From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 6 Apr 2004 08:07:35 +0100 From: thornber@redhat.com Message-ID: <20040406070735.GA25046@nimzo> References: <20040405212919.GA12443@osdl.org> Mime-Version: 1.0 Content-Disposition: inline In-Reply-To: <20040405212919.GA12443@osdl.org> Subject: [linux-lvm] Re: LVM2 stripe performance Reply-To: LVM general discussion and development List-Id: LVM general discussion and development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , List-Id: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Dave Olien Cc: thornber@redhat.com, linux-lvm@redhat.com On Mon, Apr 05, 2004 at 02:29:19PM -0700, Dave Olien wrote: > > I'm running tiobench on an LVM2 striped logical volume. I'm comparing > the LVM2 striping performance with a similar striping configuration > using a hardware RAID controller. > > With hardware RAID, I'm able to get about 66 megabytes per second on > sequential writes, and 52 megabytes per second with random writes. > > With LVM2, I can get only 23 megabytes per second with either random > or sequential writes. > > Read performance on LVM is about 10% less than with hardware RAID. > e.g. with 48 threads, hardware RAID sequental reads are 105 megabytes > per second, LVM2 RAID sequential reads are 95 megabytes per second. > > This is using an ext2 file system, mounted with noatime. > > I think I recall seeing email with similar problems to this on this > mail list. If I recall, the cause had to do with ordering of requests > submitted to disk. > > Has there been any work to address this? Try the latest -mm kernel - Joe