An isolated node is an auto triggered node with no ancestor node. An isolated node cannot be scheduled automatically. If other nodes depend on an isolated node, your business may be severely affected.
Every auto triggered node in a workspace must have at least one ancestor node, except for the root node of the workspace. Nodes without ancestor nodes cannot be scheduled as expected.
Diagnose and fix an isolated node
When you receive an isolated node alert, work through the following checks in order to identify and resolve the root cause.
Step 1: Check the instance generation mode of the ancestor node
Symptom: A newly created node with Instance Generation Mode set to Immediately After Deployment depends on another newly created node with Instance Generation Mode set to Next Day. No instance is generated for the ancestor node on the current day, so the dependent node becomes isolated.
Fix: Modify the Instance Generation Mode of the ancestor node and redeploy it. For details, see Scenario 3: Configure different instance generation modes for an auto triggered node and its ancestor node.
Step 2: Check whether the ancestor node is within its validity period
Symptom: DataWorks does not generate instances for nodes outside their scheduled validity period. If the ancestor node's validity period has expired or has not yet started, no instances exist for it, and the dependent node becomes isolated.
Fix: Open the ancestor node's configuration tab, go to Properties > Schedule, and update the Validity Period. For details, see Effective date.
Step 3: Check whether the ancestor node's output has changed
Symptom: When the ancestor node's business logic changes, its output may be renamed or removed. This breaks the dependency between the ancestor node and its dependent node, making the dependent node isolated.
Fix: Reconfigure the dependencies for the isolated node. For details, see Impacts of removing or changing the output of a node.
Step 4: Check cross-workspace dependency and periodic scheduling status
Symptom: The isolated node and its ancestor node reside in different workspaces. If Periodic scheduling is not turned on in the workspace where the ancestor node resides, no instances are generated for that ancestor node, and the dependent node becomes isolated.
This scenario requires coordination with the owner of the ancestor node's workspace. Before contacting them, confirm the workspace name and the ancestor node's location in Operation Center.
Fix: Ask the workspace owner to turn on Periodic scheduling for their workspace. Alternatively, remove the cross-workspace dependency. For instructions on enabling periodic scheduling, see Scheduling settings.
Step 5: Check whether an intermediate task is missing from a data backfill
Symptom: This cause applies only when you have recently run a data backfill. If the scheduling time of an intermediate task falls outside the backfill time range, no instance is generated for that task, and its downstream nodes become isolated.
Example:
-
Task A and Task C run every hour (hourly schedule).
-
Task B runs once a day at
02:00. -
Task C depends on Task B; Task B depends on Task A.
If you backfill data for the period 00:00–01:00, Task B's scheduled time (02:00) falls outside that range. DataWorks generates no instance for Task B in this period, so the node running Task C becomes isolated.
Fix: Run a separate backfill for the intermediate task (Task B) that covers the same time range (00:00–01:00). For details, see Backfill data for an auto triggered node and view data backfill instances generated for the node.
Monitoring and alerting
DataWorks generates instances for auto triggered nodes every night and schedules them to run the following day. Built-in alert rules monitor all auto triggered nodes to ensure instances are generated and run on time.
What triggers an alert: If an isolated node is detected, DataWorks sends an alert notification automatically. Handle the alert as soon as possible to prevent downstream pipeline failures.
Scan schedule: DataWorks scans auto triggered nodes at 09:00, 12:00, and 16:00 every day. If an exception occurs within 10 minutes before a scan, it falls outside the current scan window and is detected at the next scheduled scan.
Who receives notifications: Alert notifications are sent to the node owner by text message or email.
How to update the alert contact: On the Rule Management page, change the alert contact for the isolated node alert rule. For details, see Create a custom alert rule.
View isolated nodes
In Operation Center, go to the Overview page and open the Workbench Overview tab. The Focus on section shows the number and details of all isolated nodes in your workspace.