In a star schema the fact table links to a number of dimension tables. The row in the fact table will have a column with an identifier mapping to a specific row in one of the dimensions. This column will be used to join back to the dimension table in SQL statements.   When the DB2 SQL Query Compiler, a.k.a. the optimizer, determines an estimated result set for a join, it will consider the cardinality of the join columns in each of the tables that are joined. The cardinality of a column is defined as the number of unique values for that column... [More]

Tags:  underestimation view db2exfmt db2caem cardinality ibm-blog statistical