The task orchestration feature of Data Management (DMS) allows you to configure a task flow dependency check node. After you configure the node, the task flow depends on the result of its previous cycle or another task flow. You can specify the task flow to depend on another task flow or a node. This topic describes how to configure a task flow dependency check node.
Scenarios
- Cross-task flow dependency: Task Flow A can run only after Task Flow B is run as expected.
For example, Task Flow A is run at 2:00 every Monday to train a recommendation model. Task Flow A depends on the business data generated by Task Flow B from last Monday to Sunday.
- Self-dependency: Task Flow A can run only after its previous cycle is complete.
For example, Task Flow A that performs data cleansing can be run only after the data cleansing of its last cycle is complete.
Procedure
Usage notes
For a new task flow that depends on the result of its previous cycle, you must make sure that at least one node has been run as expected.- Manually triggered task flows: Right-click a descendant node of the task flow dependency check node and click Test run current node or Start test run from current to skip the task flow dependency check node and generate a record for the node that is run as expected.
- Auto triggered task flows: Click the Try Run and select Dry Run. After that, a record is generated for the auto triggered task flow that is dry run. icon next to