Import a pre-built model

Video introduction

Permissions

Only super administrators and system administrators can import models.

Compute engine description

You can import models only when the compute engine is set to MaxCompute in the compute settings.

Import a model

1. On the Dataphin homepage, click Management Hub in the top menu bar.

2. On the Management Hub page, click Model Marketplace in the navigation pane on the left to open the Model Marketplace page.

3. On the Model Marketplace page, click Import Model to open the model configuration page.

   Click Download Introduction Document to download the model introduction document. This document provides information about data warehouse plans and model building for related industries.

4. On the Import Model page, complete the Initialization and Import Model steps.

   1. Configure the initialization settings.

      Parameter	Description
      Import Method	You can use a MaxCompute project provided by Dataphin or your own MaxCompute project. Use a MaxCompute project provided by Dataphin: Dataphin provides a public trial MaxCompute project to help you try the product and reduce trial costs. If you use the provided MaxCompute project, you do not need to configure connection information. Important Because the public trial MaxCompute project shares resources with other trial customers, the following limits apply: Only table reads are supported. Data Definition Language (DDL) operations and data writes are not supported. When you create a table, a table with the same name might exist. Add a suffix to the table name to avoid conflicts. Data import into shared source tables is not supported. Data exporting is not supported. Use your own MaxCompute project: To use all Dataphin features, use your own MaxCompute project. Note If you use your own MaxCompute project, first create a MaxCompute project. Then, enable Decimal 2.0 in Management->Basic Properties. When you import a model, the required data blocks, projects, and compute sources are automatically created in your MaxCompute project. Therefore, provide the connection information for the compute source. The template import creates tables in the compute source (MaxCompute project), which can cause conflicts and overwrite existing tables. Use an empty MaxCompute project.
      Endpoint	The Endpoint from the compute settings. Click Go to Modify Endpoint to change it. For more information, see Set the compute engine of a Dataphin instance to MaxCompute.
      AccessKey ID, AccessKey secret	Enter the AccessKey ID and AccessKey secret of an account that can access the MaxCompute project data. Obtain the AccessKey ID and AccessKey secret on the User Information Management page. Important To ensure a successful connection between the Dataphin project and the MaxCompute project, enter the AccessKey of a MaxCompute project administrator.
      Compute Source Type	Only the MaxCompute compute engine is supported.
      Data	The name of the data block to import.
      Project Name	The name of the project to import. Both a production project and a development project are created. All imported content is in the development project and is not published to the production environment.
      Select MaxCompute Project	Dev-Prod project: Enter the MaxCompute project for the development (DEV) and production (PROD) projects. Basic project: Enter the MaxCompute project for the Basic project.

   2. Click Validate and Import.

   3. Import the model data.

      • The model import process consists of four steps:

        • Initialize Space: Imports data blocks, compute resources, and projects.

        • Import Physical Tables: Imports physical tables and their data.

        • Import Standard Definitions: Imports common business logic, subject areas, and business entities.

        • Import Models and Nodes: Imports logical dimension tables, logical fact tables, computing nodes, atomic metrics, business filters, and derived metrics.

          Note

          • Models and nodes are imported into the development environment. They are not published to the production environment. This prevents periodic scheduling from consuming compute resources.

          • During the import of physical tables, if a physical table or view with the same name already exists in the project, the system automatically renames the table or view being imported to avoid conflicts.

      • If an error occurs in any step, that step fails. For example, if an object or node fails to import during the Import Models and Nodes step, the entire step fails.

   4. After the Import Model step is complete, click Enter Model to open the Data Architecture page.