DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta;
	h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references;
	b=UmcUSSQr/qPhjOkI/BI+b8m/jDh6gMYHa/dU8hatUwrFLYBobX1Cie7/LsrNlFrkBvtOj8nYZlEarl1SYEM6ux463uqYBojkj1Q28LUyJUuF0/+Bfn8IpdpBeauYtUGKvKn1eUbF0X96ms8AnvMDEM3+DUife1EtKPqGLhORtw4=
Message-ID: <b048c51b0708220802u24d78e15i916bf7c52a303280@mail.gmail.com>
Date: Wed, 22 Aug 2007 19:02:03 +0400
From: "Aleksandr Vinokurov" <aleksandr.vin@gmail.com>
To: "Scott Marlowe" <scott.marlowe@gmail.com>
Subject: Re: Partial index and query plan
Cc: pgsql-sql@postgresql.org
In-Reply-To: <dcc563d10708220743h37f85ae5scb183bb9e6e85073@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
References: <b048c51b0708220557g18cc9a6p7bf2ab195539c3bd@mail.gmail.com>
	<dcc563d10708220743h37f85ae5scb183bb9e6e85073@mail.gmail.com>

Scott Marlowe wrote:
>
> I don't see the issue here.  The index being used is the same partial
> index you created.  Maybe it's just a question of semantics?
>

As I understand final filter is:
	a)  pointed at the index creation
	b)  is redundant as all the indexed records have action <> 0.

So checking of it is a time wasting.

And the plan should be this:

>> Limit  (cost=0.00..3.05 rows=1 width=8) (actual time=4.798..4.798
>> rows=0 loops=1)
>>    ->  Index Scan using indx_date_action12_uid_user_his on
>> user_history  (cost=0.00..6.10 rows=2 width=8) (actual
>> time=4.791..4.791 rows=0 loops=1)
>>          Index Cond: ((date > '2007-08-18 14:33:40.60664'::timestamp
>> without time zone) AND (date <= '2007-08-18 16:30:00'::timestamp
>> without time zone) AND (uid = 138658))


I suggest that this 'Filter' check will not be noticed as it always
return TRUE, and so will be checked only once, -- because of the
"limit 1". :)

But thanks, Aleksandr.