All Products
Search
Document Center

Data Management:Manage task flows and task flow instances in Task Orchestration Operation Center

Last Updated:Jun 21, 2024

This topic describes how to edit a task flow, view the execution duration of the task flow, and limit the number of SQL statements that are concurrently executed in the task flow in Task Orchestration Operation Center of Data Management (DMS).

Go to the Task Orchestration Operation Center page

  1. Log on to the DMS console V5.0.
  2. In the top navigation bar, choose O&M > Task Orchestration Operation Center.

    Note

    If you use the DMS console in simple mode, move the pointer over the 2023-01-28_15-57-17.png icon in the upper-left corner and choose All functions > O&M > Task Orchestration Operation Center.

Manage task flows

After you go to the Task Orchestration Operation Center page, perform the following operations to manage a task flow: Click Task Flow Management. On the Task Flow Management page, find the task flow that you want to manage and click Edit, Published Tasks, or Running History in the Operation column to view the directed acyclic graph (DAG), published versions, or runtime history of the task flow. You can also view the execution duration trend charts of task flows that are scheduled to or manually executed and publish or unpublish multiple task flows at a time.

Manage the execution of task flows

After you go to the Task Orchestration Operation Center page, click Instance Management. On the Instance Management page, you can perform the following operations:

  • In the Instance Statistics section, view the total number of task flow instances that are run, including the number of task flow instances that are successfully run and fail to be run.

    Note

    The preceding statistics on task flow instances are displayed based on the time when the instances are run. If you want to view the statistics on task flow instances that are displayed based on the business date, select Business Date to the right of the Owner field.

  • Filter the task flow instances that you want to view by instance state, including the Running, Success, and Fail states.

  • View the status of a task flow.

  • On the Instance Management page, find the task flow that you want to manage and click Execution History in the Operation column. You can view the execution duration trend charts of task flows that are scheduled to and manually executed.

image.png

View the scheduling statistics on task flows

After you go to the Task Orchestration Operation Center page, you can display the scheduling time distribution of multiple task flows in the Gantt chart by selecting the execution time and the task flows that you want to view in the Scheduling Statistics section. This helps you check whether the scheduling time of the task flows is appropriate.

Limit the number of SQL statements that are concurrently executed in a task flow

You can use the monitoring and assurance feature of DMS to limit the number of SQL statements that are concurrently executed on a database in multiple task flows. This feature ensures the stability, security, and reliability of databases.

  1. Go to the Task Orchestration Operation Center page.

  2. Click Monitoring and Assurance. On the Concurrency Configuration tab, click New.

  3. In the Add Concurrency Configuration dialog box, configure the required parameters such as Resource Type, Resources, and Resource Quota, and click OK.

    For example, if you set the Resource Quota parameter to 20, a maximum of 20 task flows can be concurrently run on the specified database.

    Note
    • The Resource Quota parameter specifies the maximum number of task flows that can be concurrently run on a database.

    • If the number of task flows that are running exceeds the value of the Resource Quota parameter, DMS runs the task flows based on the specified priority queue.

    • Valid values of the priority are 1 to 10. The greater the value, the higher the priority of a task flow. If a task waits longer than the maximum wait time that is specified for the task node, the task stops queuing and is marked as a failed task.