From: Tom Lane <tgl@sss.pgh.pa.us>
To: Ayub Khan <ayub.hp@gmail.com>
cc: Pgsql Performance <pgsql-performance@lists.postgresql.org>
Subject: Re: slow query
In-reply-to: 
 <CAHdeyE+Rhk646X4E2_dH6g-CD_nU8LJ=15XYvb7jjf8gid71Yw@mail.gmail.com>
References: 
 <CAHdeyE+Rhk646X4E2_dH6g-CD_nU8LJ=15XYvb7jjf8gid71Yw@mail.gmail.com>
Comments: In-reply-to Ayub Khan <ayub.hp@gmail.com>
	message dated "Tue, 08 Jun 2021 19:03:48 +0300"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1120999.1623172468.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Tue, 08 Jun 2021 13:14:28 -0400
Message-ID: <1121000.1623172468@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1121000.1623172468%40sss.pgh.pa.us>
Precedence: bulk

Ayub Khan <ayub.hp@gmail.com> writes:
> I checked all the indexes are defined on the tables however the query se=
ems
> slow, below is the plan. Can any one give any pointers to verify ?

You might try to do something about the poor selectivity estimate here:

>                                       ->  Index Scan using
> menu_item_restaurant_id on menu_item a  (cost=3D0.42..177.31 rows=3D1
> width=3D87) (actual time=3D0.130..3.769 rows=3D89 loops=3D1)
>                                             Index Cond: (restaurant_id =3D=
 1528)
> "                                            Filter: ((active =3D
> 'Y'::bpchar) AND (is_menu_item_available(menu_item_id, 'Y'::bpchar) =3D
> 'Y'::bpchar))"
>                                             Rows Removed by Filter: 194

If the planner realized that this'd produce O(100) rows not 1,
it'd likely have picked a different plan.  I'm guessing that
the issue is lack of knowledge about what is_menu_item_available()
will do.  Maybe you could replace that with a status column?

			regards, tom lane