From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1MT9Nq-00021C-EA for qemu-devel@nongnu.org; Tue, 21 Jul 2009 03:03:30 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1MT9Nl-0001zx-Cu for qemu-devel@nongnu.org; Tue, 21 Jul 2009 03:03:30 -0400 Received: from [199.232.76.173] (port=57550 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1MT9Nl-0001zT-6t for qemu-devel@nongnu.org; Tue, 21 Jul 2009 03:03:25 -0400 Received: from mx20.gnu.org ([199.232.41.8]:56255) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1MT9Nk-0002GK-IY for qemu-devel@nongnu.org; Tue, 21 Jul 2009 03:03:24 -0400 Received: from fwil.voltaire.com ([193.47.165.2] helo=exil.voltaire.com) by mx20.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1MT9Nj-0007wK-07 for qemu-devel@nongnu.org; Tue, 21 Jul 2009 03:03:23 -0400 Message-ID: <4A656824.7070100@Voltaire.com> Date: Tue, 21 Jul 2009 10:03:00 +0300 From: Or Gerlitz MIME-Version: 1.0 Subject: Re: [Qemu-devel] [PATCH] net: add raw backend - some performance measurements References: <20090701162115.GA4555@shareable.org> <4A4CA747.1050509@Voltaire.com> <20090703023911.GD938@shareable.org> <4A534EC4.5030209@voltaire.com> <20090707145739.GB14392@shareable.org> <4A54B0F1.3070201@voltaire.com> <20090715203806.GF3056@shareable.org> <4A647B72.5090404@Voltaire.com> <20090720155308.GA9327@gondor.apana.org.au> In-Reply-To: <20090720155308.GA9327@gondor.apana.org.au> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Herbert Xu Cc: Mark McLoughlin , "Michael S. Tsirkin" , Dor Laor , qemu-devel@nongnu.org, Jan Kiszka Herbert Xu wrote: > Is netfilter enabled on the bridge? If so you need to turn it off okay, when setting net.bridge.bridge-nf-call-iptables to zero, the VM TX / tap+bridge packet rate climbs from 170K to 195K but it still way beyond the 240K rate achieved by the raw mode --> we have now a clear sign on the performance gain this approach provides. Or.