Subject: Re: DISTINCT Question
From: Scott Marlowe <smarlowe@g2switchworks.com>
To: Y Sidhu <ysidhu@gmail.com>
Cc: pgsql-performance@postgresql.org
In-Reply-To: <b09064f30705081252i27df3236hb12c13b923d0f45f@mail.gmail.com>
References: <b09064f30705081252i27df3236hb12c13b923d0f45f@mail.gmail.com>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Message-Id: <1178656101.7497.29.camel@state.g2switchworks.com>
Mime-Version: 1.0
Date: Tue, 08 May 2007 15:28:22 -0500

On Tue, 2007-05-08 at 14:52, Y Sidhu wrote:
> Does using DISTINCT in a query force PG to abandon any index search it
> might have embarked upon?

 explain analyze select distinct request from businessrequestsummary
where lastflushtime between now() - interval '30 minutes' and now();
                                                                                   QUERY PLAN                                                                                                                                                     
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 Unique  (cost=3.04..3.04 rows=1 width=17) (actual time=110.565..162.630
rows=75 loops=1)
   ->  Sort  (cost=3.04..3.04 rows=1 width=17) (actual
time=110.555..135.252 rows=6803 loops=1)
         Sort Key: request
         ->  Index Scan using businessrequestsummary_lastflushtime_dx on
businessrequestsummary  (cost=0.01..3.03 rows=1 width=17) (actual
time=0.063..59.674 rows=6803 loops=1)
               Index Cond: ((lastflushtime >= (now() -
'00:30:00'::interval)) AND (lastflushtime <= now()))
 Total runtime: 163.925 ms
(6 rows)

I'd say no.