Date: Mon, 24 Jan 2011 21:07:36 +0100
From: hubert depesz lubaczewski <depesz@depesz.com>
To: Dimi Paun <dimi@lattica.com>
Cc: "pgsql-performance@postgresql.org" <pgsql-performance@postgresql.org>
Subject: Re: How to use indexes for GROUP BY
Message-ID: <20110124200735.GA23732@depesz.com>
Reply-To: depesz@depesz.com
References: <1295893741.9380.984.camel@dimi.lattica.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <1295893741.9380.984.camel@dimi.lattica.com>
User-Agent: Mutt/1.5.21 (2010-09-15)

On Mon, Jan 24, 2011 at 01:29:01PM -0500, Dimi Paun wrote:
> Hi folks,
> 
> I have a table like so:
> 
> create table tagRecord (
>     uid             varchar(60) primary key,
>     [bunch of other fields]
>     location      varchar(32),
>     creationTS    timestamp
> );
> create index idx_tagdata_loc_creationTS on tagRecord(location, creationTS);
> 
> The number of individual values in location is small (e.g. 2).
> 
> I want to simply get the latest "creationTS" for each location,
> but that seems to result in a full table scan:
> 
> tts_server_db=# explain analyze select location, max(creationTS) from tagrecord group by location;
>                                                        QUERY PLAN                                                        
> -------------------------------------------------------------------------------------------------------------------------
>  HashAggregate  (cost=5330.53..5330.55 rows=2 width=18) (actual time=286.161..286.165 rows=3 loops=1)
>    ->  Seq Scan on tagrecord  (cost=0.00..4771.35 rows=111835 width=18) (actual time=0.059..119.828 rows=111739 loops=1)
>  Total runtime: 286.222 ms

you can use technique described in here:
http://www.depesz.com/index.php/2009/07/10/getting-list-of-unique-elements/

Best regards,

depesz