PostgreSQL 8 serial sequence name for long table names (and columns) does not use standard format ------------------------------------------------------------------------------------------------- Key: HHH-2428 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2428 Project: Hibernate3 Type: Bug Versions: 3.2.0.ga Environment: PostgresSQL 8.1, Windows Reporter: Andrew Wheeler Priority: Minor Tables with serial columns where the both the table and the column are long do not conform to the standard sequence format of [table]_[column]_seq CREATE TABLE incident_property_relationship ( incident_property_relationship_id int4 NOT NULL DEFAULT nextval('incident_property_relationshi_incident_property_relationshi_seq'::regclass), . . } This table was created using the statement: create table incident_property_relationship { incident_property_relationship_id serial not null, . . } I have fixed the issue by creating a new dialect for postgreSQL (PostgreSQL8Dialect) and changed the function getIdentitySelectString to: public String getIdentitySelectString(String table, String column, int type) { return new StringBuffer().append("select currval(pg_get_serial_sequence('") .append(table) .append("','") .append(column) .append("'))") .toString(); } This should also work where the serial column has a custom sequence generator. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators.... - For more information on JIRA, see: http://www.atlassian.com/software/jira