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: Indexes on very small tables

Re: Indexes on very small tables

From: Jonathan Lewis <jonathan_at_jlcomp.demon.co.uk>
Date: Mon, 12 Jul 1999 22:37:02 +0100
Message-ID: <931815913.8248.0.nnrp-09.9e984b29@news.demon.co.uk>


Keep an eye on my website late tomorrow for the first items on a page of 'old wives tables'.

--

Jonathan Lewis
Yet another Oracle-related web site: www.jlcomp.demon.co.uk

mpir_at_compuserve.com wrote in message <7mdkvq$uvi$1_at_nnrp1.deja.com>...
>One consideration is if the table fits in 1 db block buffer (usually
>10K in 5 two K buffers, but it varies by instance and platform). If it
>does, a table scan of one buffer is usually faster than an index range
>scan followed by a fetch by rowid.
>Of course, if you need a unique index for referential integrity, this
>factor is irrelevant.
>
>In article <19990709124923.09630.00008399_at_ng-fg1.aol.com>,
> rniemic_at_aol.com (Rniemic) wrote:
>> The index breakpoint seems to be in the area of 1000 rows but note
>that is
Received on Mon Jul 12 1999 - 16:37:02 CDT

Original text of this message

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