Our external forecasting tool pushes demand forecasts to Azure Data Lake, which is integrated with D365 10.0.38 for supply planning. The Data Lake sync process completes successfully with no errors reported in the integration logs, but when we run master planning, it still uses the old forecast data from two weeks ago instead of the updated forecasts.
We’ve verified that the new forecast data is present in Data Lake with correct timestamps and formatting. The Synapse pipeline shows successful execution, and we can query the updated forecast values directly in the Data Lake storage. However, the supply planning forecast in D365 remains completely unchanged, showing the same numbers from our last successful update on July 28th.
Data Lake Sync: Completed Successfully
Forecast Records Updated: 1,247
Master Planning Forecast: Still showing 07/28 data
The planning team is making decisions based on outdated forecasts, which is causing inventory issues. Has anyone successfully integrated external forecast data through Data Lake with D365 supply planning?
Data Lake sync completing doesn’t automatically import the data into D365 tables. You need a separate data entity import job that reads from Data Lake and writes to the demand forecast tables. The sync just makes the data available - check if you have the “Import forecast from Data Lake” recurring batch job configured and running.
The waiting status usually means either the batch job doesn’t have proper execution permissions or there’s a dependency issue. Check if the job is configured to run under a service account that has access to both Data Lake and the demand forecast entities in D365. Also verify that the Data Lake connection string in the job parameters is correct and includes the necessary SAS tokens or managed identity credentials.
Also consider the forecast date ranges. Master planning only considers forecasts within a specific time fence configured in the coverage group. If your external tool is generating forecasts with dates outside this time fence, D365 will import them but planning will ignore them. Check the forecast time fence settings in your item coverage groups.
Beyond the batch job, you need to ensure the forecast model is correctly mapped. D365 can have multiple forecast models, and the Data Lake import needs to target the specific model that master planning uses. Check your master planning parameters to see which forecast model is set as default, then verify your import job is writing to that same model. A mismatch here would explain why data imports but planning doesn’t see it.