RE: OICA and Oracle choosing the wrong index

From: Jesse, Rich <Rich.Jesse_at_quadtechworld.com>
Date: Tue, 11 May 2004 10:00:42 -0500
Message-ID: <FBE1FCA40ECAD41180400050DA2BC54004E93944@qtiexch2.qgraph.com>

Yes, for those of you in that wonderful Fantasy Land where schemas have = some measure of normalization and referential integrity, apps use bind = variables, proper indexing exists, and the instance doesn't need to do = an unhealthy mix of OLTP and DW, and the beer flows like wine.

Things are different here on the Front. I wish they weren't, but that's = life...

I'm Rich Jesse and I approve of this message (such as it is).

Rich Jesse                        System/Database Administrator
rich.jesse_at_quadtechworld.com      QuadTech, Sussex, WI USA

> -----Original Message-----
> From: oracle-l-bounce_at_freelists.org
> [mailto:oracle-l-bounce_at_freelists.org]
> Sent: Tuesday, May 11, 2004 9:12 AM
> To: oracle-l_at_freelists.org
> Subject: Re: OICA and Oracle choosing the wrong index
>
> based optimizer. There is no valid reason why would someone=20
> want to set OICA to 20 and make
> index I/O 5 times cheaper then the table one. How can things=20
> go wrong? Very simple: if you
> have a report that would be best executed by using the full=20
> table scan, OICA=3D20, might=20
> force the use of an index, which is suboptimal. You can aid=20
> CBO by telling that 75% of=20
> your index blocks are cached, that is normally true, but=20
> don't engage in "price fixing".=20

Please see the official ORACLE-L FAQ: http://www.orafaq.com

To unsubscribe send email to: oracle-l-request_at_freelists.org put 'unsubscribe' in the subject line.

--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------

Received on Tue May 11 2004 - 10:06:13 CDT