TVL Managed Superset
Get started

Connect Databricks to Apache Superset 2026

Tutorial to connect Databricks SQL Warehouse to Apache Superset: driver, token, optimization, Unity Catalog.

· Tuvalu Tech OÜ

Databricks has become in 2026 the reference lakehouse platform, particularly among data-mature organizations. Connecting Apache Superset to a Databricks SQL Warehouse is now straightforward with the official driver. This guide details the procedure and optimizations.

1. Prerequisites

  • An accessible Superset instance;
  • A Databricks workspace with an active SQL Warehouse;
  • A Databricks personal access token (or service principal);
  • The databricks-sql-connector driver installed.

If you want to get started quickly, TVL Managed Superset includes Databricks drivers by default.

2. Retrieve the hostname and HTTP path

In the Databricks UI:

  1. SQL → SQL Warehouses → select your warehouse;
  2. Connection Details tab;
  3. Copy Server hostname, HTTP path, Port (443).

3. Build the URI

databricks+connector://token:<personal_access_token>@<hostname>:443?http_path=<http_path>&catalog=<catalog>&schema=<schema>

Example:

databricks+connector://token:dapi***@adb-1234567890.10.azuredatabricks.net:443?http_path=/sql/1.0/warehouses/abc123&catalog=main&schema=analytics

4. Add to Superset

  1. UI → Settings → Database Connections → + Database;
  2. Type: Databricks;
  3. Paste the URI;
  4. Test → Save.

5. Unity Catalog

With Unity Catalog (centralized Databricks governance), specify catalog in the URI or let Superset choose during dataset creation. Allows isolating schemas per environment (dev / prod).

This configuration is applied by default on TVL Managed Superset, which follows community best practices.

6. Optimization

  • Serverless SQL Warehouse: startup in seconds vs minutes;
  • Auto-stop enabled (5-10 min) to save costs;
  • Photon Engine enabled for 2-5x performance;
  • Aggressive Superset cache (24h) on stable dashboards;
  • Materialized views Databricks for frequent aggregations.

7. Security

  • Service principal instead of PAT for production;
  • IP allowlist Databricks;
  • Unity Catalog grants at table/column level;
  • Audit logs enabled.

8. Common pitfalls

  • Expired token: PATs have a lifespan, to rotate;
  • Incorrect HTTP path: very common, verify it's properly copy-pasted;
  • Paused warehouse: first query takes 30-60s to start;
  • Sub-query latency: Databricks prefers simple queries on very large volumes.

9. Conclusion

Databricks + Apache Superset offers a powerful combination for organizations that have already invested in the lakehouse. The Serverless SQL Warehouse makes the experience close to a classic cloud data warehouse, without the overhead.

Want the benefits of Apache Superset without the friction of installation and maintenance? Deploy your instance in 3 clicks with TVL Managed Superset, hosted in Europe (OVHcloud, Roubaix, France).

For more: connect Snowflake, connect BigQuery, connect ClickHouse.