When will the commited Changes progate to Datafiles ?

Question

Hi Experts,
I have gone through the bellow link .
In other words, you are absolutely correct to say that the data is already
in the data files. But DBWR flushes regardless of whether something is committed
or not. Therefore the data files during normal operation of the datafiles will contain
BOTH committed and uncommitted data. It's the redo stream that allows us to work
out which is which. So you still end up needing the redo logs to make sense of what's in the datafiles Source : Re: Commit/Rollback and Redo Log files.... HJR sir's post.
So we have all the committed changes only in Redo logs and data files contains both committed and un-committed data (Because of DBWR flushes all modified data in buffer cache during checkpoint (regard less of commit) .
When we did the clean instance shut down : either everything is committed (all committed data in redo applied to data files and redo logs are no more required to do the instance recovery . ITS IN CASE OF CLEAN SHUTDOWN.
But if the shutdown did not happen  how the changes in redo log files propagated to data files. What are the responsible back ground processes and what are the steps in this case ?
Waiting for your replies
Thanks in Advance.
Vijay
Edited by: Vijay.Cherukuri on Jul 3, 2012 8:56 PM
Edited by: Vijay.Cherukuri on Jul 3, 2012 9:05 PM

Answer 1

No - there is no parameter that controls how
often you shutdown and restart your database.You mean to say instance recovery happens instance crash only.? ( this is my assumption in this case)
My question is how the changes in redo propagated to data files (assume instance never brought down in last 1 month) ?
And Single node instance. I think in this case bellow points are applicable
Oracle Database performs this recovery automatically on two occasions:
At the first database open after the failure of a single-instance
database or all instances of an Oracle RAC database (crash recovery).
When some but not all instances of an Oracle RAC configuration fail (instance recovery).
The recovery is performed automatically by a surviving instance in the configuration.Please correct me If i did any false statements ?
-thanks
Vijay