From owner-freebsd-performance@FreeBSD.ORG Sun Mar 24 19:58:50 2013 Return-Path: Delivered-To: freebsd-performance@freebsd.org Received: from mx1.freebsd.org (mx1.FreeBSD.org [8.8.178.115]) by hub.freebsd.org (Postfix) with ESMTP id D5B43109 for ; Sun, 24 Mar 2013 19:58:50 +0000 (UTC) (envelope-from daniel.bilik@neosystem.cz) Received: from mail.neosystem.cz (mail.neosystem.cz [94.23.169.88]) by mx1.freebsd.org (Postfix) with ESMTP id A05331D0 for ; Sun, 24 Mar 2013 19:58:50 +0000 (UTC) Received: from mail.neosystem.cz (unknown [127.0.10.15]) by mail.neosystem.cz (Postfix) with ESMTP id 65619476D; Sun, 24 Mar 2013 20:58:47 +0100 (CET) X-Virus-Scanned: amavisd-new at mail.neosystem.cz Received: from leon.sn.neosystem.cz (unknown [IPv6:2001:41d0:2:5ab8::60:14]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mail.neosystem.cz (Postfix) with ESMTPSA id 3578F4767; Sun, 24 Mar 2013 20:58:45 +0100 (CET) Date: Sun, 24 Mar 2013 20:55:45 +0100 From: Daniel Bilik To: Adam Vande More Subject: Re: FreeBSD 9.1 vs CentOS 6.3 Message-Id: <20130324205545.b357801d4b13fdb8024a9423@neosystem.cz> In-Reply-To: References: <514C1E5F.8040504@contactlab.com> <20130323213406.93cc3baddf69d5d71f10365e@neosystem.cz> Organization: neosystem.cz X-Mailer: Sylpheed 3.2.0 (GTK+ 2.24.12; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Mailman-Approved-At: Sun, 24 Mar 2013 20:45:13 +0000 Cc: freebsd-performance@freebsd.org X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 24 Mar 2013 19:58:50 -0000 On Sun, 24 Mar 2013 11:01:05 -0500 Adam Vande More wrote: > These are interesting results. Did you try tuning any of the jemalloc > options in /etc/malloc.conf? No tuning, jemalloc was tested "out of the box" just for curiosity. > I think increasing the number of arenas may help the contention, eg "ln > -s 3N /etc/malloc.conf" Luckily the test system is still running, so I could just apply suggested tuning, restart mysqld with jemalloc and launch benchmark sets. Results for read-only transactions are practically identical to previous ones. Read-write transactions test is still running but from numbers for 1 to 16 threads I can tell there is some (very) small improvement and the results are more stable. -- Daniel Bilik neosystem.cz