Re: Index rebuilding behavior

From: <Tim_Cuthbertson_at_sonat.com>
Date: Thu, 05 Aug 1999 18:29:18 GMT
Message-ID: <7ocl5d$n5v$1@nnrp1.deja.com>

In article <FFyG1o.DEp_at_news.decus.org>, skubiszewski_at_Eisner.DECUS.Org wrote:
> A curiousity question --
>
> I've been playing with the rebuild option and noticed this
> behavior.
>
> Index has 24 extents all different. 5 each of size 80K,
> 120K, 160K, 200K and 4 of size 240K. One is in tablespace
> A and 23 are in tablespace B. Pctincrease was 1.
>
> I explicitly alter the index to rebuild with storage params
> initial 256K next 256K pctincrease 0 in tablespace B.
>
> End result: 6 extents, 1 is in tablespace A, 80K
> and 4 are in tablespace B, size 280K and one is in
> tablespace B with size 320K.
>
> I thought a new index would be created using
> the old index and the old index would be dropped
> and then replaced with the new.
> This is clearly not happening though.
>
> How does rebuild work?
>
> Rose
>

I am sorry, but I cannot understand what you are describing. An index is a segment, and a segment can only have extents in one tablespace (unless it is a partioned index in Oracle 8 +). Is it a partitioned index?

Tim

Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't. Received on Thu Aug 05 1999 - 13:29:18 CDT