Re: online index rebuilds
- From: "Jonathan Lewis" <jonathan@xxxxxxxxxxxxxxxxxx>
- To: <tanel.poder.003@xxxxxxx>, <roger_xu@xxxxxxxxxxx>, <joseph@xxxxxxxxxxxxxx>, "Bobak, Mark" <Mark.Bobak@xxxxxxxxxxxxxxx>, "ORACLE-L" <oracle-l@xxxxxxxxxxxxx>
- Date: Mon, 30 Jan 2006 20:06:29 -0000
You can actually get an index full scan
(rather than fast full scan) for an offline
index rebuild - and I thnk I decided the
choice was cost-based even back in 8i.
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk/faq/ind_faq.html
The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/cbo_book/ind_book.html
Cost Based Oracle: Fundamentals
http://www.jlcomp.demon.co.uk/appearances.html
Public Appearances - schedule updated 10th Jan 2006
----- Original Message -----
From: "Tanel Põder" <tanel.poder.003@xxxxxxx>
To: <roger_xu@xxxxxxxxxxx>; <joseph@xxxxxxxxxxxxxx>; "Bobak, Mark"
<Mark.Bobak@xxxxxxxxxxxxxxx>; "ORACLE-L" <oracle-l@xxxxxxxxxxxxx>
Sent: Monday, January 30, 2006 7:58 PM
Subject: Re: online index rebuilds
Hi,
Thinking about why full table scan is required for online index rebuild... Maybe
the potential index block splits are the reason why index FFS can't be used for
online index rebuild? Consistent reads should overcome this issue though, but
maybe there is something we don't know...
Btw, with offline index rebuild the decision whether to go with FTS or index FFS
is also cost based (at least 10.2). If I have only single-column table indexed
then the offline rebuild may do an FTS instead of fasf full index scan. (also,
even in the case of fast full index scan, rows have to be sorted anyway, because
index fast full scan doesn't guarantee an ordered result).
Tanel.
--
//www.freelists.org/webpage/oracle-l
Other related posts: