We recently modified our ION workflow for expense approval routing to add an additional approval step for amounts over $5000. The workflow is processing correctly and approvals are working as expected, but our analytics dashboards in Infor Data Lake are still showing the old approval metrics. The ETL process seems to be running on schedule, but the output mapping from ION doesn’t appear to be feeding the updated approval stage data into our analytics model.
I’ve checked the ION workflow output and the new approval step is generating events, but they’re not reflected in our spend visibility reports. Has anyone dealt with analytics data model alignment issues after modifying ION workflows? I’m wondering if there’s a specific configuration step I’m missing to update the ETL process for the new workflow structure.
I ran into this same scenario last quarter with a client. Here’s a complete approach to resolve your analytics alignment issue:
Step 1: Verify ION Workflow Output Mapping
First, confirm your modified ION workflow is publishing the complete approval stage data. In ION Desk, navigate to your workflow and check the Document Flow configuration. Ensure the new approval step is mapped to output fields in the ExpenseApproval document type. The key fields you need are: ApprovalStageID, ApprovalStageName, ApproverUserID, ApprovalTimestamp, and ApprovalAmount.
Step 2: Update Data Lake ETL Process
In Data Lake Administration, go to ETL Mappings > Expense Analytics. You need to update three components:
Dimension Table Update: Add the new approval stage to the DIM_APPROVAL_STAGE table. In Mapping Studio, select the dimension table and add a new row with your stage details (StageID, StageName, StageOrder, StageDescription).
Fact Table Mapping: Update FACT_EXPENSE_APPROVAL to include mappings for the new stage. Add field mappings from your ION document to the fact table columns. The critical mappings are:
ION.ApprovalStageID → FACT.APPROVAL_STAGE_KEY
ION.ApprovalTimestamp → FACT.APPROVAL_DATE_KEY
ION.ApproverUserID → FACT.APPROVER_KEY
Analytics Data Model Alignment: Update your analytics view definitions to include the new approval stage in queries. In Data Lake, modify the VW_EXPENSE_APPROVAL_METRICS view to join the new dimension values and calculate metrics that include the additional approval tier.
Step 3: ETL Process Update and Testing
Once mappings are updated, you need to refresh the ETL:
In Data Lake ETL Jobs, locate the Expense Analytics job
Run a validation test with a small date range (last 2-3 days)
Check the job logs for any mapping errors or field mismatches
If validation succeeds, run a full reload for the period since your workflow change
Step 4: Data Validation
After the ETL completes, validate your analytics:
Query FACT_EXPENSE_APPROVAL directly to confirm new stage data is populated
Check your Birst or analytics dashboards to verify the new approval metrics appear
Compare approval counts in your operational system vs analytics to ensure accuracy
Regarding Historical Data: If your ION events are still in the Data Lake staging area (typically 30-90 day retention), you can backfill. Otherwise, you’ll need to extract historical approval data from CloudSuite Financials tables directly and perform a one-time data load to populate the historical analytics.
Common Pitfall: Make sure your ETL job sequence includes a dimension load before the fact table load. If the new approval stage isn’t in the dimension table when the fact table ETL runs, you’ll get orphaned records or referential integrity errors.
This complete update ensures your ION workflow output mapping, ETL process, and analytics data model are fully aligned for accurate spend visibility reporting.
This draft is based on general Infor CloudSuite knowledge. It has not been verified against your specific version and environment. Practitioners: verify the steps and share your experience below.
I’ve seen this exact issue before. When you modify ION workflows, the output mapping doesn’t automatically update the downstream analytics schemas. You need to explicitly update the ETL mappings in Data Lake to recognize the new approval step fields. Check your ION API Gateway logs to confirm the new events are being published with the additional approval stage data. The workflow might be working, but if the field mapping isn’t updated in your Data Lake connection point, those new data points just get dropped during ingestion.
Adding to what Mike said - you’ll want to review your Data Lake ETL job definitions. In ICS 2023.1, there’s a specific configuration file for expense analytics that maps ION workflow stages to the analytics dimension tables. The file is usually under your Data Lake tenant configuration. You need to add the new approval stage as a dimension value and update the fact table mapping to capture the new approval timestamp and approver fields. Without this update, your ETL will continue using the old schema definition and ignore the new workflow output.
Thanks both. I found the ETL configuration in Data Lake and I can see the workflow stage mappings. However, I’m not entirely clear on how to add the new approval stage dimension. Do I need to manually edit the JSON configuration, or is there a UI-based tool in Data Lake for updating these mappings? Also, once I update the mapping, will it backfill the historical data from when we made the workflow change, or will it only apply to new transactions going forward?
For ICS 2023.1, you can use the Data Lake Mapping Studio interface to update dimension mappings without directly editing JSON files. Navigate to Data Lake Administration > ETL Mappings > Expense Analytics. Look for the ApprovalStage dimension table and add your new stage value there. Then update the fact table mapping to include the new fields from your ION workflow output.
Regarding backfill - it depends on your Data Lake retention settings. If the raw ION events are still in your staging area, you can trigger a historical reload for the affected date range. Otherwise, you’ll only see the corrected data from the point of the mapping update forward.
One thing to watch out for - make sure your ION workflow is using consistent field names in the output document. I’ve seen cases where the workflow designer created slightly different field names for the new approval step, and even though the mapping was updated, the ETL couldn’t match the fields properly. Use ION Message Monitoring to inspect an actual approval event and verify the exact field names being published. Compare those to what you’re mapping in Data Lake to ensure they match exactly, including case sensitivity.
Tested this on CloudSuite Financials 2023.x — verifying the ExpenseApproval document type output mapping in ION Desk immediately resolved our stale approval stage analytics.