Date: Fri, 9 Jun 2017 09:53:47 -0500
From: Kenneth Marshall <ktm@rice.edu>
To: Frits Jalvingh <jal@etc.to>
Cc: Scott Marlowe <scott.marlowe@gmail.com>,
	Andreas Kretschmer <andreas@a-kretschmer.de>,
	"pgsql-performance@postgresql.org" <pgsql-performance@postgresql.org>
Subject: Re: Improving PostgreSQL insert performance
Message-ID: <20170609145347.GE26506@aart.rice.edu>
References: 
 <CAKhTGFX-ChBSjqENrAv8uqUR_H5PTvf14jG0cOAq6c-ami_7sQ@mail.gmail.com>
 <9e2a0a34-dd83-43af-d41a-791b34ff76e1@a-kretschmer.de>
 <20170609132857.GC26506@aart.rice.edu>
 <CAKhTGFVTvFxsfngDNgQV-ciWYFn6aXROwGj6s4Gqkqa0bTusJw@mail.gmail.com>
 <CAOR=d=1CYLvtesjcCmC95rWZSzJ02pYZyfJ8SA9jfx-CCekx3A@mail.gmail.com>
 <CAKhTGFXiU_t_EiB37q-H897pFqQY9a=hDAuAZV7T3dLMTsvSiw@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CAKhTGFXiU_t_EiB37q-H897pFqQY9a=hDAuAZV7T3dLMTsvSiw@mail.gmail.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Precedence: bulk
Sender: pgsql-performance-owner@postgresql.org

On Fri, Jun 09, 2017 at 02:39:37PM +0000, Frits Jalvingh wrote:
> Hi all,
> 
> Thanks a lot for the many responses!
> 
> About preparing statements: this is done properly in Java, and pgsql does
> it by itself. So that cannot be done better ;)
> 
> I tried the copy command, and that indeed works quite brilliantly:
> Inserted 24000000 rows in 22004 milliseconds, 1090710.7798582076 rows per
> second
> 
> That's faster than Oracle. But with a very bad interface I have to say for
> normal database work.. I will try to make this work in the tooling, but it
> needs some very special code to format all possible values properly, and to
> manage the end of the copy, so it is not usable in general which is a pity,
> I think.
> 
> So, I am still very interested in getting normal inserts faster, because
> that will gain speed for all work.. If Oracle can do it, and Postgres is
> able to insert fast with copy- where lies the bottleneck with the insert
> command? There seems to be quite a performance hit with the JDBC driver
> itself (as the stored procedure is a lot faster), so I can look into that.
> But even after that there is quite a gap..
> 
> Regards,
> 
> Frits

Hi Frits,

Have you looked at UNLOGGED tables and also having more that 1 insert
stream running at a time. Sometimes multiple parallel inserts can be
faster.

Regards,
Ken


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