It appears that there is a bug when using DISTINCT COUNT from nested data cubes. I found this ticket Nested Data Cube DistinctCount Bug
that brings up this bug and I have also seemed to confirm it. I had a data cube that had data cubes as sources. I created a dashboard with DISTINCT COUNT on different dimensions and they were not accurate. Then I created a new data cube using non-cube sources (the same data, basically reproducing what the nested data cubes were doing) and then the DISTINCT COUNT on the fields worked. Was there any resolution to the original ticket and/or is this logged as a bug already with Dundas? Thanks.
Bug with DISTINCT COUNT in nested data cubes
Hi Kelly,
I can see something similar and have passed it on for further investigation. I might ask you for more details later - will give you an update very soon.
Thank you very much, Upasana.