Share and Consume Data with Data Share in Autonomous AI Database

Use Data Share in Data Studio to share Oracle data and metadata from Autonomous AI Database with other databases and non-database tools. Data Share supports provider and consumer workflows: a provider creates a share, adds objects and recipients, and publishes the share; a consumer subscribes to the provider and creates database objects that make the shared data available for querying.

This topic provides an entry point for understanding when and how to use Data Share with Autonomous AI Database. For detailed overview of the Data Share page and the available sharing types, see the Data Studio Documentation.

When to use Data Share

Use Data Share when you want a governed way to make data available to other users, databases, applications, or tools without building a custom export-and-delivery process.

For example, use it to:

• Consume a share when another provider has made data available to your Autonomous AI Database.

• Share versioned data when recipients should see a published snapshot and receive changes only when the provider publishes a new version.

• Share live data when Autonomous AI Database recipients need access to the latest committed source data through a direct database connection.

• Use open Delta Sharing-based access for tools and APIs that support Delta Sharing, or use Live Share for near-real-time access between supported Autonomous AI Database instances.

• Automate sharing workflows with the DBMS_SHARE PL/SQL package when you want to implement provider or consumer process using scripts instead of the Data Studio UI.

How Data Share Works with Autonomous AI Database

Data Share runs from Database Actions for Autonomous AI Database. Database Actions is bundled with each Autonomous AI Database instance and provides development, data, administration, monitoring, and download features.

For non-administrator users, Database Actions hides tool cards when the user does not have the required permissions. Data Studio tools require the DWROLE database role. For more information, see Connect to Autonomous AI Database on Dedicated Exadata Infrastructure with Database Actions

Data Share on Autonomous Database on Dedicated Exadata Infrastructure supports the Consumer role, which:

• Subscribes to a share provider.
• Registers available shares.
• Creates external tables or views for the shared objects.
• Queries the shared data from Autonomous AI Database.

A Versioned Share publishes data as a known snapshot. The provider publishes the share to an object storage location, and recipients use a profile to register and consume the shared data. Recipients see changes when the provider publishes a new version. See Provide Versioned Share for more information.

A Live Share lets an Autonomous AI Database recipient access the latest data directly from the provider database. Live Share uses cloud links and is intended for Oracle AI Database consumption; for live shares in the same region, recipients see current data at query time. See Provide Live Share for more information.

Before You Begin

Before using Data Share, verify the following:

• The user can access Database Actions and has the required Data Studio role, such as DWROLE, unless using the ADMIN user.
• For Live Share providers, configure Share Provider Identity in Data Studio Settings so recipients can identify the provider.
• Configure SMTP in Data Studio Settings if you want Data Share to send recipient activation links or share notifications by email. See Data Studio Settings for more details.

Open Data Share

To open Data Share:

1. Open Database Actions for your Autonomous AI Database.
2. Select the Data Studio tab.
3. Select Data Share.
4. Select Consume Share to subscribe to shares made available by a provider.

What You Can Do in Data Share

• Consume Share: Subscribe to a share provider, register available shares, create external tables or views, and query the shared objects from SQL. See Consume Share for detailed consumer workflow for subscribing to versioned and live share providers and creating queryable objects from shared data.

  • Consume Versioned Share: Upload or paste a Delta Sharing JSON profile, subscribe to a provider, select shares, and link shared objects for use in Autonomous AI Database.

  • Consume Live Share: Subscribe to an available Live Share provider and create views of tables included in the live share.

• Manage shares with DBMS_SHARE: Use PL/SQL to discover shares, and monitor or troubleshoot sharing workflows. See Manage Shares Using DBMS_SHARE to automate PL/SQL workflow for provider and consumer sharing operations.

Data Share Limitations with Autonomous AI Database on Dedicated Exadata Infrastructure

• Oracle Database version 19c and later supports both Versioned Share and Live Share. Data Share also supports a defined set of column types, and Live Share providers can create shares with a maximum of four objects. If multiple source tables are joined together, Oracle recommends creating a view that performs the joins and sharing the view instead. See Data Share Limitations on Autonomous AI Database to review the list of supported column types.

• You can only consume a versioned share on Autonomous AI Database on Dedicated Exadata Infrastructure. Producing versioned shares is not supported.

• A live share data provider can create shares with a maximum of four objects.

• If the live share data provider is sharing multiple tables which would be joined together, it is recommended that the producer create a view that performs the joins and then only shares the view.

• For Live Share performance, increasing ECPUs alone may not improve complex query performance. The Troubleshooting Guide explains how an administrator can configure degree of parallelism through resource management settings.

• When Autonomous AI Database consumes Delta Sharing data, it does not support Delta tables that have Deletion Vectors enabled. If a shared Delta table uses Deletion Vectors, ask the provider to remove Deletion Vectors or provide a compatible share. See Shared Table Query Attempt Failure for guidance.