How does RAC Consolidate the redo logs while recovery

I have one doubt regarding RAC recovery on StandBy Site (or in general recovery mechanism of RAC), i am not able to figure how the consolidation of the redo log files from 'N' instances is done and applied at target standby site.

Theortical Case

Lets assume there are 2 instances in a RAC environment, 2 session running on different instances.

T1 Session1 Makes an update (and commit) on Tab1.Col1 to vaule x on instance 1. Lets assume SCN1 at this time.
T2 Session2 Makes an update (and commit) on Tab1.Col1, Tab1.Col2 to vaule y and z resp on instance 2. Lets assume SCN2 at this time.
T3 Session1 Makes an update (and commit) on Tab1.Col1 to vaule a on instance 1. Lets assume SCN3 at this time.
T4 Switch on Instance 1, thread 1 redo 1 would be shipped to standby site
T5 Switch on Instance 2, thread 2 redo 1 would be shipped to standby site

Now at standby site the archive/stand by redo (in case of online apply) would be applied as and when archive logs are shipped, even if it automatically selects which log to apply first and second, how will it handle the SCNs distributted across the logs?

Rgds
Abhay.

In the order of the SCN. :rolleyes:

I guess every one knows that, but asking that very question of how does it consolidate.

Quote:

---

Originally Posted by abhaysk

...Now at standby site the archive/stand by redo (in case of online apply) would be applied as and when archive logs are shipped, even if it automatically selects which log to apply first and second, how will it handle the SCNs distributted across the logs?

---

Oracle applies redo data on the standby database using standard recovery techniques of an Oracle database, redo logs are NOT applied when archive logs shipped, but when all logs required are present. ;)

PS: RAC may have many nodes configured, but is one single "physical" database (data is shared among these nodes).

I understand that all ** necessary ** logs should be present, however i am interested to know the mechanism of resloving SCN's accross the log files. I am interested to know how it will resolve cross over SCNs, ofcourse in detail. I know it will for sure consolidate the logs and then apply, but the question was consolidation process itself. Again we all know its single database.