Preset Partner Connect Schema Changes

[Dave_Nithio]

When using partner connect to connect Serverless Databricks to my BI tool Preset, you must manually define the schema that Preset has access to. In my case, I individually selected all databases currently in my hive_metastore:

The problem is, once created, you don't have the option to add additional schema to this setup. If I create a new database that I want to connect, it looks like I have to delete the partner connect, re-add partner connect, then re-select all schema available. Partner connect does create a service principal called PRESET_USER that has a personal access token which links Preset to Databricks. In user management we can generate secrets and add permissions, but not explicitly add a new schema to its access. Looking at the serverless warehouse, there are also no explicit settings related to schema allowance. I also looked at the JDBC connection, including its connection arguments, but that does not explicitly define schema either.

The only other thought I had to update available schema is in Workspace Settings for SQL warehouses and serverless compute. There is an option for Data Security that allows us to enter key-value pairs for data access. These seem to be spark settings though and not schema settings.

Does anyone know if there is a way to update schema for this sql serverless warehouse connection?

[mark_ott]

No, there is currently no simple, direct way to add new schema access to an existing Serverless Databricks SQL warehouse connection through Partner Connect for Preset—neither through Databricks UI, BI tool configuration, nor the Databricks service principal’s direct permissions .

What’s Happening Behind the Scenes

• When you use Partner Connect, you select which schemas (databases) Preset’s service principal (PRESET_USER) can access at creation time.

• This access is tied to that specific principal and warehouse, and Databricks does not expose an interface to modify allowed schemas after initial setup.

• The only way to update schema access is to delete the Partner Connect integration and recreate it, carefully selecting the full list of desired schemas—including any new ones.

Why Schema Modification Isn’t Possible

• SQL warehouses serverless settings, including Data Security KV pairs, relate to Spark settings and general security, not fine-grained schema accessibility.

• The service principal’s personal access token does not natively support updating schema access lists post-creation; its permissions are governed by the original setup and Databricks’ internal policy model.

• JDBC connection arguments do not target schema-level access—these settings handle security, performance, and connection properties, not database discovery or inclusion.

Alternatives and Workarounds

• The recommended method is to delete and recreate the Partner Connect configuration, making sure to include all needed schemas.

• If you anticipate needing more dynamic schema additions, consider creating a service principal manually and managing its roles/permissions directly via Databricks ACLs and workspace permissions, but keep in mind that Databricks BI integrations still don’t generally allow dynamic runtime schema expansion.

• As of November 2025, there’s no supported endpoint, CLI, or undocumented hack to add schemas to a Partner Connect warehouse–Preset connection after initial creation.

---

In summary:
You must delete and re-create the Partner Connect integration to add newly created databases (schema) for Preset access via Databricks SQL serverless warehouses. No UI, warehouse, or connection string setting exists to update schema inclusion after setup—this is a known limitation due to how Databricks manages partner BI access and service principal security.

---

Modification Method	Able to Add Schema After Creation?
Partner Connect Setup	No (must re-create)
Service Principal Permissions	No (cannot add schema directly)
Workspace SQL Warehouse Settings	No
JDBC Connection String Arguments	No