From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mx3.redhat.com (mx3.redhat.com [172.16.48.32])
	by int-mx1.corp.redhat.com (8.13.1/8.13.1) with ESMTP id l8OGYmLB024898
	for <linux-lvm@redhat.com>; Mon, 24 Sep 2007 12:34:48 -0400
Received: from shuttlerun.csg.stercomm.com (scidalsmtp02.csg.stercomm.com
	[204.214.3.25])
	by mx3.redhat.com (8.13.1/8.13.1) with ESMTP id l8OGYkUH012521
	for <linux-lvm@redhat.com>; Mon, 24 Sep 2007 12:34:46 -0400
Message-ID: <46F7E720.4060304@stercomm.com>
Date: Mon, 24 Sep 2007 11:34:40 -0500
From: Chris Cox <chris_cox@stercomm.com>
MIME-Version: 1.0
Subject: Re: [linux-lvm] LVM Overhead
References: <1C8CF1EA1A5B5940B81B0710B2A4C9381DF6018911@an-ex.ActiveNetwerx.int>	<46F7D242.3030707@stercomm.com>
	<46F7D49A.2090206@wpkg.org>
In-Reply-To: <46F7D49A.2090206@wpkg.org>
Content-Transfer-Encoding: 7bit
Reply-To: LVM general discussion and development <linux-lvm@redhat.com>
List-Id: LVM general discussion and development <linux-lvm.redhat.com>
List-Unsubscribe: <https://www.redhat.com/mailman/listinfo/linux-lvm>,
	<mailto:linux-lvm-request@redhat.com?subject=unsubscribe>
List-Archive: <https://www.redhat.com/archives/linux-lvm>
List-Post: <mailto:linux-lvm@redhat.com>
List-Help: <mailto:linux-lvm-request@redhat.com?subject=help>
List-Subscribe: <https://www.redhat.com/mailman/listinfo/linux-lvm>,
	<mailto:linux-lvm-request@redhat.com?subject=subscribe>
List-Id: <linux-lvm.redhat.com>
Content-Type: text/plain; charset="us-ascii"
To: LVM general discussion and development <linux-lvm@redhat.com>

Tomasz Chmielewski wrote:
> Chris Cox schrieb:
>> Joseph L. Casale wrote:
>>> Hi,
>>> I am wanting to use LVM behind an iet iscsi target and I am wondering
>>> how much overhead if any does LVM add? Would the performance degrade
>>> any, currently I am exporting full unpartitioned discs and am happy with
>>> the performance. From what I am reading now, I would likely partition
>>> the physical discs and use that in LVM then export a volume.
>>
>> Overhead is negligible.  I haven't seen any impact at all.
> 
> Quite the contrary - unless the default settings are not changed.
> 
> 
> By default, the readahead values for LVM volumes is quite low / for
> iSCSI, I always have to change it to get acceptable performance (one
> might want to replace the $HOSTNAME with the name of the PV, or define a
> separate list; "setra" values might need some experimenting, too):
> 
> 
> LVMS=$(/bin/ls /dev/$HOSTNAME/)
> 
> for LVM in $LVMS
>     do
>         blockdev --setra 16384 /dev/$HOSTNAME/$LVM
>     done
> 
> 

Hmmm, perhaps for iSCSI.  But using fibre SAN, I saturate my
2Gbps link (almost 200MB/sec).


-- 
Chris Cox
Sr. Unix Sys Admin