Step 4 - Stack Additional Workspace and Model Datasets

🔍 What You’re Doing

The workflow template arrives configured for one workspace and model combination. If your audit covers more than one, every additional combination needs its own Imports, Exports, Processes, and Actions datasets, and each of those has to be appended to the Stack Actions supporting pipeline before its action names appear in the audit output.

Note

This step is only needed if your audit covers more than one workspace and model combination. A single combination is already handled by the template.

Create the additional datasets first. If you have not already built Imports, Exports, Processes, and Actions for each additional workspace and model combination, return to Step 1 and create them now. Use the same Export Type settings, choosing the workspace and model for that combination, and reduce each one to id and name. Everything below assumes those datasets already exist.

With them in place, the work here is in two parts: add each new dataset to the Workflow, then add a Stack Rows transformation to the end of the Stack Actions pipeline for each one.

---

✅ Summary of Steps

1. Open the Workflow you just created.

2. Search for and select the Dataset.

3. Click Add to Workflow.

4. Click Pipelines.

5. Open the Stack Actions Pipeline.

6. Select the last step in the Pipeline.

7. Click the + below the last step.

8. Select Advanced.

9. Choose Stack Rows as the Transformation Type.

10. Name the transformation after the Dataset you are stacking.

11. Click Dataset to Stack.

12. Select the Dataset for the additional workspace and model.

13. Click Save Transformation.

14. Confirm the stacked actions appear in the output.

---

Repeat for Every Combination

Note

Repeat this process for each remaining dataset and each remaining workspace and model combination. Every combination you want in the audit needs all four of its datasets stacked: Imports, Exports, Processes, and Actions.

This is what fills in the names of those items on the event records. A combination you leave unstacked does not produce an error. Its events still appear in the audit log, but the action name resolves to nothing, so you are left with IDs instead of readable names.

Because all four dataset types were reduced to the same two columns, id and name, in Step 1, every set stacks onto the others without further shaping. A dataset that still carries its extra columns will not line up.

---

Back | Next