Date: Fri, 2 Nov 2007 08:56:04 +0100
From: hubert depesz lubaczewski <depesz@depesz.com>
To: Palle Girgensohn <girgen@pingpong.net>
Cc: pgsql-performance@postgresql.org
Subject: Re: select max(field) from table much faster with a group by clause?
Message-ID: <20071102075604.GA16757@depesz.com>
Reply-To: depesz@depesz.com
References: <D26563C7B9CCEB0A338FF08E@rambutan.pingpong.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <D26563C7B9CCEB0A338FF08E@rambutan.pingpong.net>
User-Agent: Mutt/1.5.13 (2006-08-11)

On Thu, Nov 01, 2007 at 02:07:55PM +0100, Palle Girgensohn wrote:
> I have a table "login" with approx 600,000 tuples, a person table with 
> approx 100000 tuples.
> When running
> select max("when") from login where userid='userid'
> it takes a second or two, but when adding "group by userid" the planner 
> decides on using another plan, and it gets *much* faster. See example below.
> pp=# explain analyze SELECT max("when") FROM login WHERE userid='girgen' ;

just do:
create index q on login (userid, "when"); and you should be fine.
if it will not help, rewrite the query as:
select "when"
from login
where userid = 'girgen'
order by userid desc, "when" desc limit 1;

depesz

-- 
quicksil1er: "postgres is excellent, but like any DB it requires a
highly paid DBA.  here's my CV!" :)
http://www.depesz.com/ - blog dla ciebie (i moje CV)