Lakewatch prerequisites
Lakewatch is in Private Preview. To request access, reach out to your Databricks account team.
This page walks you through the features you must enable in the workspace you plan to use with Lakewatch.
Unity Catalog
Unity Catalog must be enabled in your Lakewatch workspace. All of the tables that Lakewatch creates go into the Unity Catalog that you configure during the installation process, and this catalog cannot be using the legacy hive metastore.
Serverless SQL warehouse
When interacting with the Lakewatch user interface, a SQL warehouse is used to provide the data seen in the UI.
Some of the reporting features of Lakewatch also use the Serverless SQL warehouse for background tasks. The installation wizard automatically creates a small SQL warehouse for use with Lakewatch so that you have separate cost attribution and control.
Serverless job compute
The jobs that Lakewatch creates use serverless job compute for lower latency and more predictable performance as your data scales up. Classic compute isn't supported.
Accounts created after December 16, 2024 (AWS and Azure) or February 3, 2026 (GCP) have serverless compute auto-enabled. Older accounts require an account admin to manually enable it in the Accounts Console.
To enable serverless compute at the account level:
- Log in to the Accounts Console as an account admin.
- Go to Settings > Feature enablement.
- Find Serverless compute and enable it.
- If prompted (AWS and GCP only), accept the serverless terms of service.
- Wait 1–2 minutes for the setting to propagate, then refresh the Lakewatch onboarding page.
The Serverless compute setting in the workspace-level Feature Previews page is unrelated. That setting controls access management and rate limits, not serverless eligibility. The account-level setting in the Accounts Console is the one required for Lakewatch.