From mboxrd@z Thu Jan  1 00:00:00 1970
From: Vladimir Bashkirtsev <vladimir@bashkirtsev.com>
Subject: Re: Poor read performance in KVM
Date: Fri, 20 Jul 2012 14:54:30 +0930
Message-ID: <5008EB8E.30708@bashkirtsev.com>
References: <5002C215.108@bashkirtsev.com>	<5003B1CC.4060909@inktank.com>	<50064DCD.8040904@bashkirtsev.com>	<5006D5FB.8030700@inktank.com>	<5007FB6E.1080307@bashkirtsev.com>	<CADvuQRGaivtgho4R=rk3c_B6eJ2GfG=tZFGNy9knoKyh2CR-zg@mail.gmail.com> <CADxhoDT8rimSwP+Y75tGVo3Nd5WYw8xHyDxBAMV8tb5no0OP4Q@mail.gmail.com> <50084EC3.60003@inktank.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Return-path: <ceph-devel-owner@vger.kernel.org>
Received: from mail.logics.net.au ([150.101.56.178]:54221 "EHLO
	mail.logics.net.au" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1751306Ab2GTFYo (ORCPT
	<rfc822;ceph-devel@vger.kernel.org>); Fri, 20 Jul 2012 01:24:44 -0400
In-Reply-To: <50084EC3.60003@inktank.com>
Sender: ceph-devel-owner@vger.kernel.org
List-ID: <ceph-devel.vger.kernel.org>
To: Mark Nelson <mark.nelson@inktank.com>
Cc: Calvin Morrow <calvin.morrow@gmail.com>, Tommi Virtanen <tv@inktank.com>, Josh Durgin <josh.durgin@inktank.com>, ceph-devel@vger.kernel.org


>
> We are seeing degradation at 64k node/leaf sizes as well.  So far the 
> degradation is most obvious with small writes.  it affects XFS as 
> well, though not as severely.  We are vigorously looking into it. :)
>
Just confirming that one of our clients has run fair amount (on 
gigabytes scale) of mysql updates in preceding few days. So it appears 
all performance degradation stems from random writes. Perhaps I will 
rebuild each osd to check if it would help to recover.