Message-ID: <42063FF5.9040707@aeccom.com>
Date: Sun, 06 Feb 2005 17:04:05 +0100
From: Dirk.Lutzebaeck@t-online.de (Dirk Lutzebaeck)
Reply-To: lutzeb@aeccom.com
User-Agent: Mozilla Thunderbird 1.0 (X11/20041206)
MIME-Version: 1.0
To: John A Meinel <john@arbash-meinel.com>
Cc: Greg Stark <gsstark@mit.edu>, pgsql-performance@postgresql.org,
	tgl@sss.pgh.pa.us
Subject: Re: query produces 1 GB temp file
References: <20050205182117.0125074@mail2.aeccom.com>
	<87y8e2kb8j.fsf@stark.xeocode.com> <42061B35.6000306@aeccom.com>
	<4206356C.9060308@arbash-meinel.com>
In-Reply-To: <4206356C.9060308@arbash-meinel.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

John,

I'm doing VACUUM ANALYZE once a night. Before the tests I did VACUUM and 
then ANALYZE.

Dirk

John A Meinel wrote:

> Dirk Lutzebaeck wrote:
>
>> Greg,
>>
>> Thanks for your analysis. But I dont get any better after bumping 
>> STATISTICS target from 10 to 200.
>> explain analyze shows that the optimizer is still way off estimating 
>> the rows. Is this normal? It still produces a 1 GB temp file.
>> I simplified the query a bit, now only two tables are involved (bi, 
>> df). I also vacuumed.
>
>
>
> Are you just doing VACUUM? Or are you doing VACUUM ANALYZE? You might 
> also try VACUUM ANALYZE FULL (in the case that you have too many dead 
> tuples in the table).
>
> VACUUM cleans up, but doesn't adjust any planner statistics without 
> ANALYZE.
>
> John
> =:->
>