When we configure a data source using Hibernate, we should add the hibernate.dialect
property (or eclipselink.target-database
if you are using Ecli
Databases implement subtle differences in the SQL
they use. Things such as data types for example vary across databases (e.g. in Oracle You might put an integer value in a number field and in SQL Server use an int field). Or database specific functionality - selecting the top n rows is different depending on the database. The dialect abstracts this so you don't have to worry about it.