Skip to main content

March 2025

These features and Databricks platform improvements were released in March 2025.

note

Releases are staged. Your Databricks account might not be updated until a week or more after the initial release date.

Databricks Runtime 16.3 is GA​

March 31, 2025

Databricks Runtime 16.3 and Databricks Runtime 16.3 ML are now generally available.

See Databricks Runtime 16.3 and Databricks Runtime 16.3 for Machine Learning.

Disable your legacy Hive metastore (Public Preview)​

March 31, 2025

You can now disable direct access to Hive metastores from your workspace, ensuring that all access and governance is provided by Unity Catalog. See Disable access to the Hive metastore used by your Databricks workspace.

Databricks jobs now allow 1,000 tasks​

March 31, 2025

A single job in your workflow can now include up to 1,000 tasks. See Orchestration using Databricks Jobs.

Mosaic AI Agent Framework and Mosaic AI Agent Evaluation are generally available (GA)​

March 31, 2025

Build and evaluate AI agents using Mosaic AI Agent Framework and Mosaic AI Agent Evaluation. Agent Framework simplifies the creation and deployment of high-quality generative AI applications. Agent Evaluation provides tools to assess the quality, cost, and latency of AI applications, helping you develop accurate, safe, and governed outputs. See Introduction to building gen AI apps on Databricks and What is Mosaic AI Agent Evaluation?.

Managed SaaS connectors support column-level selection (API-only)​

March 28, 2025

Managed connectors for enterprise applications (such as Salesforce, Workday, and ServiceNow) now support column-level selection and deselection. You can also specify whether to automatically ingest future columns as they're added to the source. See Managed connectors in Lakeflow Connect.

Lakehouse Monitoring is GA​

March 26, 2025

Databricks Lakehouse Monitoring is now generally available.

Legacy system schemas deprecated​

March 26, 2025

The following legacy system schemas are deprecated and will be removed from system catalog. For a list of current system schemas and tables, see Monitor account activity with system tables.

  • operational_data
  • lineage
  • access_preview
  • billing_preview

Claude Sonnet 3.7 now supported on Mosaic AI Model Serving​

March 26, 2025

Anthropic's Claude Sonnet 3.7 is now supported on Mosaic AI Model Serving. Claude Sonnet 3.7 is available using:

Group clusters are generally available (GA)​

March 25, 2025

The ability to assign all-purpose and jobs compute resources to a group using Dedicated access mode is now generally available. See Assign compute resources to a group.

New CAN VIEW permission available for SQL warehouses (Public Preview)​

March 25, 2025

The CAN VIEW permission allows users to monitor SQL warehouses, including the associated query history and query profiles. These users cannot run queries on the SQL warehouse without being granted additional permissions. See SQL warehouse ACLs.

New column explorer​

March 24, 2025

You can now use the column explorer when working with tabular notebook outputs. The column explorer allows you to search, show or hide, pin, and rearrange columns. See Explore columns.

AI Gateway now supports pay-per-token endpoints​

March 24, 2025

Mosaic AI Gateway now supports Foundation Model APIs pay-per-token endpoints.

You can now enable the following governance and monitoring features on your pay-per-token endpoints:

  • Permission and rate limiting to control who has access and how much access.
  • Payload logging to monitor and audit data being sent to model APIs using inference tables.
  • Usage tracking to monitor operational usage on endpoints and associated costs using system tables.
  • AI Guardrails to prevent unwanted and unsafe data in requests and responses.

Databricks Connector for Google Sheets (Public Preview)​

March 20, 2025

The Databricks Connector for Google Sheets allows you to connect to your Databricks workspace from Google Sheets. See Connect to Databricks from Google Sheets.

Web terminal activity is now considered compute activity​

March 20, 2025

Compute resources configured to automatically terminate after a period of inactivity will continue to run if activity is detected in a Databricks web terminal connected to that compute, including through a notebook. See Automatic termination.

Budget policies renamed to serverless budget policies​

March 19, 2025

Serverless budget policies (Public Preview) have been renamed from budget policies. See Attribute usage with serverless budget policies.

Delta Sharing supports renaming tables​

March 19, 2025

Delta Sharing now supports renaming shared tables. If a provider renames a shared table, the table on the recipient side automatically reconciles without erroring. See Update shares.

OIDC federation for Databricks-to-open Delta Sharing (Public Preview)​

March 17, 2025

Data providers in Databricks can now enable recipients to use their own identity provider (IdP) to establish access to Delta Sharing shares created in Databricks. This authentication flow uses Open ID Connect (OIDC) token federation to exchange JSON web tokens (JWT) generated by the recipient's IdP for short-lived OAuth tokens generated by Databricks. This Databricks-to-open sharing authentication flow is for recipients who do not have access to a Unity Catalog-enabled Databricks workspace.

OIDC federation is an alternative to using long-lived Databricks-issued bearer tokens to connect non-Databricks recipients to providers. It enables fine-grained access control, supports MFA, and reduces security risks by eliminating the need for recipients to manage and secure shared credentials.

Databricks-to-open sharing with OIDC federation supports user-to-machine (U2M) authentication for recipients on Power BI and Tableau, and machine-to-machine (M2M) OAuth Client Credentials authentication from Python client apps registered in a recipient's IdP.

This feature is in Public Preview. Contact your Databricks representative to request access.

See Use Open ID Connect (OIDC) federation to enable authentication to Delta Sharing shares (open sharing).

AutoML supports clusters in dedicated access mode assigned to groups​

March 17, 2025

AutoML can now be used on clusters in dedicated access when assigned to groups. Previously, AutoML was only available on dedicated access clusters in single-user mode. See Requirements.

The compute system tables are enabled by default​

March 17, 2025

The system.compute schema is now enabled by default in all Unity Catalog workspaces. See Compute system tables reference.

Hive metastore federation is generally available​

March 14, 2025

Hive metastore federation enables you to use Unity Catalog to access and govern data that is registered in a Hive metastore. This includes both externally-managed Hive metastores and legacy internal Databricks Hive Hive metastores.

GA adds support for:

  • Hive metastore 3.1 (in addition to versions 0.13 and 2.3, which were supported in the Public Preview).
  • Postgres and SQL Server (in addition to MySQL, which was supported in the Public Preview).
  • Applying row filters and column-level masks to Hive tables that are accessed using Hive metastore federation.

In addition, this release includes preview support for:

  • Reading Delta shallow clones that are defined in the Hive metastore (Public Preview).
  • Reading DBFS-mount-backed tables across workspaces (Private Preview: contact your Databricks account representative to sign up).

See Hive metastore federation: enable Unity Catalog to govern tables registered in a Hive metastore.

Compute logs can now be delivered to volumes (Public Preview)​

March 13, 2025

You can now deliver a compute resource's driver, worker, and event logs to a Unity Catalog volume. See Compute log delivery.

Download as Excel is now supported in notebooks connected to serverless compute​

March 13, 2025

For notebooks connected to serverless compute, you can now download tabular cell results as an Excel file. Download as Excel is also supported for notebooks connected to SQL warehouses. See Download results.

Bind storage credentials, service credentials, and external locations to specific workspaces (GA)​

March 12, 2025

The ability to bind storage credentials, service credentials, and external locations to specific workspaces is now generally available. Workspace binding prevents access to those objects from other workspaces. This feature is especially helpful if you use workspaces to isolate user data access, for example, if you have separate production and development workspaces or a dedicated workspace for handling sensitive data.

For more information, see (Optional) Assign an external location to specific workspaces, (Optional) Assign a storage credential to specific workspaces and (Optional) Assign a service credential to specific workspaces.

Databricks Runtime 16.3 (Beta)​

March 12, 2025

Databricks Runtime 16.3 and Databricks Runtime 16.3 ML are now available as Beta releases.

See Databricks Runtime 16.3 and Databricks Runtime 16.3 for Machine Learning.

Databricks JDBC Driver (OSS) is GA​

March 11, 2025

The Databricks JDBC Driver (OSS) is now generally available (GA) and will soon be open source. Download the Databricks JDBC Driver (OSS) from the Maven Repository.

Monitor and revoke personal access tokens in your account (Public Preview)​

March 11, 2025

Token report is now in Public Preview. Token report enables account admins to monitor and revoke personal access tokens (PATs) in the account console. Databricks recommends you use OAuth access tokens instead of PATs for greater security and convenience. See Monitor and revoke personal access tokens in the account.

Genie Conversation API is now in Public Preview​

March 11, 2025

The Genie Conversations API is now in Public Preview. You can now ask questions, retrieve generated SQL and query results, or get the details associated with a Genie space using the Databricks REST API. To learn how to use the API, see the Genie REST API reference.

Perform batch inference using AI Functions (Public Preview)​

March 10, 2025

Mosaic AI now supports batch inference using AI Functions. This functionality is in Public Preview. See the following guides for more details:

ServiceNow connector (Public Preview)​

March 10, 2025

You can now ingest from ServiceNow using Databricks Lakeflow Connect. See Configure ServiceNow for Databricks ingestion. To participate in the preview, contact your Databricks account team.

External models now supports OpenAI-compatible LLMs​

March 10, 2025

External models in Mosaic AI Model Serving now supports a custom provider option. You can use this option to specify alternate providers that are not explicitly listed as a model provider or models behind custom proxies that are OpenAI API compatible. See Model providers.

Delta Sharing support for query optimizer statistics is GA​

March 4, 2025

Data recipients accessing Delta tables shared via Delta Sharing can now benefit from optimizer statistics to improve query performance. The data provider must either enable predictive optimization or manually run the ANALYZE command on the shared Delta table.

Performance improvements include the following:

  • Same-region Databricks-to-Databricks queries are 11% faster.

  • Cross-region Databricks-to-Databricks queries are 28% faster.

See Predictive optimization for Unity Catalog managed tables and ANALYZE TABLE.

AI Gateway now supports fallbacks for external models​

March 4, 2025

You can now enable fallbacks on your model serving endpoints using AI Gateway. If your request to an external model fails, you can specify that the request be redirected to other external models served on that endpoint. See Configure AI Gateway using the UI.

Control the location of Databricks Assistant with the new docking feature​

March 4, 2025

You now have more control over the placement of the Databricks Assistant pane and the ability to minimize it with the new docking feature.

See Control the location of Databricks Assistant.

Last updated on