when use data guard, primary propagate redoLog and standby apply redo log.
- the transaction inner redo log was applied on standby will has same SCN with primary ?
- the apply action will cause the standby's SCN increased bigger also ?
thanks:)
Add a comment
|
1 Answer
The SCN is the internal clock of a database. These SCN are recorded in the datablock if a row changes (Explain SCN and ORA_ROWSCN) and so it also must be recorded in the redo logs and on some other locations. If the redo is applied to a physical standby in a dataguard configuration, the same SCNs are recorded at the same places where they are recorded in the primary database. The standby should be an exact copy of the primary. Whan we switch over to the standby it is in the same state as the old primary.
For a logical standby this is different. I think a logical standby will have it is own SCN numbers. But I have not deep knowledge about logical standby database.
-
standby has its internal private schema or system level tablespace which maybe not replicated from primary ? if so the SCN come from local SGA and redolog will not be consistent, or redolog's SCN will update standby's local SCN in its SGA ? Commented Nov 10, 2021 at 3:53
-
@Chinaxing I was writing about physical standby, where the standby database is an exact copy of the primary database. For such a physical standby there is part that is not exactly the same as the primary database. For a logical standby this is different. I think a logical standby will have it is own SCN numbers. But I have not deep knowledge about logical standby database. Commented Nov 10, 2021 at 11:24
-
thanks. data guard support both of physical and logical standby ? Commented Nov 11, 2021 at 3:21
-
1@Chinaxing yes, it support both. Oracle Data Guard, Concepts and Administration, 19c Commented Nov 11, 2021 at 6:07