Question : Setting Primary Key Clustered with foriegn Key cascade - SQL

I havent done any DBA in two years and I am rusteee...

The scenario is - we have added a custom rolesinroles table that consists of RoleId and Parent RoleId - I am trying to do the following, which may be quite simple

Create a Primary clustered key from both of the Ids - is this correct ?


/****** Object:  Index [PK__RolesInRoles__1B29035F]    Script Date: 07/19/2010 22:14:18 ******/
ALTER TABLE [dbo].[RolesInRoles] ADD PRIMARY KEY CLUSTERED
(
      [RoleId] ASC,
      [ParentRoleId] ASC
)WITH (PAD_INDEX  = OFF, STATISTICS_NORECOMPUTE  = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS  = ON, ALLOW_PAGE_LOCKS  = ON) ON [PRIMARY]


I would like both the RoleId and ParentRoleId to reference another table's primary Key "Roles", and do a cascading delete when that "Role" is deleted if the ParentRoleId = Role.RoleId
or if RoleId = Role.RoleId

Any help is greatly appreciated

Also, while I have your ear - if I try to add a duplicate of the clustered key, is there an elegant way to catch it ?

Thanks!

Answer : Setting Primary Key Clustered with foriegn Key cascade - SQL

>> if I try to add a duplicate of the clustered key, is there an elegant way to catch it ?

You could always create INSTEAD OF INSERT/UPDATE triggers and catch the duplicate (non-trivial) and provide a more friendly message or fail silently.