An HTTP trigger node lets you trigger a DataWorks workflow node and its downstream nodes from an external environment — such as a local system or a cross-tenant environment — by calling the TriggerSchedulerTaskInstance OpenAPI operation. This is useful when your orchestration logic lives outside DataWorks and you need DataWorks scheduling to wait for an external signal before proceeding.
How it works
An HTTP trigger node is a zero-load node: it carries no compute logic and exists solely as a gating point in your workflow. DataWorks generates recurring instances for the node on schedule. Each instance waits in the Waiting for Trigger state until your external system calls TriggerSchedulerTaskInstance. Once triggered, the node releases its downstream nodes, which then run in sequence.
Two conditions must both be satisfied before the HTTP trigger node releases its downstream nodes:
All upstream nodes that the HTTP trigger node depends on have run successfully.
Your external system has called
TriggerSchedulerTaskInstancefor that specific instance.
The order in which these two conditions are met does not matter. If the trigger call arrives before the upstream nodes finish, DataWorks holds the instruction and proceeds once the upstream nodes complete.
The trigger instruction is retained for only 24 hours. If the upstream nodes do not finish within 24 hours, the instruction is discarded and the trigger call must be repeated.
Each instance can be triggered only once. Calling TriggerSchedulerTaskInstance on an already-triggered instance has no effect.
Limitations
HTTP trigger nodes are available in DataWorks Enterprise Edition and higher. For a version comparison, see Editions of DataWorks.
HTTP trigger nodes are zero-load nodes and cannot perform compute tasks. Place the actual compute node downstream of the HTTP trigger node.
A recurring instance for the HTTP trigger node must have been generated and must be in the Waiting for Trigger state. You can verify this on the Auto Triggered Instances page in Operation Center.
The scheduled time for the recurring instance must have arrived.
The HTTP trigger node must not be in a frozen state when the trigger call is made.
The scheduling resource group must have sufficient resources at trigger time.
Prerequisites
Before you begin, make sure you have:
A RAM user with the Develop or Workspace Administrator role in your workspace. For details, see Add members to a workspace.
The Workspace Administrator role grants broader permissions than required for this task. Assign it only when necessary.
A serverless resource group associated with your workspace. See Use serverless resource groups.
Create and publish an HTTP trigger node
Create the node
For step-by-step creation instructions, see Create a recurring task.
Configure scheduling
After creating the node, open the Scheduling panel on the right side of the node editing page and configure the following parameters. For all other scheduling options, see Node scheduling.
| Parameter | Description |
|---|---|
| Resource group | Select the serverless resource group associated with your workspace. |
| Instance generation mode | T+1 Next Day: instances are generated the following day. Generate Immediately After Publishing: instances are generated right after the node is published. |
When no upstream node is configured, the root node of the business flow is used as the default upstream node.
Publish the node
Submit and publish the node to the production environment. For details, see Publish nodes/workflows.
After publishing, the node runs on its configured schedule. To monitor it, go to Operation Center > Node O&M > Auto Triggered Node O&M > Auto Triggered Nodes. For an overview of Operation Center, see Get started with Operation Center.
Trigger the node from an external environment
Use the following workflow to call TriggerSchedulerTaskInstance from your external system:
Create a workflow in Data Development that contains an HTTP trigger node, configure node dependencies, and publish to Operation Center.
The system generates recurring instances on schedule. Retrieve the instance's task ID and scheduled time from Operation Center.
Pass those values to
TriggerSchedulerTaskInstanceusing Java, Python, or the API debugging tool. The node releases its downstream nodes, which run in sequence.
For TriggerTime: use a fixed value to trigger only that specific instance; use a dynamic variable to trigger all matching instances.
Step 1: Get the instance task ID and scheduled time
When you can view instances depends on the instance generation mode:
T+1 Next Day: go to Operation Center the next day.
Generate Immediately After Publishing: instances are available right away.
Log on to the DataWorks console. In the top navigation bar, select your region. In the left-side navigation pane, choose Data Development and O&M > Operation Center, select your workspace, and click Go to Operation Center.
In the left navigation bar, click Auto Triggered Node O&M > Auto Triggered Instances.
Find your HTTP trigger node instance and record its Task ID and Scheduled Time.
Hover over the instance name to see the Task ID.
Step 2: Call TriggerSchedulerTaskInstance
Call the API using one of the following methods.
Step 3: Verify that the trigger succeeded
After calling TriggerSchedulerTaskInstance, confirm that the node was triggered successfully:
In Operation Center, go to Auto Triggered Node O&M > Auto Triggered Instances.
Find your HTTP trigger node instance. If the trigger succeeded, its state changes from Waiting for Trigger to a running or success state, and its downstream nodes begin running in sequence.
What's next
For a complete walkthrough of using HTTP trigger nodes in cross-tenant scenarios, see Use HTTP trigger nodes to trigger node execution across tenants.

