An Apache Spark-based analytics platform optimized for Azure.
The current documentation answers part of the question directly, but it is important to separate what Databricks has officially announced from what has not been announced.
Starting September 26, new workspaces will be provisioned without access to certain legacy features, including the legacy Hive Metastore. However, Databricks explicitly states that existing workspaces and their workflows are not impacted by this change. Therefore, this date should not be interpreted as an end-of-support or shutdown date for Hive Metastore in existing production workspaces. (Azure Databricks documentation)
With respect to the specific questions:
- How long can existing production workspaces continue using Hive Metastore?
Based on the currently published documentation, Databricks has not announced an end-of-support date, retirement date, or removal date for Hive Metastore in existing workspaces. The September 30, 2026 milestone applies to the provisioning of new workspaces, not the operation of existing ones.
- Will existing workspaces be automatically converted to Unity Catalog?
I could not find any documentation indicating automatic migration. Databricks recommends that customers migrate to Unity Catalog, but the migration process remains customer-managed and requires planning, validation, and execution by workspace administrators.
- Can Hive Metastore still be used after enabling Unity Catalog?
Yes. Databricks documentation describes Unity Catalog as an additive governance layer. In Unity Catalog-enabled workspaces, the legacy Hive Metastore can still be accessed through the hive_metastore catalog. Databricks also documents migration approaches that allow both environments to coexist during transition periods. (Work with Hive Metastore alongside Unity Catalog)
- Is there a published date for complete removal of
hive_metastoreaccess?
I could not find any official Databricks announcement that specifies a date when direct access to hive_metastore will be removed from existing workspaces. Current documentation continues to describe coexistence and migration scenarios, while recommending eventual migration away from the legacy model.
Migration and rollback considerations
One practical point for production migrations is that Databricks currently supports a phased transition model. Unity Catalog can be introduced while existing Hive Metastore assets remain available, allowing teams to validate governance, permissions, workloads, and operational processes before removing legacy dependencies.
Databricks also documents Hive Metastore Federation as a migration option for organisations that want a gradual transition path instead of a large-scale cutover. The key objective should be to use Hive Metastore as a temporary compatibility mechanism during migration rather than as a long-term governance strategy.
To sum-up, the published September 30, 2026 date is a new-workspace provisioning change, not a documented retirement date for Hive Metastore in existing workspaces. As of the current documentation, there is no announced end-of-support date, no automatic migration requirement, and no published date for complete removal of hive_metastore access in existing workspaces.
Help make this community better for everyone: if this answer resolved your issue, please accept it or leave an upvote. If not, share more details in a comment so we can continue the discussion and find the right solution.