[ https://issues.apache.org/jira/browse/RANGER-1820?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Pradeep Agrawal reopened RANGER-1820: ------------------------------------- Re-opening this issue as an entry of new SQL patch 029 need to be added in optimized SQL server script as well. > Duplicate entries should be deleted before creation of unique index on > x_group and x_group_users table > ------------------------------------------------------------------------------------------------------ > > Key: RANGER-1820 > URL: https://issues.apache.org/jira/browse/RANGER-1820 > Project: Ranger > Issue Type: Bug > Components: Ranger > Affects Versions: 0.5.0, 0.6.0, 0.7.0, 1.0.0 > Reporter: Pradeep Agrawal > Assignee: Pradeep Agrawal > Fix For: 1.0.0, 0.7.2 > > Attachments: RANGER-1820-1.patch > > > Currently, concurrent request to create a group and group-users mapping is > causing duplicate entries of the same group in the x_group table and > user-group mapping entry in the x_group_users table. Duplicate entry creation > has been restricted through the RANGER-1765 patch which creates unique > constraints on the x_group and x_group_users table. But unique index creation > may fail if there is already duplicate entries(likely to happen in upgrade > case) so we should delete duplicate entries of these tables before creating > the unique index. If the reference of duplicate group id is being used in > child table then the id of the duplicate group shall be updated in the child > table. In case of duplicate entries of the same group name; group name having > the minimum id(created first) shall be given higher preference and will not > be deleted from the x_group table. -- This message was sent by Atlassian JIRA (v6.4.14#64029)