Skip to main content
Version control shows the live deployment alongside published and historical blueprint versions. Use it to understand a change before deployment or restore a known version as a new published version.

​
Open version control

From a published blueprint’s detail page, open More actions and select Version Control. You can also select a linked deployment status from the Catalog or open a blueprint from Lakehouse > Deployments. The version rail marks important states:
  • Live is the version currently deployed.
  • Staging is being prepared.
  • Staged is ready to deploy.
  • Historical versions are available for comparison when the entity has published history.

​
Compare versions

Select a version in the rail. DataX compares it with Current deployment and summarizes:
  • changed columns;
  • tag changes;
  • description changes;
  • SQL changes.
Use the Overview, Columns, SQL, and Sample rows tabs for detail. A Breaking change indicator means the comparison found a potentially incompatible change. Review the actual diff and downstream lineage; the indicator cannot know every consumer’s expectations.

​
Choose an action

The available button depends on the selected version and deployment state.
Restoring does not move the version pointer backward. It creates a new version from the historical definition, preserving the audit trail.
Only one target can be staged at a time. If another target is staged, deploy or discard it before you choose a different version.

​
Safe deployment checklist

  1. Confirm the selected version and the Live version.
  2. Review column additions, removals, type changes, and constraints.
  3. Compare the SQL and sample rows.
  4. Check entity lineage for downstream consumers.
  5. Confirm that you have the required blueprint and deployment access.
  6. Run the deployment action and wait for the progress state to settle.
If the page stays in In Progress, refresh the deployment list before taking another action. If a staged target is unexpected, do not replace it until you identify who created it and why. See Lakehouse deployments for list-level filtering and status guidance.
Last modified on September 16, 2026