xfs
[Top] [All Lists]

Re: XFS IO multiplication problem on centos/rhel 6 using hp p420i raid c

To: stan@xxxxxxxxxxxxxxxxx
Subject: Re: XFS IO multiplication problem on centos/rhel 6 using hp p420i raid controllers
From: Dennis Kaarsemaker <dennis.kaarsemaker@xxxxxxxxxxx>
Date: Thu, 07 Mar 2013 14:52:56 +0100
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
Dkim-signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=booking.com; s=bk; h=Mime-Version:Content-Transfer-Encoding:Content-Type:References:In-Reply-To:Date:Cc:To:From:Subject:Message-ID; bh=1PU8X6dO3390c3WdTD1cIim1JyEcflpEsA5dO0E51c0=; b=ot370Hc5XAzfePHVFZitqVKE8hO7+19vXkWxw2zj0x3UbWDx1FupPP1zHu4fVKUp9elLfiUqW61gy8p0eQt0mINRB3olqg/Co3ux7g3/MJb++cwLR54/M+fAJ/+hzYRVc8IFPoaY2i2sdcMrPM6fkVncx6SOCkmEu2J742Jaslw=;
In-reply-to: <513899D3.80907@xxxxxxxxxxxxxxxxx>
Organization: Booking.com
References: <1362060736.1247.30.camel@seahawk> <20130228194023.GQ5551@dastard> <1362577992.1247.84.camel@seahawk> <20130307035737.GC6369@dastard> <1362651128.16657.13.camel@seahawk> <513891D7.4060801@xxxxxxxxxxxxxxxxx> <1362663188.16657.20.camel@seahawk> <513899D3.80907@xxxxxxxxxxxxxxxxx>
On Thu, 2013-03-07 at 07:44 -0600, Stan Hoeppner wrote:
> On 3/7/2013 7:33 AM, Dennis Kaarsemaker wrote:
> 
> > The "ten times as many bytes" is constant, so it is constantly writing
> > ten times as much data.
> 
> Where in the IO stack are you counting the bytes?  Are 10x as many bytes
> actually hitting the AGs?  I.e. are you consuming 10x more (real--not
> counting preallocation) space in the filesystem?  Or are you simply
> seeing larger IOs in flight per unit time, but the same total amount of
> IO bytes?

I'm counting them with vmstat, iostat and collectl as can be seen in the
various outputs I've sent. I hope that answers your question as I have
no idea how to see what's hitting the AG's.

I am not using ten times as much space, as that wouldn't fit on the
volume.
-- 
Dennis Kaarsemaker, Systems Architect
Booking.com
Herengracht 597, 1017 CE Amsterdam
Tel external +31 (0) 20 715 3409
Tel internal (7207) 3409

<Prev in Thread] Current Thread [Next in Thread>