This has been an issue with SSAS since forever, but it's been a pain and I don't understand why it happens.
On the first build of a cube, I will get an error like the following:
Errors in the OLAP storage engine: A duplicate attribute key has been found
when processing: Table: 'ServiceLocation', Column: 'location_facility',
Value: 'Acute'. The attribute is 'Facility'.
This is by design. It is standard practice to have "duplicate attributes" in a dimension (e.g. The value "Monday" will be found hundreds or thousands of times in a date dimension).
If I process the dimension ONCE to ignore key errors, the dimension will then process correctly forever thence, even after reverting to normal processing mode (i.e. not ignoring key errors). But until then, it will not complete processing.
This seems perverse: SSAS doesn't seem to be understanding the nature of a completely normal design until it has already processed, and then it does. This is a pain, and means that first deployments of a cube onto a new server will have to be a semi-manual, nursed, step-at-a-time process.
Does anyone understand why this happens? Is there metadata that I'm not setting correctly? Is there a workaround for it? This has been SSAS behavior since I've been designing with it (2004), and it's always been a pain.
