Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Taking tablespace and datafiles offline

Re: Taking tablespace and datafiles offline

From: Richard Foote <richard.foote_at_bigpond.com>
Date: Sat, 17 May 2003 00:01:30 +1000
Message-ID: <FW5xa.35548$1s1.515544@newsfeeds.bigpond.com>


"Ryan" <rgaffuri_at_cox.net> wrote in message news:Y9Ywa.14883$823.8441_at_news1.east.cox.net...
>
> "Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message
> news:KaXwa.35169$1s1.509161_at_newsfeeds.bigpond.com...
> >
> > "Ryan" <rgaffuri_at_cox.net> wrote in message
> > news:3dXwa.14795$823.10427_at_news1.east.cox.net...
> > > So where is the functionality of being able to take a datafile offline
> > while
> > > keeping others owned by the tablespace online? Users will still not be
> > able
> > > to access some data?
> > >
> >
> > Er, I said that right from the beginning!
> >
> > The functionality is that if you happen to want data in one of the files
> > that remain online, you can still get it.
> >
> > We're having a similar debate in Australia at the moment. Apparently, a
> tax
> > cut that's only worth $4 a week isn't worth having, and the government
> > shouldn't have bothered making it.
> >
> > Personally, I think it better than a tax hike.
> >
> > Whaddya want? Blood??
> >
> > HJR
>
> what kind of system would you use a design where you would take certain
> datafiles in a tablespace offline?
> any rules of thumb in this?
>

Hi Ryan,

You need to ask yourself "why" would you want to take an individual datafile offline ?

The answer that most readily springs to (my) mind is recovery. The datafile is stuffed and I want to recover it.

Now if we took the approach of having an entire tablespace worth of data (whatever that may represent) stored in one huge datafile and the file become stuffed, we would have to take the datafile/tablespace offline (meaning *none* of the data would be available) restore this huge file (time, time, time, time), and then apply all the necessary redo to this huge file (time, time, time time).

But if we were to store the same data, but in several datafiles within the tablespace and just one file became stuffed, then we could take just the naughty datafile offline (meaning all data in the other files could still be accessible), restore just this one smaller file (less time) and apply redo to just this one smaller file (less time).

Sounds like a useful design and rule of thumb ;)

Cheers

Richard Received on Fri May 16 2003 - 09:01:30 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US