Message-ID: <4D4D0F14.7020500@2ndquadrant.com>
Date: Sat, 05 Feb 2011 03:49:24 -0500
From: Greg Smith <greg@2ndquadrant.com>
User-Agent: Thunderbird 2.0.0.24 (X11/20100317)
MIME-Version: 1.0
To: Tobias Brox <tobixen@gmail.com>
CC: pgsql-performance@postgresql.org
Subject: Re: table partitioning and select max(id)
References: <AANLkTi=pzBtH4i_WYrbR9JEyq+49vPPVJf-q08z_Bhu=@mail.gmail.com>
	<4D4CDA0C.20000@2ndquadrant.com>
	<AANLkTimU90yx4NEvHRE+otSDjWCDmG5Zjyhxecr-AGZ5@mail.gmail.com>
In-Reply-To: <AANLkTimU90yx4NEvHRE+otSDjWCDmG5Zjyhxecr-AGZ5@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Tobias Brox wrote:
> I did test "select id from table order by id desc limit 1" on my parent table
> yesterday, it would still do the seq-scan.  Even adding a
> where-restriction to make sure only one partition was queried I still
> got the seq-scan.
>   

Right; you actually have to direct the query toward the specific 
partition by name, nothing run against the parent table will work.  The 
new logic for 9.1 essentially splits the query into this alternate form, 
runs it against every partition individually, then combines the 
results.  If you can afford to wait for 9.1, that is certainly the easy 
path here.  It just works out of the box in that version.

-- 
Greg Smith   2ndQuadrant US    greg@2ndQuadrant.com   Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support  www.2ndQuadrant.us
"PostgreSQL 9.0 High Performance": http://www.2ndQuadrant.com/books