All Products
Search
Document Center

Dataphin:Manage modeling tasks

Last Updated:Jan 13, 2026

Modeling tasks are field-oriented Operations and Maintenance (O&M) features in Dataphin for logical table tasks created using standardized modeling. These tasks include logical dimension tables, logical fact tables, and logical aggregate tables. This topic describes how to view and manage modeling tasks.

Background information

  • Modeling tasks include logical dimension table tasks, logical fact table tasks, and logical aggregate table tasks. Each logical table task corresponds to a logical table node. The system automatically configures the schedule based on parameters that you set, such as the statistical period. You can perform O&M on logical table tasks from a table or field perspective.

    Note

    You can view logical table tasks only in projects that support standardized modeling.

  • Modeling tasks automatically generate multiple materialized nodes based on the materialization logic of standardized modeling. Each logical table node is a combination of these materialized nodes. Multiple fields or modeling task nodes can belong to the same materialized node.

Usage notes

Standardized modeling uses a dynamic materialization logic. Therefore, you can perform O&M from a physical or logical perspective:

  • Physical perspective: You may need to view the latest code or operational logs of a materialized node to understand how an operation on the node affects the associated fields or logical table task nodes. For example, to backfill data for a modeling task node, you must first identify the materialized nodes to which the logical table task node belongs. Then, you must backfill data for each materialized node one by one. This process is complex and cannot ensure data consistency.

  • Logical perspective: You only need to select the fields or logical table task nodes that require O&M. The system automatically calculates the corresponding materialized nodes based on the current materialization method. Based on the principles of data consistency and correctness, the system prompts you about related fields that require associated operations. You do not need to be concerned about the physical implementation. This helps you shorten the operational path, reduce O&M costs, and improve O&M efficiency.

Access the modeling task page

  1. On the Dataphin home page, choose Develop > Task O&M from the top menu bar.

  2. In the navigation pane on the left, choose Task O&M > Recurring Task.

  3. In the top menu bar, select the production or development environment.

  4. On the Recurring Task page, click the Modeling Task tab.

Operations in the modeling task list

After a modeling task is submitted to the Operation Center for scheduling, it appears in the Recurring Task > Modeling Task list. The list displays the task object, schedule, priority, owner, related baselines, project, last update time, tags, and supported operations.

  • Recurrence: The time when the recurring task is scheduled to run in the specified time zone.

  • Priority: The priority of the task. For a logical aggregate table, the priorities of all metrics in the table are displayed. Therefore, a logical aggregate table can have multiple priorities.

  • Operation Owner: The O&M owner of the logical table task. For a logical aggregate table, the owners of all metrics in the table are displayed. Therefore, a logical aggregate table can have multiple owners.

  • Related Baselines: The baselines to which the node belongs as a guaranteed leaf node and the related baselines that include the node as an ancestor node.

    Note

    This column is not displayed if the baseline feature is disabled.

  • Project: The project to which the task belongs. The project is displayed in the Project English Name (Project Chinese Name) format. If a logical aggregate table spans multiple projects, the names of the projects are displayed and separated by commas (,).

The following operations are available in the modeling task list.

Operation

Description

DAG

Click image to view the DAG of the modeling task. For more information, see Operations for modeling task DAG nodes.

View recurring instances

View the recurring instances that are generated when a task runs. You can also perform O&M on recurring instances.

Edit development node

Go to the task editing page in the Dev project to edit the task. This operation is applicable only to the Dev-Prod development mode.

Edit node

Go to the task editing page to edit the task. This operation is applicable only to the Basic mode.

View production node

Go to the Prod project to view the task configuration.

Note

This feature is not supported for tasks that are not published to the production environment in Basic mode or Dev-Prod development mode.

Field and batch operations

Go to the field and batch operations list to perform operations on fields. Supported operations include viewing the data generation lineage, viewing the data consumption lineage, and backfilling data. For more information, see Manage modeling task fields.

View production lineage

Click to view the data generation lineage of the table. For more information, see Manage modeling task fields.

View consumption lineage

Click to view the data consumption lineage of the table. For more information, see Manage modeling task fields.

View materialization code

View the materialization code of the logical table task.

View data backfill instances

View and perform O&M on instances generated by data backfill operations.

Data backfill

Backfill data for logical table task nodes. For more information, see Data backfill for auto triggered tasks.

Download ancestor and descendant nodes

Download the list of ancestor and descendant nodes for the current node. The list includes all columns, even those that are not displayed.

After you click Download Ancestor/Descendant Nodes, in the Download Ancestor/Descendant Nodes dialog box, select the levels for ancestor and descendant nodes. You can select from Level 1~Level 10, or All Levels. The default is Level 1 for both. After you select the levels, click OK to download the Excel file. The file is named {task_name}_ancestor_descendant_nodes_{timestamp}.xlsx.

Modify owner

Click to modify the owner of the logical table.

Note
  • This operation is applicable only to the Basic mode and the Prod environment of the Dev-Prod mode. You cannot modify the owner in the Dev environment.

  • Details table: To modify the owner of this table, you must have the O&M permissions on the current project.

  • Aggregate table: To batch modify the owners of metrics in this aggregate table, you can modify only the metrics for which you have O&M permissions in the corresponding project. The system automatically filters out the metrics for which you do not have O&M permissions.

  • For a details table, if you modify the owner of the parent dimension logical table, the owner of the child dimension logical table is automatically changed to the owner of the parent dimension.

Modify priority

For a logical table task that is not on a baseline, click to modify the priority of the logical table.

Note
  • This applies only to the Prod environment in Basic and Dev-Prod modes. The Dev environment cannot be modified.

  • The priority of a logical table task on a baseline is determined by the baseline priority. Adjust the priority on the baseline.

  • Details table: To modify the priority of this table, you must have the O&M permissions on the current project.

  • Aggregate table: To batch modify the priorities of metrics in this aggregate table, you can modify only the metrics for which you have O&M permissions in the corresponding project. The system automatically filters out the metrics for which you do not have O&M permissions.

  • If the baseline feature is enabled for the current tenant, you can set the task priority only to Lowest, Low, or Medium. The configuration takes effect only for materialized nodes that are not on a baseline and their corresponding fields. To set a higher priority, configure it on the baseline.

  • If your compute engine type is MaxCompute, the correspondence between Dataphin task priorities and MaxCompute job priorities is as follows: lowest (9), low (7), medium (5), high (3), and highest (1). For more information about MaxCompute job priorities, see Job priority.

  • The priority of a Spark SQL task takes effect only if different priority queues are configured for tasks in HDFS of the Hadoop compute engine.

Configure monitoring and alerting

Configure monitoring rules for task runs. For more information, see Overview of offline task monitoring.

Note
  • This operation is applicable only to Prod and Basic projects.

  • Only accounts that have permissions to configure task monitoring in the current project can configure monitoring and alerting.

Operations for modeling task DAG nodes

The Directed Acyclic Graph (DAG) shows the dependencies between the task node and its ancestor and descendant nodes. You can also perform O&M on the ancestor and descendant nodes. By default, the DAG shows the main node (the selected node) and its first-level ancestor and descendant nodes. You can select a modeling task node to perform O&M operations on the task.

Dataphin supports cross-project O&M. To perform O&M operations on a cross-project modeling task node, you must have view and operation permissions for the project to which the metric belongs. For an aggregate table task, you can perform operations only on metrics for which you have the required permissions.

image.png

  • Operations in the DAG

    Operation

    Description

    Expand Parent Nodes

    Expand the dependency nodes at different levels of the main node in the DAG.

    Expand Child Nodes

    Expand Materialized Nodes

    Expand the materialized nodes of the current logical table task node to view the names of the materialized nodes, the fields in the materialized nodes, and the materialization code.

    image

    View Operation Logs

    View the operation logs of the current node.

  • Operations for modeling nodes

    Hover over a DAG node to view its name, type, schedule, owner, and description. The operations for modeling DAG nodes are the same as the operations in the modeling task list. For more information, see Operations in the modeling task list.

    Note

    Because an aggregate table task does not belong to a specific project, you cannot configure monitoring and alerting for cross-project metrics and their aggregate tables within the current project.

Batch operations for modeling tasks

Modeling tasks support the following batch operations:

Operation

Description

Modify O&M Owner

Batch modify the owners of modeling tasks.

Note
  • This applies only to the Prod environment in Basic and Dev-Prod modes. You cannot make modifications in the Dev environment.

  • To batch modify the owners of selected tables or metrics, you must have the O&M permissions on the corresponding project.

Modify priority

Batch modify the priorities of modeling tasks.

Note
  • This applies only to the Prod environment in the Basic and Dev-Prod patterns. Modifications are not supported in the Dev environment.

  • To batch modify the priorities of selected tables or metrics, you must have the O&M permissions on the corresponding project.

  • If the baseline feature is enabled for the current tenant, you can set the task priority only to Lowest, Low, or Medium. The configuration takes effect only for materialized nodes that are not on a baseline and their corresponding fields. To set a higher priority, configure it on the baseline.

  • If your compute engine type is MaxCompute, the correspondence between Dataphin task priorities and MaxCompute job priorities is as follows: lowest (9), low (7), medium (5), high (3), highest (1). For more information about MaxCompute job priorities, see Job priority.

  • The priority of a Spark SQL task takes effect only if different priority queues are configured for tasks in HDFS of the Hadoop compute engine.

Data Backfill

By default, data is backfilled for the selected tasks and their ancestor and descendant nodes. The backfill operation uses the All Fields and Large Volume Mode - Specify Node ID methods. You can customize the basic information and other configurations for data backfill, except for the timestamp range. For more information about the parameters, see Backfill data for the current node and its ancestor and descendant nodes.