If you have deceided to concatenate the fact-tables it's really so simple like here:
Load * From a
Load * From b
But it could be that other solutions like joining or mapping maybe combined with concatenating led to better (in which way: load-time, ram-ressources, gui-calculations ?) results. Concatenating is the easiest way and aren't there really performance issues it should be enough. If not you will need test some different approaches and examine the log- and mem-files to optimize your app regarding to your likeliest bottle-neck.
Thank you Marcus for your suggestions.
Looking into the scenario in my case (described above), what could be the possible way to do? I meant do I need to take the Key Fields into consideration? If YES, could you please let me know HOW?
The problem is that, when I am loading the Individual Tables at a time separately with dimensions, I am getting the proper results. Once I am doing a simple concatenate, the association does not happen properly.
For example, For Fact A and Fact B (if loaded separately; one at a time), I am able to filter the metrics by Month and Supplier ID
But once I load both Fact and Fact B (concatenated), the association fails and there is no Month association (or some other dimension association) with the metrics of Fact A.
Could you please help more?
The key-fields from the fact-tables need to be absolutely part from the concatenation and then it shouldn't be issues with associations. Maybe this is helpful in your case: