How to annotate MYSQL autoincrement field with JPA annotations How to annotate MYSQL autoincrement field with JPA annotations java java

How to annotate MYSQL autoincrement field with JPA annotations


To use a MySQL AUTO_INCREMENT column, you are supposed to use an IDENTITY strategy:

@Id @GeneratedValue(strategy=GenerationType.IDENTITY)private Long id;

Which is what you'd get when using AUTO with MySQL:

@Id @GeneratedValue(strategy=GenerationType.AUTO)private Long id;

Which is actually equivalent to

@Id @GeneratedValueprivate Long id;

In other words, your mapping should work. But Hibernate should omit the id column in the SQL insert statement, and it is not. There must be a kind of mismatch somewhere.

Did you specify a MySQL dialect in your Hibernate configuration (probably MySQL5InnoDBDialect or MySQL5Dialect depending on the engine you're using)?

Also, who created the table? Can you show the corresponding DDL?

Follow-up: I can't reproduce your problem. Using the code of your entity and your DDL, Hibernate generates the following (expected) SQL with MySQL:

insert into    Operator    (active, password, username) values    (?, ?, ?)

Note that the id column is absent from the above statement, as expected.

To sum up, your code, the table definition and the dialect are correct and coherent, it should work. If it doesn't for you, maybe something is out of sync (do a clean build, double check the build directory, etc) or something else is just wrong (check the logs for anything suspicious).

Regarding the dialect, the only difference between MySQL5Dialect or MySQL5InnoDBDialect is that the later adds ENGINE=InnoDB to the table objects when generating the DDL. Using one or the other doesn't change the generated SQL.


Using MySQL, only this approach was working for me:

@Id @GeneratedValue(strategy=GenerationType.IDENTITY)private Long id;

The other 2 approaches stated by Pascal in his answer were not working for me.


For anyone reading this who is using EclipseLink for JPA 2.0, here are the two annotations I had to use to get JPA to persist data, where "MySequenceGenerator" is whatever name you want to give the generator, "myschema" is the name of the schema in your database that contains the sequence object, and "mysequence" is the name of the sequence object in the database.

@GeneratedValue(strategy= GenerationType.SEQUENCE, generator="MySequenceGenerator")@SequenceGenerator(allocationSize=1, schema="myschema",  name="MySequenceGenerator", sequenceName = "mysequence")

For those using EclipseLink (and possibly other JPA providers), it is CRITICAL that you set the allocationSize attribute to match the INCREMENT value defined for your sequence in the database. If you don't, you'll get a generic persistence failure, and waste a good deal of time trying to track it down, like I did. Here is the reference page that helped me overcome this challenge:

http://wiki.eclipse.org/EclipseLink/Examples/JPA/PrimaryKey#Using_Sequence_Objects

Also, to give context, here is what we're using:

Java 7Glassfish 3.1PostgreSQL 9.1PrimeFaces 3.2/JSF 2.1

Also, for laziness' sake, I built this in Netbeans with the wizards for generating Entities from DB, Controllers from Entities, and JSF from Entities, and the wizards (obviously) do not know how to deal with sequence-based ID columns, so you'll have to manually add these annotations.