From: Korisk <Korisk@yandex.ru>
To: pgsql-performance@postgresql.org
Subject: hash aggregation speedup
MIME-Version: 1.0
Message-Id: <739641349631199@web11d.yandex.ru>
Date: Sun, 07 Oct 2012 21:33:19 +0400
Content-Transfer-Encoding: 7bit
Content-Type: text/plain

I have table:
create table hashcheck(id serial, name varchar, value varchar);
and query:
hashaggr=# explain  analyse verbose select name, count(name) as cnt from hashcheck group by name order by name desc;
                                                                  QUERY PLAN                                                                  
----------------------------------------------------------------------------------------------------------------------------------------------
 Sort  (cost=471979.50..471985.40 rows=2362 width=9) (actual time=19642.938..19643.184 rows=4001 loops=1)
   Output: name, (count(name))
   Sort Key: hashcheck.name
   Sort Method: quicksort  Memory: 343kB
   ->  HashAggregate  (cost=471823.53..471847.15 rows=2362 width=9) (actual time=19632.256..19632.995 rows=4001 loops=1)
         Output: name, count(name)
         ->  Seq Scan on public.hashcheck  (cost=0.00..363494.69 rows=21665769 width=9) (actual time=49.552..11674.170 rows=23103672 loops=1)
               Output: id, name, value
 Total runtime: 19643.497 ms
(9 rows)

without indexes. 
Indexes don't speedup the query much
For hash  Total runtime: 17678.225 ms
For btree Total runtime: 14188.484 ms
I'm don't know how to use Gin and gist this way. 

So the question is there any way to speed up the "group by" query? Or may be there does exists any other way to count histogram?

Thank you.


PS:
hashaggr=# select version();
                                                   version                                                   
-------------------------------------------------------------------------------------------------------------
 PostgreSQL 9.1.6 on x86_64-redhat-linux-gnu, compiled by gcc (GCC) 4.7.2 20120921 (Red Hat 4.7.2-2), 64-bit