MIME-Version: 1.0
References: <7AC7104E-E233-4CCA-B99E-12197D61A4DD@americanefficient.com>
 <20201026170405.GT9241@telsasoft.com>
 <DF40D416-6F79-4912-9306-F012D0BED09A@americanefficient.com>
In-Reply-To: <DF40D416-6F79-4912-9306-F012D0BED09A@americanefficient.com>
From: Michael Lewis <mlewis@entrata.com>
Date: Mon, 26 Oct 2020 11:20:01 -0600
Message-ID: 
 <CAHOFxGpcN+=Vo10=2roxOOm8UXu8zstdaggRq0vvu00wcC6YjQ@mail.gmail.com>
Subject: Re: Understanding bad estimate (related to FKs?)
To: Philip Semanchuk <philip@americanefficient.com>
Cc: Justin Pryzby <pryzby@telsasoft.com>,
	postgres performance list <pgsql-performance@postgresql.org>
Content-Type: multipart/alternative; boundary="00000000000075248305b29625cd"
Precedence: bulk

--00000000000075248305b29625cd
Content-Type: text/plain; charset="UTF-8"

On Mon, Oct 26, 2020 at 11:14 AM Philip Semanchuk <
philip@americanefficient.com> wrote:

> >> The item I'm focused on is node 23. The estimate is for 7 rows, actual
> is 896 (multiplied by 1062 loops). I'm confused about two things in this
> node.
> >>
> >> The first is Postgres' estimate. The condition for this index scan
> contains three expressions --
> >>
> >> (five_uniform = zulu_five.five_uniform) AND
> >> (whiskey_mike = juliet_india.whiskey_mike) AND
> >> (bravo = 'mike'::text)
>

Are the columns correlated? Have you tried to create extended statistics
and see if the estimate changes? I believe that extended stats will not
directly help with joins though, only group bys and perhaps choosing an
index scan vs table scan when comparing the correlated columns to static
values rather than joining up tables. Wouldn't be much effort to try it
though.

--00000000000075248305b29625cd
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr">On Mon, Oct 26, 2020 at 11:14 AM Philip S=
emanchuk &lt;<a href=3D"mailto:philip@americanefficient.com">philip@america=
nefficient.com</a>&gt; wrote:<br></div><div class=3D"gmail_quote"><blockquo=
te class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px =
solid rgb(204,204,204);padding-left:1ex">&gt;&gt; The item I&#39;m focused =
on is node 23. The estimate is for 7 rows, actual is 896 (multiplied by 106=
2 loops). I&#39;m confused about two things in this node.<br>
&gt;&gt; <br>
&gt;&gt; The first is Postgres&#39; estimate. The condition for this index =
scan contains three expressions --<br>
&gt;&gt; <br>
&gt;&gt; (five_uniform =3D zulu_five.five_uniform) AND<br>
&gt;&gt; (whiskey_mike =3D juliet_india.whiskey_mike) AND<br>
&gt;&gt; (bravo =3D &#39;mike&#39;::text)<br></blockquote><div><br></div><d=
iv>Are the columns correlated? Have you tried to create extended statistics=
 and see if the estimate changes? I believe that extended stats will not di=
rectly help with joins though, only group bys and perhaps choosing=C2=A0an =
index scan vs table scan when comparing the correlated columns to static va=
lues rather than joining up tables. Wouldn&#39;t be much effort to try it t=
hough.</div></div></div>

--00000000000075248305b29625cd--