Re: SCN doesn't increment after DROP TABLE

In article <8gm7k2$bv2$1_at_nnrp1.deja.com>,   kal121_at_my-deja.com wrote:
> Are you saying the DDL statements do not generate SCNs? What types of
> transactions generate SCNs?
>
> In article <8glu1a$4a6$1_at_nnrp1.deja.com>,
> Mark D Powell <markp7832_at_my-deja.com> wrote:
> > In article <8gk2nb$qms$1_at_nnrp1.deja.com>,
> > kal121_at_my-deja.com wrote:
> > > How come DROP TABLE statements aren't recorded in the archive
 logs?
 Or
> > > rather, why isn't the SCN number incrementing after I do a DROP
 TABLE?
> > >
> > > This information must be recorded somewhere - or else when you do
> > > recovery, it wouldn't know to drop tables that had been previously
> > > dropped!
> > >
> > Just because Oracle apparently does not update the scn number for a
> > drop table operation does not mean that the DDL has not been
 recorded
> > in the redo log from which all recovery originates. After all when
 you
> > drop a table some information is written to rollback and rollback is
> > written to redo and it would be the scn number of the last update to
> > the table that would be important in a recovery to a time prior to
 the
> > drop.
> > --
> > Mark D. Powell -- The only advice that counts is the advice that
> > you follow so follow your own advice --
> >

You made the comment that the scn number was not being updated by DDL. I just pointed out that it does not matter to recovery as the necessary information is in the redo; I know because I have ran several recoveries that covered object creation and deletion and the database was how it should have been at the end of recovery.

I can not remember the details of when the scn is updated. DML statements cause it and I imagine it is tied to the commit statement, but the details excapes me right now.

--
Mark D. Powell  -- The only advice that counts is the advice that
 you follow so follow your own advice --


Sent via Deja.com http://www.deja.com/
Before you buy.