A Cartesian square is the connection
of all elements of a data set, e.g. a database table, with all elements of another data set. The Cartesian square is an important phenomenon in database technology because it is usually to be avoided. If two tables are linked in a database, all records of the first table are linked to all records of the second table without an explicit link rule. Two tables would thus produce resulting recordset with axb rows, where "a" is the row count of table 1 and b is the row count of table 2. For example, a table with 100 records and another with 200 records could potentially create a recordset with 20,000 records. Obviously, this can quickly create extremely large recordsets.
In order to control the amount of data, care is therefore taken with all links between database tables to ensure that only exactly the links that are actually required are implemented by applying a link rule.
There are a few cases where a Cartesian square is desired, for example when a table is to be created with the same default values for each month of the year.