In old version of Oracle database, it is easy to control the join path by reordering the tables in table list after the FROM clause, but it is getting difficult to control the join path in today’s cost based SQL optimizer, I am going to introduce method that is applicable in most databases such as Oracle, Sybase, DB2 and SQL Server in the following:

To control a join path, we cannot tell the database SQL optimizer which path is the best one to select. Instead, we add something to the syntax of the SQL statement that causes an increase to the cost of the current join path selected by database SQL optimizer. Let’s take a look at the following example of two table join scenarios.