Re: Accessing rows in large tables

From: Peter Y. Hsing <hsing_at_ix.netcom.com>
Date: 1996/05/23
Message-ID: <4o28pa$bn3_at_dfw-ixnews4.ix.netcom.com>#1/1


On a table of that size, you HAVE to use indices to have any reasonable performance.

  1. Index fields which will be used often in queries.
  2. Index fields which have high selectivity, i.e., one particular value in the field does not account for more than roughly 15% of all values (guideline).

-Peter

On Wed, 22 May 1996 18:15:49 +0300, Tsagarakis Manolis <tzagara_at_cti.gr> wrote:

>
>Hi there,
>
>
>I am currently working on a project, where we will have large tables. There
>will be a table, with maximal 1 million rows. We don't know the exact
>number of rows it will have. Assuming the worst case (1 million rows) we
>are looking for ways, to access rows fast. I've never worked with large
>tables in ORACLE, so I have the following questions:
>
> 1. Is it efficient to use indexes ?
> I red somewhere, that building indexes is efficient
> only for tables having number of rows in a particular range
> (eg. from 3000 to 300,000 rows).
>
> 2. Can someone point me sites, where I can find about large tables in
> ORACLE ?
>
>
>Thank you very much in advance
>
>
>PS. We have ORACLE 7.2 on a DEC Alpha 5/300
>
>************************************************************************
>Emmanuel Tzagarakis, tzagara_at_cti.gr
>Computer Technology Institute (CTI)
>University of Patras
>Patras
>Greece
>*************************************************************************
>
>
>
>
>
>
>
Received on Thu May 23 1996 - 00:00:00 CEST

Original text of this message