[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-3352?page=c... ] Donnchadh O Donnabhain commented on HHH-3352: --------------------------------------------- Interestingly, in my case moving the join to my subclasses eliminated the temporary table. i.e. I changed the following: <hibernate-mapping package="com.company.project.domainmodel.articles"> <class name="AbstractArticle" table="ARTICLES"> <id name="id"> <generator class="native" /> </id> <discriminator column="article_type" /> ... <join table="ARTICLE_BODY" optional="true"> <key column="article_id" not-null="true" /> <property name="body" /> </join> <subclass name="PublishedArticle" discriminator-value="PUBLISHED"> ... </subclass> <subclass name="UnpublishedArticle" discriminator-value="UNPUBLISHED"> ... </subclass> </class> </hibernate-mapping> to: <hibernate-mapping package="com.company.project.domainmodel.articles"> <class name="AbstractArticle" table="ARTICLES"> <id name="id"> <generator class="native" /> </id> <discriminator column="article_type" /> ... <subclass name="PublishedArticle" discriminator-value="PUBLISHED"> ... <join table="ARTICLE_BODY" optional="true"> <key column="article_id" not-null="true" /> <property name="body" /> </join> </subclass> <subclass name="UnpublishedArticle" discriminator-value="UNPUBLISHED"> ... <join table="ARTICLE_BODY" optional="true"> <key column="article_id" not-null="true" /> <property name="body" /> </join> </subclass> </class> </hibernate-mapping> and changing the update statement to refer to the base class rather than the subclass (as it happens it was reasonable to do this in our case). The temporary tables (especially with HHH-3919 ) were causing havoc with MySQL 5.0 statement-based replication. -- 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