Databricks
Databricks SQL lets you query lakehouse tables through a SQL warehouse. Fresco BI connects to a SQL warehouse with a personal access token, lists Unity Catalog catalogs, schemas, tables and views, and imports the tables you pick into a Data Table.
| Feature | Supported |
|---|---|
| Import | Yes |
| DirectQuery | No |
| SSL/TLS | Always on (verified) |
| SSH tunnel | No |
| Cloud | Yes |
| On-premises (via Gateway) | No |
Prerequisites
Section titled “Prerequisites”- A Databricks workspace on AWS (
*.cloud.databricks.com), Azure (*.azuredatabricks.net) or Google Cloud (*.gcp.databricks.com). - A SQL warehouse. Fresco BI connects only to SQL warehouses, not to all-purpose clusters.
- A workspace admin who can add a user or service principal and grant Unity Catalog privileges.
Step 1: Create a read-only identity
Section titled “Step 1: Create a read-only identity”Use a dedicated service principal (recommended) or user, here called fresco-reader.
- Add
fresco-readerto the workspace with the Databricks SQL access entitlement. - On the SQL warehouse, open Permissions and give
fresco-readerCan use. - Grant read-only Unity Catalog privileges in the SQL editor:
GRANT USE CATALOG ON CATALOG <catalog> TO `fresco-reader`;GRANT USE SCHEMA ON CATALOG <catalog> TO `fresco-reader`;GRANT SELECT ON CATALOG <catalog> TO `fresco-reader`;Privileges granted on a catalog also cover schemas and tables created in it later. To limit access to one schema instead:
GRANT USE CATALOG ON CATALOG <catalog> TO `fresco-reader`;GRANT USE SCHEMA ON SCHEMA <catalog>.<schema> TO `fresco-reader`;GRANT SELECT ON SCHEMA <catalog>.<schema> TO `fresco-reader`;- Create a personal access token for
fresco-readerand copy it. Give it an expiry date and put a reminder to rotate it.
Never use an admin’s token. Fresco BI only reads, so the identity needs nothing beyond the grants above.
Step 2: Allow network access
Section titled “Step 2: Allow network access”Fresco BI connects to the workspace host on port 443 over the internet. If the workspace uses IP access lists, add Fresco BI’s outbound IPs to the allow list: Fresco BI’s outbound IP addresses (the setup guide beside the data source form lists them). Databricks has no On-premises (via Gateway) or SSH tunnel option.
Step 3: Fill in the form
Section titled “Step 3: Fill in the form”Copy Server hostname and HTTP path from the SQL warehouse’s Connection details tab.
| Field | What to enter |
|---|---|
| Server hostname | Workspace host only, for example dbc-a1b2c3d4-e5f6.cloud.databricks.com. No https://, port or path |
| HTTP path | /sql/1.0/warehouses/<warehouse-id> |
| Catalog | One catalog to list, or empty to list every catalog the identity can see |
| Schema | One schema to list, or empty to list every schema in the catalog(s) |
| Access token | The token from Step 1 |
Click Test connection: it runs SELECT current_version() on the warehouse. A stopped warehouse is started first, which can take a few minutes. Then click Save.
Fresco BI reads results directly from the workspace host (Cloud Fetch is off), so no cloud storage access is needed.
SSL modes
Section titled “SSL modes”Databricks connections are always encrypted and the server certificate is always verified against the system trust store. There is no SSL mode to choose.
Troubleshooting
Section titled “Troubleshooting”- The server rejected the user name or password. The access token is wrong, expired or revoked. Create a new token and paste it.
- The database user cannot read this object. The workspace IP access list blocks Fresco BI (add Fresco BI’s outbound IP addresses (the setup guide beside the data source form lists them)), the identity lacks Can use on the SQL warehouse, or it lacks USE CATALOG, USE SCHEMA or SELECT on that object. Check Steps 1 and 2.
- The connection settings are not valid: check the marked fields. The HTTP path does not point to an existing SQL warehouse, or the server hostname is not a Databricks workspace host. Copy both again from Connection details.
- The database or object was not found on the server. The catalog, schema or table was renamed or dropped, or the identity cannot see it.
- The host name could not be resolved (DNS). Check the Host field. Check the workspace host spelling.
- No connection could be opened to this host and port. Check the host, the port, and that the firewall allows Fresco BI’s IP. The workspace could not be reached, or the connection dropped while reading. Check the host and try again.
- The server did not answer in time. Usually a firewall that drops packets, or a wrong host or port. The warehouse took more than 5 minutes to start, or a statement ran too long. Start the warehouse first or use a larger size.
- The SSL connection could not be set up with the chosen mode. Check the SSL mode and the CA certificate. A proxy is intercepting TLS to the workspace. Let Fresco BI reach Databricks directly.
- The source returned a data type this connector does not support. BINARY, INTERVAL, GEOGRAPHY, GEOMETRY and VOID columns cannot be imported; leave them out of the Data Table. STRUCT, ARRAY, MAP and VARIANT columns are imported as JSON text.

