Re: Standby on different platforms

"Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message news:1lCJ9.995$jM5.2873_at_newsfeeds.bigpond.com...
>
> "Paul Brewer" <paul_at_paul.brewers.org.uk> wrote in message
> news:3df6f0fa_2_at_mk-nntp-1.news.uk.worldonline.com...
> > "Richard Foote" <richard.foote_at_bigpond.com> wrote in message
> > news:8bkJ9.454$jM5.1352_at_newsfeeds.bigpond.com...
> > > Hi Steven,
> > >
> > > The answer is no.
> > >
> > > BUT ORACLE, IF YOU'RE LISTENING, THIS WOULD BE A REALLY HANDY FEATURE
!!
> > >
> > > Maybe one day with the Logical SB.
> > >
> > Richard,
> >
> > What's a logical SB?
>
> With traditional standby, we ship redo change vectors ("Change this column
> for rowid x.x.x.x and set value to Y") -which means that every row must be
> in exactly the same spot on both the primary and the standby (because file
> number, for example, is part of the rowid -so if it's file 6 on the
primary,
> it's got to be file 6 on the standby).
>
> New in 9i is the logical database, where we convert the redo change vector
> into a simple SQL statement ("update emp set sal=8000 where ename='BOB'").
> Meaning that the EMP table can be in file 6 on ythe primary, but file 67
on
> the standby for all we care: SQL is SQL and works, wherever the EMP table
> is.
>
> Now don't ask me why the logical standby doesn't work across different
> platforms, because in my book, SQL issued on NT should be identical to the
> SQL issued on Unix.... but it doesn't. But I suspect there's no
fundamental
> reason that it can't be *made* so to work in future releases, so I'm
> confidently expecting Richard's wish to come true in the future. Sometime.
>
> But what do I know??
>
> Regards
> HJR
>
>
>
>
>
> >I'd have thought we'd need something like a logical
> > redo log to transport and apply.
> > I must be missing something here.
> > Serious question, btw, even if stupid.
> >
> > Regards,
> > Paul
> >
> >
> >
>

Thanks for the explanation, Howard.

Regards,
Paul Received on Wed Dec 11 2002 - 14:24:50 CST