Support
Troubleshooting
This page covers the most common issues encountered while demonstrating or testing NNIPA. For each problem you'll find the most likely cause and the recommended resolution.
Common Issues
"Save as Table" never finishes
Trino or the lakehouse write service is slow or unavailable.
Check the data-query-service and lakehouse-storage-service logs, then rerun the operation.
Transform dropdown or dataset list is empty
No source table has been ingested or no catalog dataset exists.
Run the ingestion workflow first and confirm the dataset appears in the Catalog.
Quality "Dataset" dropdown is empty
The current tenant has no registered datasets.
Register or ingest a dataset before creating a Data Quality suite.
Workflow stays queued
Airflow scheduler is unhealthy or isn't picking up the DAG.
Check docker logs for nnipa-airflow, verify the DAG exists in the shared volume, and confirm the REST API is healthy at :8088/health.
Frontend API returns 404 (production only)
The Kong route is missing.
Add the route to kong.yml and redeploy Kong using deploy:kong-only.
Requests hang on Windows
IPv6 localhost resolution.
Use curl -4 or 127.0.0.1 instead of localhost.
"Save as Table" fails for a Viewer account
The transform:materialize permission is required.
Use an Operator or Administrator account.
- Most demo issues are caused by missing datasets or services that are not running.
- Verify service health before debugging application behavior.
- Make sure the current user has the required permissions for the action being performed.
- If a workflow depends on ingested data, complete the ingestion process before continuing.