All my entities use this type of @Id
@Id @GeneratedValue(strategy = GenerationType.SEQUENCE, generator = "MYENTITY_SEQ") @SequenceGenerator(name = "MYENTITY_SEQ", sequenceName = "MYENTITY_SEQ") @Column(name = "MYENTITY", nullable = false) private Long id; or
@Id @Column(name = "MYENTITY") I find that an Oracle sequence named hibernate_sequence is always created. Why is this so? And how can I avoid this?
I am using JPA1 with Hibernate 3 and the Oracle 10g dialect.
The HIBERNATE_SEQUENCE is used with REVINFO-entity for creating revision numbers. If you want to use different sequence you should create your custom revision entity.
Help with that: http://docs.jboss.org/hibernate/envers/3.5/reference/en-US/html/revisionlog.html
I see the following code in org.hibernate.id.SequenceGenerator:
public void configure(Type type, Properties params, Dialect dialect) throws MappingException { ObjectNameNormalizer normalizer = ( ObjectNameNormalizer ) params.get( IDENTIFIER_NORMALIZER ); sequenceName = normalizer.normalizeIdentifierQuoting( PropertiesHelper.getString( SEQUENCE, params, "hibernate_sequence" ) ); parameters = params.getProperty( PARAMETERS ); if ( sequenceName.indexOf( '.' ) < 0 ) { final String schemaName = normalizer.normalizeIdentifierQuoting( params.getProperty( SCHEMA ) ); final String catalogName = normalizer.normalizeIdentifierQuoting( params.getProperty( CATALOG ) ); sequenceName = Table.qualify( dialect.quote( catalogName ), dialect.quote( schemaName ), dialect.quote( sequenceName ) ); } else { // if already qualified there is not much we can do in a portable manner so we pass it // through and assume the user has set up the name correctly. } this.identifierType = type; sql = dialect.getSequenceNextValString( sequenceName ); } Where the third parameter of PropertiesHelper.getString(String, Properties, String) is the default property value.
So I'm tempted to say that, somewhere, you have an Id not "properly" annotated. Maybe you should perform a little debugging session.
