Re: 'Index Full Scan' for Index Scan without Index Cond - Mailing list pgsql-patches

From Simon Riggs
Subject Re: 'Index Full Scan' for Index Scan without Index Cond
Date
Msg-id 1149606224.2621.515.camel@localhost.localdomain
Whole thread Raw
In response to Re: 'Index Full Scan' for Index Scan without Index Cond  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: 'Index Full Scan' for Index Scan without Index Cond
List pgsql-patches
On Tue, 2006-06-06 at 10:31 -0400, Tom Lane wrote:
> Simon Riggs <simon@2ndquadrant.com> writes:
> > Perhaps we should call it 'Ordered Scan' so that it is completely
> > different from 'Index Scan'?
>
> It's not an either/or proposition; the planner could be using the index
> for both selectivity and ordering.

Good point, but Itagaki's aim was to look at the ordering-only
situation, which I think would benefit from separate wording.

> This discussion is also overlooking
> the possibility that a partial index is being used for its predicate
> (again, possibly in combination with explicit index quals and/or
> interest in the sort order).

Assuming its possible, of course. I wasn't able to verify that code just
by eyeballing it, so it doesn't surprise me if you say there are issues.

--
  Simon Riggs
  EnterpriseDB   http://www.enterprisedb.com


pgsql-patches by date:

Previous
From: Tom Lane
Date:
Subject: Re: Why do we have a WAL record for CLOG page extension?
Next
From: Tom Lane
Date:
Subject: Re: 'Index Full Scan' for Index Scan without Index Cond