How to view and manage manual instances-DataWorks(DataWorks)-阿里云帮助中心

Go to the manual instance page

Log on to the DataWorks console. In the target region, click Data Development and O&M > Operation Center in the left-side navigation pane. Select a workspace from the drop-down list and click Go to Operation Center.
In the navigation pane on the left, click Manually Triggered Task O&M > Manually Triggered Node Instance to open the manual instance page.

Manage instances

You can view and manage manual instances, trigger-based workflow instances, and manually triggered workflow instances.

Manage manual instances

Running a manual task from the Manage manual tasks page generates a manual instance on this page.

Go to the Manually Triggered Node Instance tab. You can filter and view information about manual instances by criteria such as Node Name, Task type, and Business Time.
Perform operations on manual instances.
- Single operation: In the Actions column for an instance, you can Terminate or rerun a manual instance, or click More to view the instance run logs and other information.
- Batch operations: Select multiple manual instances, and then click Stop or rerun at the bottom of the page.
- Export query results: First, filter the instances based on your criteria. After the search is complete, click Export Query Results above the list to export the filtered results to your local device. This is an asynchronous operation. You can monitor the progress and download the file from the View Operation Center action records page.
- DAG operations: For more information, see Manage the DAG for manual instances.

Manage trigger-based workflow instances

You can run a workflow manually in Manage trigger-based workflows or run it by using an event trigger. When the workflow is triggered, a corresponding workflow instance appears on this page.

Go to the Triggered Workflow Instance tab. You can filter and view information about trigger-based workflow instances by criteria such as Node Name, Trigger, and Business Time.
Note
- Manual trigger: If a workflow instance was triggered manually, its Trigger column displays -.
- Event trigger: If a workflow instance was triggered by an event, its Trigger column displays the trigger name.
Perform operations on trigger-based workflow instances.
- Individual actions: You can Terminate or rerun a trigger-based workflow instance in the Actions column of the instance, or click More > Edit Node to edit it in Data Studio.
- Batch operations: You can select multiple trigger-based workflow instances and then click Terminate, Batch Rerun, Set as Successful, Change Priority, Change Resource Group, or Refresh Instance at the bottom of the page.
- Export query results: First, filter the instances based on your criteria. After the search is complete, click Export Query Results above the list to export the filtered results to your local device. This is an asynchronous operation. You can monitor the progress and download the file from the View Operation Center action records page.
- DAG operations: For more information, see Manage the DAG for manual instances.
The icons below the task instance show its run status.
- : The number of running instances.
- : The number of successful instances.
- : The number of failed instances.
- : The number of instances that have not been run.

Manage manually triggered workflow instances

Running a manually triggered workflow in Manage manually triggered workflows generates a manually triggered workflow instance on this page.

Go to the Manual Workflow Instance tab. You can filter and view information about manually triggered workflow instances by criteria such as Workflow, Running status, and Business Time.
Perform operations on manually triggered workflow instances.
- Single operation: You can Batch Terminate or rerun a manually triggered workflow instance in its Actions column.
- Batch operations: You can select multiple manually triggered workflow instances and then click Terminate, Rerun, Set as Successful, Change Priority, Change Resource Group, or Refresh Instance at the bottom of the page.
- Export query results: First, filter the instances based on your criteria. After the search is complete, click Export Query Results above the list to export the current filtered results to your local device. This is an asynchronous operation. You can monitor the progress and download the file from the View Operation Center action records page.
- DAG operations: For more information, see Manage the DAG for manual instances.
The icons below the manually triggered workflow show the run status of its instances.
- : The number of running instances.
- : The number of successful instances.
- : The number of failed instances.
- : The number of instances in other states.

Automated follow-up actions for batch operations

Important

This feature is currently supported only for batch operations on trigger-based workflows.

To improve O&M efficiency, DataWorks provides optional Post-O&M Actions for four batch operations: Change Priority, Modify Resource Group, Refresh Instance, and Stop. This feature lets you define a chained action that runs automatically after the primary operation succeeds. This simplifies the process and ensures that changes take effect immediately.

Use cases

When you modify the properties of an instance that is already in the scheduling queue (for example, with a status of Waiting for Time, Waiting for Resources, or Running), the changes do not immediately apply to the currently queued or running instance. Traditionally, you would have to stop the instance, apply the changes, and then rerun it. The follow-up action feature automates this process, allowing you to complete the entire stop-modify-rerun chain in a single operation.

1. Batch change priority, modify resource group, and refresh instance

When you perform a Change Priority, Modify Resource Group, or Refresh Instance operation on trigger-based workflow instances, you can specify a follow-up action to ensure your modifications take effect immediately.

Procedure

On the Triggered workflow instance page, select one or more instances you want to modify.
Click Change Priority, Change Resource Group, or Refresh Instance at the bottom of the page.
In the dialog box that appears, you will see a new Follow-up O&M Actions section in addition to the main operation settings.

Follow-up O&M action options

You can choose one of the following follow-up actions based on your needs:

Stop instances that are Waiting for Time, Waiting for Resources, or Running:
- Effect: The system first applies the property change you specified (such as changing the priority) and then automatically stops any target instances with a status of 'Waiting for Time', 'Waiting for Resources', or 'Running'.
- Use case: Use this option when you only need to stop the currently running instances, allowing you to decide whether to rerun them later. This prevents instances with old configurations from continuing to consume resources.
Rerun instances:
- Effect: After applying the property change, the system immediately reruns all selected instances.
- Use case: Use this option when you want the changes to take effect immediately and rerun the task with the new configuration (new priority, new resource group, or refreshed code).

2. Batch stop

When performing a Batch Stop operation, you can also define a follow-up action, such as an automatic rerun, to quickly recover from failures after the instances are stopped.

Procedure

On the Triggered workflow instance page, select one or more instances you want to stop.
Click the Terminate button at the bottom of the page.
In the dialog box that appears, you will see a new Post-O&M Actions section.

In the Batch Stop dialog box, you can set Task Scope to All Tasks (stops all tasks within the selected trigger-based workflow instances) or Specified Tasks (stops only specified tasks). You can also filter by Instance Status by selecting Waiting for Time, Waiting for Resources, Running, or Checking. In the Follow-up O&M Actions section, you can select Rerun Instances and enable the Reset instance status before rerunning option.

Follow-up O&M action options

Rerun Instance:
- Effect: After stopping the selected instances, the system immediately reruns them.
- Use case: This option is useful when you detect an issue with a task run (such as it being stuck or running too long) and want to stop it and then immediately restart it with the original configuration.
Reset instance status before rerunning:
- Effect: This is a more forceful stop-and-rerun operation. The system first forces the status of all selected instances to "Not-run" (even running tasks are terminated) and then reruns them.
- Use case: This is suitable for scenarios that require a complete reset of the instance's current state to ensure it starts from a clean slate. For example, use this when an instance's internal state becomes corrupted and a simple stop-and-rerun fails to resolve the issue.

Manage manual instance DAG

Click the instance name, or click Actions in the DAG column to open the DAG for that instance. In the DAG view, you can right-click an instance to perform operations. Right-click a task node, such as ODPS SQL, in the DAG. The context menu provides the following options: show ancestor nodes, show descendant nodes, view runtime log, instance diagnosis, view code, edit node, view lineage, Trigger DQC Check, Stop, rerun, and set to successful.

Note

In the list of manually triggered workflows, click a workflow to display its DAG on the right. The DAG may contain multiple tasks.

Actions	Description
Show Ancestor Nodes	Expands up to six levels of upstream nodes for the selected node. Note You can expand ancestor nodes only for manually triggered workflow instances.
Show Descendent Nodes	Expands up to six levels of downstream nodes for the selected node. Note You can expand descendant nodes only for manually triggered workflow instances.
View Runtime Log	Displays the runtime log for the current instance, which includes statuses such as Running, Succeeded, and Failed. Note You cannot manually delete instances. The platform automatically deletes instances after they expire, which is typically around 30 days. For tasks that run on the shared resource group for scheduling, instances are retained for one month and logs are retained for one week. If the log of a completed instance is larger than 3 MB, it is cleared at a scheduled time each day. For tasks that run on an exclusive resource group for scheduling, instances and logs are retained for one month.
Run Diagnostics	DataWorks Operation Center provides an instance diagnosis feature to help you understand the end-to-end information of a task run and quickly locate issues. For more information, see intelligent diagnosis.
View Code	View the code of the current node.
Edit Node	Opens the workflow in DataStudio for editing.
View Lineage	Displays the data lineage of the current instance.
Terminate	Stops the task from running. This operation is valid only for the current instance.
Rerun	Reruns the selected task. This is useful for handling failed or missed nodes. Note You can rerun only tasks that are in the Succeeded or Failed state.