Re: Transportable Tablespaces and Deferred segment creation "bug"

From: Niall Litchfield <niall.litchfield_at_gmail.com>
Date: Mon, 26 Nov 2012 15:49:57 +0000
Message-ID: <CABe10sZKC_z1trydmJGmdPgL6N5SMoxbFW-FbY3CtsOBW3e_kw_at_mail.gmail.com>

I don't buy the space saving thing (there might be dictionary performance improvements of course). I'm pretty certain that each of your empty tables *if *they had an initial segment created would "waste" 64k (8 8k blocks) of space, that's 625mb for every 10,000 empty objects (it'll probably be a bit more with bitmap extent maps filling I suppose). It strikes me that people who worry about a wasted few hundred mb in a schema that size are probably either overly obsessive, unbelievably effective or just looking in the wrong place for space savings. :)

On Mon, Nov 26, 2012 at 11:17 AM, Norman Dunbar <oracle_at_dunbar-it.co.uk>wrote:

> Hi Dominic,
>
> On 26/11/12 11:13, Dominic Brooks wrote:
> > I know you have a workaround on your blog and the main point is that
> > dbms_tt should probably warn about these, but just wanted to point out
> > that you can materialise these deferred segments using
> > dbms_space_admin.materialize_deferred_segments.
>
> Yes, I know about this, thanks all the same, but in a huge database,
> which this one isn't, imaging the result of materiali[zs]ing all those
> empty tables again. The point of the deferred segment creation is to
> save space - which it does, admirably - seems a shame to have to undo
> all that good work when we need to downgrade/migrate it to a Standard
> Edition database.
>
>
> Cheers,
> Norm.
>
> --
> Norman Dunbar
> Dunbar IT Consultants Ltd
>
> Registered address:
> Thorpe House
> 61 Richardshaw Lane
> Pudsey
> West Yorkshire
> United Kingdom
> LS28 7EL
>
> Company Number: 05132767
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>

-- 
Niall Litchfield
Oracle DBA
http://www.orawin.info


--
http://www.freelists.org/webpage/oracle-l

Received on Mon Nov 26 2012 - 16:49:57 CET