From: Tom Lane <tgl@sss.pgh.pa.us>
To: "=?utf-8?B?WHVhbiBDaGVu?=" <cx0222@vip.qq.com>
cc: "=?utf-8?B?cGdzcWwtcGVyZm9ybWFuY2U=?=" <pgsql-performance@postgresql.org>
Subject: Re: Question: Is it valid for a parent node's total cost to be lower
 than a child's total cost in EXPLAIN?
In-reply-to: <tencent_343A4605F05A95847792E921D527E54B5C09@qq.com>
References: <tencent_343A4605F05A95847792E921D527E54B5C09@qq.com>
Comments: In-reply-to "=?utf-8?B?WHVhbiBDaGVu?=" <cx0222@vip.qq.com>
	message dated "Sat, 19 Jul 2025 19:03:14 -0400"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <159696.1752976670.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Sat, 19 Jul 2025 21:57:50 -0400
Message-ID: <159697.1752976670@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/159697.1752976670%40sss.pgh.pa.us>
Precedence: bulk

"=3D?utf-8?B?WHVhbiBDaGVu?=3D" <cx0222@vip.qq.com> writes:
> In some cases, I noticed that a parent path node's total cost is less th=
an that of one of its child path nodes. I initially expected that the tota=
l cost of a node should be at least as large as the sum of its child nodes=
=E2=80=99 total costs, or at least not smaller than any single child.

That is true if we expect the parent node to run its child node(s) to
completion, but there are cases where we don't.  An obvious example
is a Limit node.  In the case at hand, the planner is betting that
the Merge will not need to run one side to completion because the
other side's largest join key value is less than the first side's
largest value.  The merge can stop as soon as either input is
exhausted.

			regards, tom lane