You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I tested this pattern with all successful items, mixed results, and all failed items. All items ran, Join waited for both Split sets, and exactly one status-update job launched afterward.
The setup below gives you two outcomes: fully successful and partial/failed.
1. Set both Split controllers’ Continue Percentage to 0.
This setting controls whether the Split can continue after its jobs finish.
100 requires every job in that Split to succeed.
0 allows it to continue regardless of how many succeeded.
Both settings wait for all jobs in the Split to finish. Setting it to 0 does not make it continue early or turn failed jobs into successful jobs.
2. Change both query-to-Join wires to “On Continue.”
Keep your two decision gates and both Splits. Change these specific wires:
Dealer Attribution Query → On Continue → Join
Competitor Attribution Query → On Continue → Join
The wires remain attached to the query nodes, rather than directly to the Split controllers.
Each query node may run multiple times, but its On Continue wire fires once after its entire controlled set finishes. Join therefore receives one completion from each branch.
The On Success wires from your initial “Status Running” job into the two decision gates can stay as they are.
3. Replace the unconditional “Status Success” job after Join.
The job after Join should no longer unconditionally mark your external workflow run successful. Reaching Join means processing finished, but some items may have failed.
You can replace that step with a simple passthrough job. For example, using the Shell Script plugin:
#!/usr/bin/env node
const{ job }=require('@pixlcore/xyops-sdk');(asyncfunction(){awaitjob.read();constdata=job.getData();// Print the joined input so you can inspect every result.console.log('Joined input:\n'+JSON.stringify(data,null,'\t'));// Preserve the joined data in this job's output.job.addData(data);job.finalSuccess('Joined input inspected');})();
This job’s success means it inspected the input successfully. The parent workflow still remembers any failed query jobs and uses those results when calculating its overall status.
4. Create a standalone status-update event.
Create a regular event using the same plugin, target, and “Update a workflow run” action as your current Status Success step. Give it an enabled Manual trigger so the workflow’s Run Event action can launch it.
For example, call it Update Workflow Completion Status.
Add these two event user parameters:
Parameter ID
Type
Purpose
external_run_id
Text
The workflow run ID in your external system
completion_status
Text
The status to assign
In that event’s plugin parameters, configure:
Plugin field
Value
Workflow Run ID
{{params.external_run_id}}
Status
{{params.completion_status}}
Use distinct user-parameter names as shown above, so they do not collide with the plugin’s own parameter names.
This event runs as a separate follow-up job, outside the completed workflow. Its parameter references therefore use params, rather than the workflow.params references from your original in-graph node.
5. Add two completion actions to the parent workflow.
Open the parent workflow’s edit page and find its workflow-level actions in the details section. These actions belong to the whole workflow, rather than either individual query node.
Add these two enabled actions:
Condition
Action type
Event to run
On Success
Run Event
Update Workflow Completion Status
On Any Error
Run Event
Update Workflow Completion Status
For the On Success action, set its User Parameters to:
Parameter
Value
external_run_id
{{parent.params.workflow_run_id}}
completion_status
succeeded
For the On Any Error action, use:
Parameter
Value
external_run_id
{{parent.params.workflow_run_id}}
completion_status
Your external system’s supported partial/failure status
Here, workflow_run_id is the parent workflow parameter shown in your screenshot. Substitute its actual parameter ID if yours differs.
The Run Event action copies the calling workflow’s parameters into the follow-up job’s parent.params, allowing that expression to resolve the original run ID.
On Any Error includes Warning. You do not need an additional On Warning action for the same status update, since that could launch a second update.
Save the standalone event and the parent workflow.
6. Test both outcomes.
First, run with every item succeeding. Then deliberately fail one query item.
In both cases, check that:
Every Split item finishes.
The passthrough job after Join runs once.
Exactly one follow-up status event launches.
That event receives the correct external run ID and completion status.
The parent workflow normally finishes with Success when all its jobs succeed, or Warning when some succeed and some fail. The completion action then selects the appropriate status update.
There are two details to keep in mind:
The overall result includes every workflow job. If your initial Status Running job succeeds but every query item fails, the parent still has mixed results and finishes with Warning. This setup therefore distinguishes “fully successful” from “partial/failed.” Separate “some query items succeeded” and “no query items succeeded” statuses would need an explicit check of the query jobs’ result codes.
A skipped decision branch does not send a Continue signal. If one decision skips its entire branch, the two-input Join will not launch its passthrough job. The workflow-level completion actions still run when the parent workflow finishes, and its result reflects the jobs that actually ran.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
I tested this pattern with all successful items, mixed results, and all failed items. All items ran, Join waited for both Split sets, and exactly one status-update job launched afterward.
The setup below gives you two outcomes: fully successful and partial/failed.
1. Set both Split controllers’ Continue Percentage to
0.This setting controls whether the Split can continue after its jobs finish.
100requires every job in that Split to succeed.0allows it to continue regardless of how many succeeded.Both settings wait for all jobs in the Split to finish. Setting it to
0does not make it continue early or turn failed jobs into successful jobs.2. Change both query-to-Join wires to “On Continue.”
Keep your two decision gates and both Splits. Change these specific wires:
The wires remain attached to the query nodes, rather than directly to the Split controllers.
Each query node may run multiple times, but its On Continue wire fires once after its entire controlled set finishes. Join therefore receives one completion from each branch.
The On Success wires from your initial “Status Running” job into the two decision gates can stay as they are.
3. Replace the unconditional “Status Success” job after Join.
The job after Join should no longer unconditionally mark your external workflow run successful. Reaching Join means processing finished, but some items may have failed.
You can replace that step with a simple passthrough job. For example, using the Shell Script plugin:
This job’s success means it inspected the input successfully. The parent workflow still remembers any failed query jobs and uses those results when calculating its overall status.
4. Create a standalone status-update event.
Create a regular event using the same plugin, target, and “Update a workflow run” action as your current Status Success step. Give it an enabled Manual trigger so the workflow’s Run Event action can launch it.
For example, call it Update Workflow Completion Status.
Add these two event user parameters:
external_run_idcompletion_statusIn that event’s plugin parameters, configure:
{{params.external_run_id}}{{params.completion_status}}Use distinct user-parameter names as shown above, so they do not collide with the plugin’s own parameter names.
This event runs as a separate follow-up job, outside the completed workflow. Its parameter references therefore use
params, rather than theworkflow.paramsreferences from your original in-graph node.5. Add two completion actions to the parent workflow.
Open the parent workflow’s edit page and find its workflow-level actions in the details section. These actions belong to the whole workflow, rather than either individual query node.
Add these two enabled actions:
For the On Success action, set its User Parameters to:
external_run_id{{parent.params.workflow_run_id}}completion_statussucceededFor the On Any Error action, use:
external_run_id{{parent.params.workflow_run_id}}completion_statusHere,
workflow_run_idis the parent workflow parameter shown in your screenshot. Substitute its actual parameter ID if yours differs.The Run Event action copies the calling workflow’s parameters into the follow-up job’s
parent.params, allowing that expression to resolve the original run ID.On Any Error includes Warning. You do not need an additional On Warning action for the same status update, since that could launch a second update.
Save the standalone event and the parent workflow.
6. Test both outcomes.
First, run with every item succeeding. Then deliberately fail one query item.
In both cases, check that:
The parent workflow normally finishes with Success when all its jobs succeed, or Warning when some succeed and some fail. The completion action then selects the appropriate status update.
There are two details to keep in mind:
The overall result includes every workflow job. If your initial Status Running job succeeds but every query item fails, the parent still has mixed results and finishes with Warning. This setup therefore distinguishes “fully successful” from “partial/failed.” Separate “some query items succeeded” and “no query items succeeded” statuses would need an explicit check of the query jobs’ result codes.
A skipped decision branch does not send a Continue signal. If one decision skips its entire branch, the two-input Join will not launch its passthrough job. The workflow-level completion actions still run when the parent workflow finishes, and its result reflects the jobs that actually ran.
All reactions