I am working on a possible scenario, where a big data changes would happen on oracle dataguard primary (a large DB) and if something is wrong with the data quality after I would like to use my standby as the fallback scenario.
My idea was:
- stop log apply on standby:
EDIT DATABASE 'standby_db' SET STATE='APPLY-OFF';
- do the 24 hours SQL updates/inserts on primary
- if data quality is not ok , make a switchover or failover from primary to standby (because standby has the data state before the updates/inserts began on primary).
No if I understand on SWITCHOVER the role transition from primary DB to standby DB , the primary would place a marker on the last change in the redo, which is sent to standby DB. Primary converts itself to standby DB. The standby would note the marker in the redo from primary. But I have stopped LOG apply , so I would not like to have the latest changes from the primary I would like to start using my DB from the data state I have "frozen" on stanbdy by stopping log apply. A failover is not an option here also , since my primary is live and kicking and not lost?
UPDATE to the question:
I guess I will be using the FAILOVER. When I stop the apply log on standby, then thh broker will not cooperate later when using Failover. So I will then kill all the sessions on primary, and make
ALTER DATABASE FAILOVER TO standby_db;
Does this sound reasonable or not? Thanks in advance.