Hi,
so far I've seen M2M dimension design from Microsoft adventure works, it is basically "main fact table join a fact/intermediate dimension (on intermediate dimension table's lowest grain); the fact/intermediate dimension then joins to intermediate fact table (mapping table), which in the end join to M2M dimension".
Our design is a bit different which I am not quite sure it would have potential problem (performance, accuracy of data aggregation). It goes like this:
Main fact table join to Customer table (intermediate dimension) on the dimension's lowest granularity (CustomerID), the Customer table then joins to a intermediate fact table (mapping table) on another attribute of the customer table (i.e. not the lowest grain), and in the end this intermediate fact table joins to M2M dimension table.
and in our dimension usage tab where we define the relationship between M2M dimension and the intermediate fact table, it had b
elow alert:
My question is what kind of side effect will this design/alert have? Yes, in the customer dimension not all attributes are related to the Geography-CD Code
-- Let us help each other to continuously improve.