All Products
Search

This Product

    DataWorks:Manage Open Data

    Document Center

    DataWorks:Manage Open Data

    Last Updated:Mar 27, 2026

    DataWorks Open Data is a centralized metadata collection for the DataWorks platform. It grants your MaxCompute compute engine access to standardized, traceable metadata—covering tables, task nodes and instances, workspaces, members, projects, Data Quality, and data assets—without requiring complex configuration.

    Workflow at a glance: View packages → Install and authorize → Query metadata → Uninstall when done

    Use cases

    Open Data upgrades the previous invitation-only Open Data into a publicly available, visual-interface experience. Instead of using the command line, you manage and query metadata directly in the DataWorks console.

    Use Open Data for the following scenarios:

    • Data inventory: Get a complete picture of data objects—tables, tasks, workspaces, owners, update frequencies, and dependencies—so you can eliminate data black boxes across your organization.

    • End-to-end lineage tracking: Run custom SQL queries on upstream and downstream lineage paths of specific tables, tracing the full data flow from source to downstream applications to quickly locate issues or assess impact.

    • Custom metadata analysis: Query and analyze metadata using SQL directly, in addition to the existing OpenAPI, to shorten the data governance lifecycle.

    Prerequisites

    Before you begin, ensure that you have:

    Limitations

    • Edition: Only DataWorks Enterprise Edition and later are supported.

    • Permissions: Only tenant owners, tenant administrators, data governance administrators, or Resource Access Management (RAM) users with the DataWorksFullAccess permission can install and uninstall Open Data.

    • Authorization scope: Metadata view authorization is only supported through MaxCompute. Other resource types are not supported.

    • Data update latency: Metadata is updated with T+1 latency—you can query metadata from the previous day. For real-time metadata, use the DataWorks OpenAPI instead.

    View Open Data

    Before installing a package, explore the available packages to understand their metadata content and target scenarios.

    1. Log on to the DataWorks console. In the top navigation bar, select the desired region. In the left-side navigation pane, choose Data Governance > Data Map. On the page that appears, click Go to Data Map.

    2. In the navigation pane on the left of the Data Map page, click the image icon to go to the Data Catalog page.

    3. In the catalog list, click DataWorks OpenData.

    4. On the Package List tab, click Details for a package. On the Table List tab, review the metadata tables and their descriptions. The two available packages serve different purposes:

      The Examples package is provided by DataWorks.
      Important

      For table schemas and field details, see Details of Open Data table schemas.

      Package name Open Data Scenarios
      MetaData Table information (table schema, field descriptions, data lineage); scheduling nodes and instances (task execution status, dependencies); management metadata (workspaces, members, projects); Data Quality and data governance details. Data asset inventory, data lineage analysis, and dependency management.
      Examples A collection of sample metric tables generated from MetaData. The actual data varies; the data displayed in the UI prevails. Combined with MetaData raw data to quickly build common analysis scenarios, such as resource utilization analysis and task health assessment.

    5. To inspect a specific metadata table, click Details for that table. On the table details page, review the Field Name and Description columns to understand the relationships between Open Data metadata tables.

    Install Open Data

    Installing a package grants your chosen MaxCompute compute engine access to all data views in that package.

    1. Go to the Open Data page. On the Package List tab, select the target package (MetaData or Examples) and click Details in the Actions column.

    2. On the package page, click Installation:

      • First-time installation: Click Installation in the upper-right corner of the package page.

      • If an installation record exists: On the Installation History tab, click Installation.

      image

    3. In the Install DataWorks Metadata dialog box, select the destination workspace and the MaxCompute project to grant permissions to.

      Note

      • The MaxCompute project you select is the compute engine bound to the destination workspace.

      • If the compute engine is already in the Installed state, skip this step.

      • For workspaces in standard mode, install and grant permissions for the MaxCompute compute engines bound to both the development and production environments.

    4. Read the installation notes, select the confirmation checkbox, and click Confirm Installation. After the installation succeeds, the new record appears on the Installation History tab.

    To use metadata from additional packages in the same compute engine, repeat the steps above for each package.

    Use Open Data

    After installing Open Data, access the authorized metadata views in Data Studio (for task development) or DataAnalysis (for ad-hoc SQL queries).

    Tool Best for
    Data Studio Writing and scheduling MaxCompute tasks that query metadata programmatically
    DataAnalysis Interactive, ad-hoc SQL queries against metadata without setting up a task node

    1. In the navigation pane on the left of the Data Map page, click the image icon to go to the Data Catalog page.

    2. In the catalog list, click MaxCompute, and then click the package prefixed with u_meta and suffixed with the region name.

      image

    3. On the details page, choose Use Now > Data Studio or Use Now > Data Analysis based on your workflow.

    Use in Data Studio

    1. In the top menu bar of Data Studio, switch to the region and workspace where Open Data is installed.

    2. Create a MaxCompute node. On the node editing page, write your query against the authorized metadata views.

    3. To verify that the installation is working, run the following sample query:

      Important

      • Use the authorized MaxCompute compute engine when running the query.

      • Replace u_meta_hangzhou with the actual package name from your MaxCompute data catalog. The package name is prefixed with u_meta and suffixed with the region name.

      SELECT  dt
        ,COUNT(*) AS database_count
FROM    u_meta_hangzhou.databases
GROUP BY dt
ORDER BY dt ASC
LIMIT   32
;

    Use in DataAnalysis

    1. Go to the SQL query page of DataAnalysis. The system provides sample metadata analysis scripts that you can modify.

    2. Click the image icon in the upper-right corner of the SQL query page. Select the authorized workspace and bind the data source that has the same name as the one generated when you bound the MaxCompute compute engine.

    3. Before running a sample script, replace REPLACE_WITH_WORKSPACE_ID with the ID of the workspace you want to query. For how to get a workspace ID, see Configure a workspace.

    Uninstall Open Data

    Uninstalling a package revokes the MaxCompute compute engine's access to the authorized metadata views and breaks any tasks that depend on this data. Verify that no active dependencies exist before proceeding.

    1. In the navigation pane on the left of the Data Map page, click the image icon to go to the data catalog page.

    2. In the catalog list, select MaxCompute and then click the package prefixed with u_meta and suffixed with the region name.

    3. On the Accessible Projects tab of the details page, find the destination project and click Uninstall in the Actions column. In the Confirm Uninstall DataWorks Metadata dialog box, read the uninstallation notes, select the confirmation checkbox, and click Confirm Uninstall.

    FAQ

    Why does metadata have a one-day delay?

    Open Data metadata is generated with T+1 latency, meaning it reflects activity from the previous day. This is a platform-level batch update constraint. If you need real-time metadata, use the DataWorks OpenAPI instead.

    Can I uninstall Open Data after installation?

    Yes. Uninstalling revokes the permissions granted to the MaxCompute compute engine and will break any tasks that query those metadata views. Before uninstalling, confirm that no active tasks or scheduled jobs depend on the views.

    How do I secure metadata?

    Use MaxCompute data access controls to restrict who can access sensitive metadata.