[ https://issues.jboss.org/browse/ISPN-2058?page=com.atlassian.jira.plugin.... ] Galder Zamarreño updated ISPN-2058: ----------------------------------- Attachment: treemove_infinispan.log -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/ISPN-2058?page=com.atlassian.jira.plugin.... ] Adrian Nistor commented on ISPN-2058: ------------------------------------- The failing test was not using write skew check so it could not work anyway. Also, the method TreeCacheImpl.move needs to read the parent and check if its 'structure' map contains the child that is being moved. This ensures the write skew check is properly detected if another TX manages to remove/move the same child concurrently. Another problem is TreeCacheImpl.move executes a batch (create child in new location, delete child from old location) but it always commits the batch regardless of exceptions thrown from the operations inside the batch. This can also lead to duplicated nodes if the creation succeeds but the delete fails because the child was already moved/removed by another TX. A new method AutoBatchSupport.failAtomic() was added to support rollback. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira