From: Tom Lane <tgl@sss.pgh.pa.us>
To: Jerry Brenner <jbrenner@guidewire.com>
cc: pgsql-performance@lists.postgresql.org
Subject: Re: Why is a sort required for this query? (IS NULL predicate on
 leading key column)
In-reply-to: 
 <CACoKFYRfJjpaG+yXAZXdTm7b7UAk4UQwV3-b2H_nmYUMS1TBgA@mail.gmail.com>
References: 
 <CACoKFYRfJjpaG+yXAZXdTm7b7UAk4UQwV3-b2H_nmYUMS1TBgA@mail.gmail.com>
Comments: In-reply-to Jerry Brenner <jbrenner@guidewire.com>
	message dated "Wed, 17 Jan 2024 06:39:06 -0800"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2027066.1705504277.1@sss.pgh.pa.us>
Date: Wed, 17 Jan 2024 10:11:17 -0500
Message-ID: <2027067.1705504277@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2027067.1705504277%40sss.pgh.pa.us>
Precedence: bulk

Jerry Brenner <jbrenner@guidewire.com> writes:
> I'm wondering why a sort is required for this query, as the index should be
> providing the required ordering to satisfy the ORDER BY clause.  Does it
> have to do with the IS NULL predicate on the leading key column in the
> index?

IS NULL is not seen as an equality condition, no.  It's pretty much
of a hack that makes it an indexable condition at all, and we don't
really do any advanced optimization with it.

			regards, tom lane