Create a StarRocks Data Source

Follow these steps to connect to StarRocks and create a StarRocks data source:

  1. Configure User Access Rights

  2. Create a StarRocks Data Source

Refer to Additional Information for additional performance tips and information about StarRocks feature support.

Configure User Access Rights

To ensure GoodData can successfully connect, discover metadata, and execute queries, you must grant permissions to the Catalog (data scope), Warehouse (compute scope), and Information Schema (metadata scope).

Run the following SQL commands as an ADMIN or ROOT user in your StarRocks cluster.

  1. Set Context (Optional)

    If you are provisioning access to an external catalog (e.g., Iceberg, Hudi) or a specific project catalog, switch to it first.

    -- Switch to the target catalog (default is 'default_catalog')
SET CATALOG {catalog_name};

  2. Create a Role

    It is a best practice to manage permissions via roles rather than direct user grants.

    CREATE ROLE {role_name};

  3. Grant Compute & Catalog Access

    GoodData requires “Usage” privileges to enter the catalog and utilize compute resources.

    -- Allow the role to access the catalog (Data Container)
GRANT USAGE ON CATALOG {catalog_name} TO ROLE {role_name};

-- Allow the role to use a compute warehouse (Compute Resources)
-- Note: 'default_warehouse' is standard, but change if using dedicated warehouses
GRANT USAGE ON WAREHOUSE {warehouse_name} TO ROLE {role_name};

  4. Grant Metadata Discovery Permissions

    GoodData needs to read the information_schema to generate Logical Data Models (LDM). Without this, schema introspection will fail.

    -- Grant access to system views for metadata discovery
GRANT SELECT ON ALL TABLES IN DATABASE information_schema TO ROLE {role_name};
GRANT SELECT ON ALL VIEWS IN DATABASE information_schema TO ROLE {role_name};

  5. Grant Data Access

    Grant read access to the specific business data.

    -- Option A: Grant access to the entire database
GRANT SELECT ON ALL TABLES IN DATABASE {database_name} TO ROLE {role_name};
GRANT SELECT ON ALL VIEWS IN DATABASE {database_name} TO ROLE {role_name};

-- Option B: Grant access to specific tables only
-- GRANT SELECT ON TABLE {database_name}.{table_name} TO ROLE {role_name};

  6. Create User

    Create the specific user and assign the role. Setting catalog and database properties is highly recommended so the user session automatically lands in the correct context.

    CREATE USER {user_name} IDENTIFIED BY '{strong_password}'
DEFAULT ROLE {role_name}
PROPERTIES (
    "catalog" = "{catalog_name}",
    "database" = "{database_name}"
);

Create a StarRocks Data Source

Once you have configured your StarRocks user’s access rights, you can proceed to create a StarRocks data source that you can then connect to.

Steps:

  1. On the home page switch to Data sources.

    The left navigation panel with the Data sources tab highlighted.

  2. Click Connect data.

    The Connect data button highlighted in the top-right corner of the Data sources screen.

  3. Select StarRocks.

    Dialog showing available data source types with the StarRocks option highlighted.

  4. Name your data source and fill in your StarRocks credentials and click Connect:

    The dialog used to enter credentials for the StarRocks data source. Fields include the Data Source Name, Connection URL, a SSL Mode selector, Username, Password, and Database Name.

  5. Click Save.

    Your data source is created!

    The list of data sources displaying the newly created StarRocks data source, including its name and unique ID.

Steps:

  1. Create a StarRocks data source with the following API call:

    curl $HOST_URL/api/v1/entities/dataSources \
-H "Content-Type: application/vnd.gooddata.api+json" \
-H "Accept: application/vnd.gooddata.api+json" \
-H "Authorization: Bearer $API_TOKEN" \
-X POST \
-d '{
    "data": {
    "type": "dataSource",
    "id": "<unique_id_for_the_data_source>",
    "attributes": {
    "name": "<data_source_display_name>",
    "url": "jdbc:mysql://<STARROCKS_HOST>:9030/<STARROCKS_DBNAME>?sslMode=<REQUIRED/VERIFY_IDENTITY>&permitMysqlScheme=true",
    "schema": "<STARROCKS_DBNAME>",
    "type": "STARROCKS",
    "username": "<STARROCKS_USER>",
    "password": "<STARROCKS_PASSWORD>"
}}}' | jq .

  2. To confirm that the data source has been created, ensure the server returns the following response:

    {
  "data": {
    "type": "dataSource",
    "id": "<unique_id_for_the_data_source>",
    "attributes": {
      "name": "<data_source_display_name>",
      "url": "jdbc:mysql://<STARROCKS_HOST>:9030/<STARROCKS_DBNAME>?sslMode=<REQUIRED/VERIFY_IDENTITY>&permitMysqlScheme=true",
      "schema": "<STARROCKS_DBNAME>",
      "type": "STARROCKS",
      "username": "<STARROCKS_USER>"
    }
  },
  "links": {
    "self": "$HOST_URL/api/v1/entities/dataSources/<unique_id_for_the_data_source>"
  }
}

Additional Information

Ensure you understand the following limitations and recommended practice.

Unsupported Features and Limitations

GoodData does not currently support the following features:

  • The following functions are not supported:
    • MEDIAN
    • PERCENTILE
    • CORREL
    • COVAR
    • RSQ
    • SLOPE
    • INTERCEPT
  • Filtering by boolean columns is not supported
  • There are known issues when using SQL datasets with a StarRocks database:
    • Converting a regular dataset into a SQL dataset generates an invalid query. You have to manually rewrite it to make it work.
    • SQL datasets are not supported if the query contains one of these data types: TINYINT, FLOAT, DECIMAL, DOUBLE UNSIGNED.
    • When converting to SQL dataset, the StarRocks YEAR data type is converted to STRING.
  • We do not currently support multiple hosts for StarRocks JDBC URLs.
MySQL
MariaDB