Details
-
Suggestion
-
Resolution: Won't Fix
-
None
-
None
Description
NOTE: This suggestion is for Confluence Server. Using Confluence Cloud? See the corresponding suggestion.
Reference material:
http://msdn.microsoft.com/en-us/library/ms151254.aspx
Administrators who wish to implement transactional replication as part of their disaster recovery framework are blocked because 2 tables in confluence do not have primary keys.
conf50od27=# \d hibernate_unique_key Table "public.hibernate_unique_key" Column | Type | Modifiers ---------+---------+----------- next_hi | integer | conf50od27=# \d cwd_application_key Did not find any relation named "cwd_application_key". conf50od27=# \d cwd_application_address Table "public.cwd_application_address" Column | Type | Modifiers -----------------------+------------------------+----------- application_id | bigint | not null remote_address | character varying(255) | not null remote_address_binary | character varying(255) | remote_address_mask | integer | not null Foreign-key constraints: "fk_application_address" FOREIGN KEY (application_id) REFERENCES cwd_application(id)
I'm not sure of any reason why these couldn't be added.
Hibernate_unique_key shouldn't be nullable anyway, and it gets updated seldomly so the extra overhead of having it be primary shouldn't be a big deal. There's only ever 1 row in that table anyway.
application_id in the cwd_application_address seems like a perfect primary key: integer based and not nullable already.
Attachments
Issue Links
- relates to
-
CONFCLOUD-28487 Implement primary keys on hibernate_unique_key and cwd_application_address to facilitate transactional replication on SQL Server
- Closed