From owner-freebsd-fs@FreeBSD.ORG Mon Mar 26 10:09:40 2012 Return-Path: Delivered-To: freebsd-fs@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A0A35106566C for ; Mon, 26 Mar 2012 10:09:40 +0000 (UTC) (envelope-from sven@crashme.org) Received: from celaeno.tauri.mw.lg.virgo.supercluster.net (celaeno.tauri.mw.lg.virgo.supercluster.net [213.252.140.33]) by mx1.freebsd.org (Postfix) with ESMTP id 5A4348FC1E for ; Mon, 26 Mar 2012 10:09:39 +0000 (UTC) Received: from miram.persei.mw.lg.virgo.supercluster.net ([213.252.140.37] helo=[192.168.20.6]) by celaeno.tauri.mw.lg.virgo.supercluster.net with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.76) (envelope-from ) id 1SC6JV-0004OP-VK for freebsd-fs@freebsd.org; Mon, 26 Mar 2012 09:34:11 +0000 Message-ID: <4F703815.8070809@crashme.org> Date: Mon, 26 Mar 2012 11:34:13 +0200 From: Sven Brandenburg User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:11.0) Gecko/20120310 Thunderbird/11.0 MIME-Version: 1.0 To: freebsd-fs@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 2.5 (++) Subject: NFSv3, ZFS, 10GE performance X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 26 Mar 2012 10:09:40 -0000 Hi, I'm toying around with a potential ZFS based Fileserver and I'm curious as to what level of performance can be expected from FreeBSD 9 and ZFS exports via NFS(v3) over 10GE. My current testing shows that while I can easily saturate the 10GE links with ttcp, I 'only' manage about 300MB/s through NFS. In my tests, the exported files should fit into L1ARC entirely (and can be copied to /dev/null on the local machine with several GB/s.. which seems about right reading from L1). I would expect something in the region of ~900MB/s via NFS in my test case. What can be expected for 10GE in terms of NFS throughput for a 9-STABLE client+server? Are there sysctl's or similar knobs to fiddle with? If helpful, I can provide more details. Thanks :) Sven