From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
---|---|
To: | Hitoshi Harada <umi(dot)tanuki(at)gmail(dot)com> |
Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Pull up aggregate subquery |
Date: | 2011-05-25 16:20:44 |
Message-ID: | BANLkTik=vohzbxFM1k=Oe4ChgW9xA3DgEQ@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Wed, May 25, 2011 at 10:35 AM, Hitoshi Harada <umi(dot)tanuki(at)gmail(dot)com> wrote:
> 2011/5/25 Hitoshi Harada <umi(dot)tanuki(at)gmail(dot)com>:
>> So I'm still
>> thinking which of pulling up and parameterized scan is better.
>
> After more investigation I came up with third idea, pushing down
> RangeTblEntry to aggregate subquery. This sounds like a crazy idea,
> but it seems to me like it is slightly easier than pulling up
> agg-subquery. The main point is that when you want to pull up, you
> must care if the final output would be correct with other join
> relations than the aggregate-related one. In contrast, when pushing
> down the target relation to agg-subquery it is simple to ensure the
> result.
>
> I'm looking around pull_up_subqueries() in subquery_planner() to add
> the pushing down logic. It could be possible to do it around
> make_one_rel() but I bet query structure changes are doable against
> Query, not PlannerInfo.
How do you decide whether or not to push down?
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2011-05-25 16:23:56 | Re: [ADMIN] pg_class reltuples/relpages not updated by autovacuum/vacuum |
Previous Message | Robert Haas | 2011-05-25 16:18:53 | Re: New/Revised TODO? Gathering actual read performance data for use by planner |