From: =?ISO-8859-1?Q?Gunnlaugur_=DE=F3r_Briem?= <gunnlaugur@gmail.com>
Subject: Re: Constraint exclusion on UNION ALL subqueries with WHERE
 conditions
Date: Thu, 22 Sep 2011 02:43:25 -0700 (PDT)
Organization: http://groups.google.com
Lines: 15
Message-ID: <6377661.339.1316684605118.JavaMail.geo-discussion-forums@yqih9>
References: 
 <bd6669a0-5da6-410c-b3a2-4a1c01777003@glegroupsg2000goo.googlegroups.com>
	<25440.1316404770@sss.pgh.pa.us>
	<12971543.2235.1316510130842.JavaMail.geo-discussion-forums@yqcd38>
	<25530.1316625060@sss.pgh.pa.us>
Reply-To: pgsql.performance@googlegroups.com
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <25530.1316625060@sss.pgh.pa.us>
Complaints-To: groups-abuse@google.com
Injection-Info: glegroupsg2000goo.googlegroups.com;
 posting-host=157.157.92.142;
	posting-account=YRSnRgoAAABkgkyuFACt4iC5RsuqV7Ez
User-Agent: G2/1.0
To: pgsql-performance@postgresql.org

Right, the view that prompted this involved subqueries; the function was ju=
st an artificial test case.

That change seems like a good one for sure.

Ideally I'd like to enable it for a particular view rather than incur the p=
lanning expense for the whole DB (something like ALTER VIEW foo WITH CONSTR=
AINT EXCLUSION), but I guess there's no support currently (and not easily a=
dded) for such per-object planner settings? The application can just issue =
SET constraint_exclusion=3Don; as needed; for my case that's fine, but for =
DBAs maybe a bit limiting.

Regards,

- Gulli