Hello. I have a newbie question regarding MSAS 2005. We have done some aggregations on the cube and we allow writeback to a relational table. I think this is the default method of MSAS. My question is if writeback is allowed, does that blow away any aggregations that were processed the night before? Will aggregations become useless on that particular dimension/fact table intersection if the report has to look at the wrtieback table and the cube?

We are trying to improve performance and the route we are taking is building more and more aggregations. I am thinking this become useless because of the writeback table.

Cube Processing time is not an issue.

Thank you for any replies.