We are currently developing a DataWarehouse in Azure where we are loading D365 out of the box data entities, as well as custom entities.
Our approach is to move the full BYOD entities into an Operational DataStore (ODS), which is in our case generated by a DataWarehouse automation tool called BIGenius. From the ODS, we are moving/loading only the needed attributes into our Core DataWarehouse (which is also generated by BIGenius and directly linked to the ODS). The BYOD export is running on incremental basis, where we also discovered issues with the export of entities on incremental basis. That's something we are monitoring closely, but still need to gain more experience. Since we are running daily loads for D365, we would have the flexibility to switch to a full load for entities which are problematic…. however, we trust in Microsoft to fix existing issues here.
I understand your fears of data changes based on the periodic release cycles. However, with the approach of the automatically generated ODS environment, you would be able to catch most changes. Of course, you still have to run updates of your DWH environment parallel to the D365 release…however if I understood correctly, you can skip a certain number of D365 releases and only need to run about 3 updates a year…. but this is something you need to validate.
I also understood that Microsoft is pushing the functionality of the Entity store export to an Azure Data Lake V2. However, last time I looked at it…it was still in Preview. Also I assume it will take some time until it is really an alternative for the BYOD export.hope this helps you make your decision. Would be very interested in hearing other strategies and experiences with D365 data access.kind regardsAndreas
If you've found this thread useful, dive deeper into User Group community content by role