[BUG] SHOW CATALOGS returns spark_catalog instead of hive_metastore for Standard workspaces

[aliakseistruh-dr]

Describe the bug
When connecting to a Databricks Standard workspace (without Unity Catalog) using the JDBC driver, the SHOW CATALOGS command returns spark_catalog as the catalog name. However, the Databricks UI displays this catalog as hive_metastore, and other Databricks tools (Python SQL Connector, REST APIs) do not recognize spark_catalog, causing integration failures.

This inconsistency breaks downstream integrations:

• Databricks UI shows hive_metastore, not spark_catalog
• Databricks Python SQL Connector does not recognize spark_catalog
• Databricks REST APIs return 422 errors when spark_catalog is passed as the catalog parameter

To Reproduce
Steps to reproduce the behavior:

1. Connect to a Databricks Standard workspace (without Unity Catalog enabled) using the JDBC driver
2. Execute the SQL query: SHOW CATALOGS
3. Observe that the returned catalog name in the catalog column is spark_catalog
4. Try to use spark_catalog with other Databricks APIs or the Python SQL Connector - it will fail

Expected behavior
For Standard workspaces (non-Unity Catalog), SHOW CATALOGS should return hive_metastore as the catalog name to be consistent with:

• The Databricks UI
• The Databricks Python SQL Connector
• The Databricks REST APIs

Screenshots
N/A

Client side logs
When executing SHOW CATALOGS on a Standard workspace, the JDBC driver returns:
catalog: "spark_catalog"

We added debug logging in our application to confirm this:
DEBUG Databricks JDBC driver returned 'spark_catalog' - bug still exists, applying workaround to use 'hive_metastore'

Client Environment (please complete the following information):

• OS: macOS (darwin 24.6.0), also reproducible on Linux
• Java version: Java 17
• Java vendor: OpenJDK
• Driver Version: 3.0.7 (also reproducible in 2.6.x, 2.7.x, and all 3.0.x versions)
• BI Tool (if used): Custom application using Spring Boot + jOOQ (enterprise)
• BI Tool version (if applicable): N/A

Additional context
Workaround: We currently have a workaround in our application that replaces spark_catalog with hive_metastore.

Impact: This issue has been present since at least March 2024 and affects any application that:

• Builds data catalog browsers displaying catalog names to users
• Passes catalog names to other Databricks APIs (e.g., preview/sync endpoints)
• Integrates with the Databricks Python SQL Connector

Note: Premium workspaces with Unity Catalog enabled correctly return catalog names like main - this issue only affects Standard workspaces.