Re: bad execution plan for subselects containing windowing-function

From: Andreas Kretschmer <akretschmer(at)spamfence(dot)net>
To: pgsql-performance(at)postgresql(dot)org
Subject: Re: bad execution plan for subselects containing windowing-function
Date: 2010-01-14 18:31:39
Message-ID: 20100114183139.GA11334@tux
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Andreas Kretschmer <akretschmer(at)spamfence(dot)net> wrote:
> > they are being done over all rows. In this particular example you
> > happen to get the same result, but that's just because "avg(foo) over
> > partition by foo" is a dumb example --- it will necessarily just yield
> > identically foo. In more realistic computations the results would be
> > different.
>
> Okay, i believe you now ;-)
>
> I will try to find a case with different results ...

I have got it!

test=# select * from values;
id | value
----+-------
1 | 10
2 | 20
3 | 30
4 | 40
5 | 50
6 | 60
7 | 70
8 | 80
9 | 90
(9 rows)

Time: 0.240 ms
test=*# select id, sum(value) over (order by id) from values where id = 5;
id | sum
----+-----
5 | 50
(1 row)

Time: 0.352 ms
test=*# select * from (select id, sum(value) over (order by id) from values) foo where id = 5;
id | sum
----+-----
5 | 150
(1 row)

Time: 0.383 ms

Andreas
--
Really, I'm not out to destroy Microsoft. That will just be a completely
unintentional side effect. (Linus Torvalds)
"If I was god, I would recompile penguin with --enable-fly." (unknown)
Kaufbach, Saxony, Germany, Europe. N 51.05082°, E 13.56889°

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Kevin Grittner 2010-01-14 18:33:50 Re: Slow "Select count(*) ..." query on table with 60 Mio. rows
Previous Message Greg Smith 2010-01-14 18:30:46 Re: a heavy duty operation on an "unused" table kills my server