From: Tom Lane <tgl@sss.pgh.pa.us>
To: Greg Rychlewski <greg.rychlewski@gmail.com>
cc: pgsql-novice@lists.postgresql.org
Subject: Re: nested loop joins
In-reply-to: 
 <CAKemG7Xi_3eS+wT97ENWbzAmpwoVHJ5=YbQQcnswo5G5qHaNMw@mail.gmail.com>
References: 
 <CAKemG7Xi_3eS+wT97ENWbzAmpwoVHJ5=YbQQcnswo5G5qHaNMw@mail.gmail.com>
Comments: In-reply-to Greg Rychlewski <greg.rychlewski@gmail.com>
	message dated "Tue, 26 Oct 2021 14:47:50 -0400"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <110340.1635274572.1@sss.pgh.pa.us>
Date: Tue, 26 Oct 2021 14:56:12 -0400
Message-ID: <110341.1635274572@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/110341.1635274572%40sss.pgh.pa.us>
Precedence: bulk

Greg Rychlewski <greg.rychlewski@gmail.com> writes:
> I'm looking to understand how nested loop joins with limits work. Say I
> have this query

> SELECT *
> FROM a  INNER JOIN b ON a.id = b.id
> LIMIT 10;

> Say this was done as a nested loop join where the rows of "a" are the outer
> loop.

> In this scenario, would all the rows of "a" be materialized or are they
> lazily evaluated?

Evaluation of the join will stop as soon as it's produced 10 rows.
Only as much of "a" will be read as necessary to get that.
(Potentially, not all of "b" would be read either, if 10 matches
to the first "a" row exist in "b".)

			regards, tom lane