From: Samed YILDIRIM <samed@reddoc.net>
To: Lazaro Garcia <lazaro3487@gmail.com>
To: pgsql-admin@postgresql.org <pgsql-admin@postgresql.org>
Subject: Re: Help with bad performance with synchronous commit
Date: Fri, 31 Mar 2017 19:36:44 +0300
Message-ID: <6346761490978204@web56g.yandex.ru> (raw)
In-Reply-To: <001f01d2aa2d$1f9f27a0$5edd76e0$@gmail.com>
References: <001f01d2aa2d$1f9f27a0$5edd76e0$@gmail.com>
List-Unsubscribe: <mailto:majordomo@postgresql.org?body=unsub%20pgsql-admin>
Hi Lazaro,<div><br /></div><div>Please initialize your test database with scale factor 50 by using pgbench than perform test again and share results to us?</div><div><br /></div><div>Best regards.<br /><br />17:45, 31 Mart 2017, Lazaro Garcia <lazaro3487@gmail.com>:<br /><blockquote><div lang="ES" link="#0563C1" vlink="#954F72"><div><p>Hello everyone</p><p> </p><p><span lang="EN-US">I wite because I have a great problem and I have no idea how resolve this:</span></p><p> </p><p><span lang="EN-US">This is the situation:</span></p><p><span lang="EN-US"> </span></p><p><span lang="EN-US">I have installed a postgresql server 9.6.2 on Ubuntu server 16.04.2 LTS with this specifications:</span></p><p><span lang="EN-US"> </span></p><p>12 cores y 30 GB de RAM.</p><p> </p><p><span lang="EN-US">I have created a database with pgbench for testing and the performance are very bad with synchronous commit. These are the results:</span></p><p><span lang="EN-US"> </span></p><p>scaling factor: 1</p><p><span lang="EN-US">query mode: simple</span></p><p><span lang="EN-US">number of clients: 50</span></p><p><span lang="EN-US">number of threads: 12</span></p><p><span lang="EN-US">duration: 30 s</span></p><p><span lang="EN-US">number of transactions actually processed: 2773</span></p><p><span lang="EN-US">latency average = 544.336 ms</span></p><p><span lang="EN-US">latency stddev = 617.941 ms</span></p><p><span lang="EN-US">tps = 91.180927 (including connections establishing)</span></p><p><span lang="EN-US">tps = 91.194080 (excluding connections establishing)</span></p><p><span lang="EN-US"> </span></p><p><span lang="EN">By analyzing the postgres log with pgbadger I could see that the updates take a long time for a table with 10 tuples only. Then I ran a explain analyze and the explain results contradict what the test throws:</span></p><p><span lang="EN-US"> </span></p><p><span lang="EN-US">Update on pgbench_tellers (cost=4.14..8.16 rows=1 width=358) (actual time=0.021..0.021 rows=0 loops=1)</span></p><p><span lang="EN-US"> -> Bitmap Heap Scan on pgbench_tellers (cost=4.14..8.16 rows=1 width=358) (actual time=0.010..0.010 rows=1 loops=1)</span></p><p><span lang="EN-US"> Recheck Cond: (tid = 5)</span></p><p><span lang="EN-US"> Heap Blocks: exact=1</span></p><p><span lang="EN-US"> -> Bitmap Index Scan on pgbench_tellers_pkey (cost=0.00..4.14 rows=1 width=0) (actual time=0.005..0.005 rows=1 loops=1)</span></p><p><span lang="EN-US"> Index Cond: (tid = 5)</span></p><p><span lang="EN-US">Planning time: 0.052 ms</span></p><p>Execution time: 0.043 ms</p><p> </p><p><span lang="EN-US">I have runned the same test in other machines with less RAM and CPU and the results are always better than this:</span></p><p><span lang="EN-US"> </span></p><p><span lang="EN-US">Any help to identify the problem could be very appreciated.</span></p><p><span lang="EN-US"> </span></p><p><span lang="EN-US">Regards</span></p></div></div></blockquote></div>
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: pgsql-admin@postgresql.org
Cc: samed@reddoc.net, lazaro3487@gmail.com
Subject: Re: Help with bad performance with synchronous commit
In-Reply-To: <6346761490978204@web56g.yandex.ru>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox