Appendix: Force rerun downstream

Limits

• The force rerun downstream feature is available only in DataWorks Professional Edition and later versions. If you are using an earlier version, you must upgrade your DataWorks edition to use this feature. For more information, see Select and purchase a DataWorks edition.

• Only workspace administrators, tenant administrators, and Alibaba Cloud accounts can initiate a force rerun downstream operation.

• When you filter descendant nodes by workspace, you can select only the workspaces for which you are an administrator.

• You can force rerun descendant nodes only for recurring instances that have a data timestamp of yesterday or the day before yesterday and are in the Succeeded or Failed state.

  If an instance does not meet these conditions, the force rerun downstream option is unavailable.

• Each tenant can perform only one force rerun downstream operation at a time in the same region.

Notes

• The force rerun downstream operation sets the state of all selected descendant nodes to Not Run. If a descendant node is running, the operation forcibly stops it.

• If a node's Rerun is set to Cannot Be Rerun After It Succeeds Or Fails, you cannot forcefully rerun its descendant instances. To modify the Rerun Property, see Rerun instructions.

Procedure

1. Select a target instance and navigate to the force rerun downstream page.

   Right-click a target instance. The instance must be a recurring instance that has a data timestamp of yesterday or the day before yesterday and is in the Succeeded or Failed state. Choose Emergency Operations > Force Heavy Run Downstream. The force rerun downstream configuration page appears.

2. Configure the force rerun downstream operation.

   1. Configure basic settings.

      A yellow warning message appears at the top of the Force Rerun Downstream dialog box: "This operation will set all downstream instances to incomplete and restart scheduling. Make sure you understand the impact and notify relevant personnel." The configuration wizard has three steps: Basic Settings, Select Workspace, and Confirm Impact.

      Parameter	Description
      Data Timestamp	Select the data timestamp of the recurring instance that you want to force rerun. You can only force rerun descendant nodes for recurring instances that have a data timestamp of yesterday or the day before yesterday.
      Start Instance	Select the start node in the current workspace for the force rerun downstream operation.
      Heavy Run Starting Point	Select whether to start the rerun from the start node or from its descendant nodes (first-level child nodes).
      Heavy Run Cross-Days Instance	If an instance for the selected data timestamp has cross-day dependencies, you can choose whether to rerun the dependent instances as well. For example, in the scenario shown in the preceding figure, Instance B has a cross-day dependency on Instance A. When you force rerun Instance A with a data timestamp of yesterday, you can choose whether to also rerun the dependent Instance B. If you choose to rerun it, Instance B with a data timestamp of today is also rerun.

   2. Click The next Step. Verify that the information is correct and click OK to complete the basic settings.

      Important

      After you click Next, you cannot modify the information on the current page. To save computing overhead and reduce your waiting time, make sure that the information is correct.

   3. Select workspaces to filter the descendant nodes to be rerun.

      In the Downstream Node the of Work Space list box on the left, select the workspaces for which you want to force a rerun. Click the icon to add them to the list box on the right. The descendant nodes in the selected workspaces are then forcibly rerun. Nodes in any unselected workspaces perform a dry-run.

      Note

      When you filter descendant nodes by workspace, you can select only the workspaces for which you are an administrator.

   4. Click The next Step. Verify the information and click OK to complete the workspace selection.

      Important

      After you click Next, you cannot modify the information on the current page. To save computing overhead and reduce your waiting time, make sure that the information is correct.

   5. Confirm the impact.

      A warning message appears at the top of the page: "This operation will set all downstream instances to incomplete and restart scheduling. Make sure you understand the impact and notify relevant personnel." After you confirm the information, select the I understand the potential impacts of this operation, such as dirty data, and confirm the operation checkbox, and then click Confirm Rerun.

      Parameter	Description
      Number of affected nodes	Shows the number of nodes that will be forcibly rerun.
      Influenced Baseline	The list shows which baselines will be affected by this operation. For more information about baseline monitoring, see Overview of intelligent baselines.
      Force Heavy Run Downstream Reason	Enter the reason for forcing the rerun of descendant nodes.

   6. Click Confirm Rerun. The system resets the status of the selected nodes to Not Run and reruns them according to your scheduling settings.

Example scenario

• Scenario: You initiate a force rerun downstream operation for Node A. You choose to rerun nodes in Workspace A but not in Workspace B.

• Details: Node B in Workspace B is not rerun. Node C in Workspace A is forcibly rerun.