregistering (D) removes the model version from the workspace entirely. Once unregistered or deleted, the model artifacts are gone, making it impossible to use for compliance review or rollback. These actions are destructive and irreversible without external backups.
- Archiving the training dataset (A) is unrelated to the model's deployability status. Archiving a dataset does not affect whether a specific model version can be deployed; it only manages dataset access.
The key boundary is between archiving (soft delete/retention) and unregistering/deleting (hard removal). Archiving is the precise mechanism for retaining historical models without allowing new deployments.
Official Reference
https://learn.microsoft.com/en-us/azure/machine-learning/how-to-manage-models
This link opens the official vendor documentation in a new tab.
Exam Strategy
When a question mentions 'compliance,' 'audit,' or 'rollback,' avoid any option that implies permanent removal like 'Delete' or 'Unregister.' Look for soft-state changes such as 'Archive' or 'Disable' which preserve data integrity while restricting access or functionality.
Frequently Asked Questions
What is the difference between archiving and unregistering a model in Azure ML?
Can I re-deploy an archived model version in Azure ML?
Community Votes
100% of anonymous learners picked answer C. Votes are pick records left by other test-takers — they are not the verified answer.
Discussion 2 contributions
Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.
From public community threads 2
-
RICHARDALEX007 👍 1 Answered CMaking a version non-deployable but retained for compliance/rollback is exactly archiving the model version (reversible, keeps it registered). Deleting (B) or unregistering (D) removes it; archiving the dataset (A) is unrelated.
-
exam_upskill 👍 1 Answered CC. Archive the model version
Reader comments & corrections
No reader corrections yet — spotted an error or have a note? Share it below.