Re: Subquery in a JOIN not getting restricted?

From: Merlin Moncure <mmoncure(at)gmail(dot)com>
To: Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov>
Cc: Jay Levitt <jay(dot)levitt(at)gmail(dot)com>, pgsql-performance(at)postgresql(dot)org
Subject: Re: Subquery in a JOIN not getting restricted?
Date: 2011-11-09 19:50:28
Message-ID: CAHyXU0zWLRquK9xu4SbTrxyOTA2kbT1b3r4_H4KdEzE-hvVgoA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On Wed, Nov 9, 2011 at 9:15 AM, Kevin Grittner
<Kevin(dot)Grittner(at)wicourts(dot)gov> wrote:
> Jay Levitt <jay(dot)levitt(at)gmail(dot)com> wrote:
>
>> I don't get why the GROUP BY in this subquery forces it to scan
>> the entire users table (seq scan here, index scan on a larger
>> table) when there's only one row in users that can match:
>
>> explain analyze
>> select questions.id
>> from questions
>> join (
>>    select u.id
>>    from users as u
>>    group by u.id
>> ) as s
>> on s.id = questions.user_id
>> where questions.id = 1;
>
>>   Total runtime: 1.262 ms
>
> Are you sure there's a plan significantly faster than 1.3 ms?

Well, this may not fit the OP's 'real' query, but the inner subquery
is probably better written as a semi-join (WHERE EXISTS).

merlin

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Kevin Grittner 2011-11-09 20:20:00 Re: Subquery in a JOIN not getting restricted?
Previous Message kzsolt 2011-11-09 19:48:13 Re: Heavy contgnous load