From: grue@hotmail.com (Grue)
Subject: Re: Efficiency of hierarchical queries
Date: 1998/02/02
Message-ID: <34d62863.22070797@128.158.254.10>#1/1
References: <34D60B3D.41D9798@inconcert.com>
X-Complaints-To: abuse@news.msfc.nasa.gov
X-Trace: news.msfc.nasa.gov 886450556 13566 (none) 128.158.97.165
Organization: http://www.msfc.nasa.gov/
Newsgroups: comp.databases.oracle.misc



The only place I've used a hierarchical query was on a table of about
250,000 records, and it returned my results instantly.  So I assumed
that it was using my indexes.  If it was doing a full table scan on
this table, I would have expected it to be *much* slower, because
without the indexes any query on the table is extremely slow.

On Mon, 02 Feb 1998 13:06:53 -0500, Keith D Gregory
<keith@inconcert.com> wrote:

>Looking at the execution plan for a hierarchical query, I see a table
>scan, no matter what I do for indexing hints. It seems to me that this
>will result in very poor performance for the table in question, which
>could contain several million rows.
>
>So, is there something magical about the CONNECT BY operation, or will
>it really scan my table?
>
>-kdg