From: Tom Lane <tgl@sss.pgh.pa.us>
To: jian xu <jamesxu@outlook.com>
cc: "pgsql-admin@lists.postgresql.org" <pgsql-admin@lists.postgresql.org>
Subject: Re: how to troubleshoot a wrong query plan
In-reply-to: 
 <BL0PR02MB456100A86F1081148D6DBE82A19FA@BL0PR02MB4561.namprd02.prod.outlook.com>
References: 
 <BL0PR02MB45615E13EF7F8CE1AF3FFB04A19FA@BL0PR02MB4561.namprd02.prod.outlook.com>
 <BL0PR02MB45616822F8516830F5900909A19FA@BL0PR02MB4561.namprd02.prod.outlook.com>
 <2746245.1703712111@sss.pgh.pa.us>
 <BL0PR02MB456100A86F1081148D6DBE82A19FA@BL0PR02MB4561.namprd02.prod.outlook.com>
Comments: In-reply-to jian xu <jamesxu@outlook.com>
	message dated "Wed, 27 Dec 2023 21:29:39 +0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2748480.1703713101.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 27 Dec 2023 16:38:21 -0500
Message-ID: <2748481.1703713101@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2748481.1703713101%40sss.pgh.pa.us>
Precedence: bulk

jian xu <jamesxu@outlook.com> writes:
> Thanks Tom, except for setting parameters, is there any other ways to sh=
ow debug information about plan generating? For example, can pg show all t=
he candidate plans it searched, and the cost of each candidate plan?

No, because rejected alternatives are not carried through to the point
of having a displayable plan.  (Trying to do so would be enormously
expensive.)  People have experimented with adding trace facilities
that would show some details about rejected plan alternatives, but
none of that has led to a facility that I'd care to use.

			regards, tom lane