From: Alessandro Ipe <Alessandro.Ipe@meteo.be>
To: Torsten Zuehlsdorff <mailinglists@toco-domains.de>
Cc: pgsql-performance@postgresql.org
Subject: Re: Excessive memory used for INSERT
Date: Wed, 17 Dec 2014 17:57:51 +0100
Message-ID: <35486133.28l5a59Zd7@snow.oma.be>
Organization: Royal Meteorological Institute of Belgium
User-Agent: KMail/4.14.3 (Linux/3.7.10-1.1-desktop; KDE/4.14.3; x86_64; ; )
In-Reply-To: <5491A0A8.7010506@toco-domains.de>
References: <2163931.4eVnXNjHa3@snow.oma.be>
 <5491A0A8.7010506@toco-domains.de>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="iso-8859-15"
Precedence: bulk
Sender: pgsql-performance-owner@postgresql.org

Hi Torsten,


Thanks for your answer.

I have modified
(SELECT * FROM upsert)
to
(SELECT * FROM upsert WHERE slot=to_timestamp('201212032145', 'YYYYMMDDHH24MI') and MSG=2)
according to your suggestion to reduce the result-set to a single row. However, the INSERT process is still consuming the same amount of RAM.


Regards,


Alessandro.


On Wednesday 17 December 2014 16:26:32 Torsten Zuehlsdorff wrote:
> Hello Alessandro,
> 
> > 2014-12-15 17:54:07 GMT DETAIL: Failed process was running: WITH upsert
> > AS (update MSG set
> > (slot,MSG,HRV,VIS006,VIS008,IR_016,IR_039,WV_062,WV_073,IR_087,IR_097,IR_1
> > 08,IR_120,IR_134,PRO,EPI,CLM,TAPE) = (to_timestamp('201212032145',
> > 'YYYYMMDDHH24MI'),2,'\xffffff','\xff','\xff','\xff','\xff','\xff','\xff','
> > \xff','\xff','\xff','\xff','\xff','\x01','\x01','\x7f','LTO5_020') where
> > slot=to_timestamp('201212032145', 'YYYYMMDDHH24MI') and MSG=2 RETURNING
> > *) insert into MSG
> > (slot,MSG,HRV,VIS006,VIS008,IR_016,IR_039,WV_062,WV_073,IR_087,IR_097,IR_1
> > 08,IR_120,IR_134,PRO,EPI,CLM,TAPE) select to_timestamp('201212032145',
> > 'YYYYMMDDHH24MI'),2,'\xffffff','\xff','\xff','\xff','\xff','\xff','\xff','
> > \xff','\xff','\xff','\xff','\xff','\x01','\x01','\x7f','LTO5_020' WHERE
> > NOT EXISTS (SELECT * FROM upsert);
> 
> How many rows is "(SELECT * FROM upsert)" returning? Without knowing
> more i would guess, that the result-set is very big and that could be
> the reason for the memory usage.
> 
> I would add an WHERE clause to reduce the result-set (an correct index
> can fasten this method even more).
> 
> Greetings,
> Torsten


-- 
Sent via pgsql-performance mailing list (pgsql-performance@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance