Loading duplicates records into SSAS 2005

This post is about what will happen when you load duplicate dimension records into SSAS 2005. A following post will deal with duplicate fact records. Will SSAS 2005 generate an error when you try the first scenario? Not by default as we shall see in a simple cube.

When you have primary keys in all dimensions, foreign keys in the fact table and a unique constraint on all the fact table foreign keys the problem with duplicate records will never occur in either the dimension tables or the fact tables. You will get errors when you load the tables in the ETL-process. In reality the data quality is not always that good or you might have performance objectives that will prevent you from having constraints in your starschema.