From owner-freebsd-chat@FreeBSD.ORG Mon Aug 2 15:08:31 2010 Return-Path: Delivered-To: freebsd-chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 642A3106564A for ; Mon, 2 Aug 2010 15:08:31 +0000 (UTC) (envelope-from bounce-4833960-378458059@online.networkworld.com) Received: from online.networkworld.com (online.networkworld.com [66.186.127.191]) by mx1.freebsd.org (Postfix) with SMTP id 1BF238FC14 for ; Mon, 2 Aug 2010 15:08:30 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; q=dns/txt; l=7725; d=online.networkworld.com;s=2008; h=from; bh=mZ8pf0iwmGLZrDzRqYEBnoQrCGRQyow2frCrxpB1OiQ=; b=ids9E51KhUvh4XQkDAz0RCHfwHuB0zDnkRpHCxxwWT1c+q3ElhxaixJoOKi5ZGtFSpHPhas3 mRzYk/LCDH8RqA==; DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=2008; d=online.networkworld.com; h=from; b=AEBKoh1F+3Pv9GvPr7yduCEl34MucBiQ7n4UVt9P7qATR2UPpHlAhZFYuu6PpNHlxCHlBR4c Vi/NzFc5LiaKLw== From: "Network World Online Resources" To: freebsd-chat@freebsd.org Date: Mon, 02 Aug 2010 10:25:00 -0400 MIME-Version: 1.0 Message-ID: X-campaignid: idg4833960 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable X-Content-Filtered-By: Mailman/MimeDel 2.1.5 Subject: 60 Minutes with Security Visionary Nir Zuk X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Aug 2010 15:08:31 -0000 60 Minutes with Security Visionary Nir Zuk Join this live webcast with Nir Zuk, network security visionary and founder= and CTO of Palo Alto Networks, for a live interactive online interview wit= h security analyst Mark Bouchard, CISSP. He was one of the developers of st= ateful inspection firewall technology, worked on the world's first intrusio= n prevention and detection appliance and is now pioneering next-generation = firewalls. Watch Nir speak about the state of the information security indu= stry and then pose your toughest questions to him. View now! = http://online.networkworld.com/t/4833960/378458059/118762/0/?3e076d18=3DZnJ= lZWJzZC1jaGF0QGZyZWVic2Qub3Jn&x=3D4e09679b= Smart Data Center Decision-Making: the Next Generation=20 Best practices for building and operating the enterprise data center of tom= orrow. Learn more. = http://online.networkworld.com/t/4833960/378458059/118586/0/?3e076d18=3DZnJ= lZWJzZC1jaGF0QGZyZWVic2Qub3Jn&x=3Db0383743= =20 SUBSCRIPTION SERVICES You are currently subscribed as=20= freebsd-chat@freebsd.org= . If you do not wish to receive future mailings from Network World Online R= esources, need to change your email or other preference, please visit: =20= http://online.networkworld.com/t/4833960/378458059/22644/0/?641c2725=3DWTdi= NzFzMXR3S0xxRkFudkVnJTJmJTJiRW1mbERzeWp2UiUyYndsNGZIdldEc0pwQSUzZA%3d%3d&x= =3Db03531ea= . If the above URL is not enabled as a link, please copy it in to your browse= r window to access our Subscription Page. View Network World's online privacy policy.=20 Copyright 2010 | Network World | 492 Old Connecticut Path | Framingham MA = 01701 | www.networkworld.com From owner-freebsd-chat@FreeBSD.ORG Mon Aug 2 16:06:07 2010 Return-Path: Delivered-To: freebsd-chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id DDB311065670 for ; Mon, 2 Aug 2010 16:06:07 +0000 (UTC) (envelope-from jdixon@omniti.com) Received: from edge.omniti.com (smtp.omniti.com [8.8.38.6]) by mx1.freebsd.org (Postfix) with ESMTP id 986DB8FC18 for ; Mon, 2 Aug 2010 16:06:06 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; d=omniti.com; s=s1024; c=relaxed/relaxed; q=dns/txt; i=@omniti.com; t=1280765165; h=From:Subject:Date:To; bh=ZBUdPoTkYjXTQ+PfVQi4OWGgYV50VdOXaTT+oMwJw5E=; b=LyUZ+I4yH5xM97ireH7AYuI39rmuTDLE4Ys+wNmx1TwE3PYHBDHUB6h2D8z0hnzX XCDqndJhw/l+KKmkR0tf/BP2zbWwSEqL3R9KLV+OEFq0lXdINKS/SdnpJNGj13fJ 4H0WcczxBzM9T5HJ5BnWVu5TnEPZ8ta0eEmofbFSkqs=; Authentication-Results: edge smtp.user=jdixon@omniti.com; auth=pass (LOGIN) Received: from [68.55.0.29] ([68.55.0.29:60044] helo=omniti.com) by edge (envelope-from ) (ecelerity 2.2.3.46 r(37468M)) with ESMTPSA (cipher=AES256-SHA) id AB/F9-03560-DECE65C4; Mon, 02 Aug 2010 12:06:05 -0400 Date: Mon, 2 Aug 2010 12:06:02 -0400 From: Jason Dixon To: freebsd-chat@freebsd.org Message-ID: <20100802160602.GQ4578@omniti.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.5.18 (2008-05-17) Subject: Register now for Surge 2010 X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Aug 2010 16:06:07 -0000 Registration for Surge Scalability Conference 2010 is open for all attendees! We have an awesome lineup of leaders from across the various communities that support highly scalable architectures, as well as the companies that implement them. Here's a small sampling from our list of speakers: John Allspaw, Etsy Theo Schlossnagle, OmniTI Rasmus Lerdorf, creator of PHP Tom Cook, Facebook Benjamin Black, fast_ip Artur Bergman, Wikia Christopher Brown, Opscode Bryan Cantrill, Joyent Baron Schwartz, Percona Paul Querna, Cloudkick Surge 2010 focuses on real case studies from production environments; the lessons learned from failure and how to re-engineer your way to a successful, highly scalable Internet architecture. The conference takes place at the Tremont Grand Historic Venue on Sept 30 and Oct 1, 2010 in Baltimore, MD. Register now to enjoy the Early Bird discount and guarantee your seat to this year's event! http://omniti.com/surge/2010/register Thanks, -- Jason Dixon OmniTI Computer Consulting, Inc. jdixon@omniti.com 443.325.1357 x.241 From owner-freebsd-chat@FreeBSD.ORG Mon Aug 2 18:36:50 2010 Return-Path: Delivered-To: chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id CD424106564A for ; Mon, 2 Aug 2010 18:36:50 +0000 (UTC) (envelope-from brett@lariat.net) Received: from lariat.net (lariat.net [66.119.58.2]) by mx1.freebsd.org (Postfix) with ESMTP id 247318FC19 for ; Mon, 2 Aug 2010 18:36:49 +0000 (UTC) Received: from WildRover.lariat.net (wagner.lariat.net [66.119.58.131]) by lariat.net (8.9.3/8.9.3) with ESMTP id MAA22246 for ; Mon, 2 Aug 2010 12:24:00 -0600 (MDT) Message-Id: <201008021824.MAA22246@lariat.net> X-Mailer: QUALCOMM Windows Eudora Version 7.1.0.9 Date: Mon, 02 Aug 2010 12:23:57 -0600 To: chat@freebsd.org From: Brett Glass Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Cc: Subject: Packet steering/SMP X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Aug 2010 18:36:50 -0000 The article at http://www.computerworld.com/s/article/9180022/Latest_Linux_kernel_uses_Google_made_protocols describes SMP optimizations to the Linux kernel (the article mistakenly calls them "protocols," but they're not) which steer the processing of incoming network packets to the CPU core that is running the process for which they're destined. (Doing this requires code which straddles network layers in interesting ways.) The article claims that these optimizations are Google's invention, though they simply seem like a common sense way to make the best use of CPU cache. The article claims dramatic performance improvements due to this optimization. Anything like this in the works for FreeBSD? --Brett Glass From owner-freebsd-chat@FreeBSD.ORG Tue Aug 3 13:57:09 2010 Return-Path: Delivered-To: freebsd-chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 4707D106566B for ; Tue, 3 Aug 2010 13:57:09 +0000 (UTC) (envelope-from jhb@freebsd.org) Received: from cyrus.watson.org (cyrus.watson.org [65.122.17.42]) by mx1.freebsd.org (Postfix) with ESMTP id 196808FC08 for ; Tue, 3 Aug 2010 13:57:09 +0000 (UTC) Received: from bigwig.baldwin.cx (66.111.2.69.static.nyinternet.net [66.111.2.69]) by cyrus.watson.org (Postfix) with ESMTPSA id B90F146B6C; Tue, 3 Aug 2010 09:57:08 -0400 (EDT) Received: from jhbbsd.localnet (smtp.hudson-trading.com [209.249.190.9]) by bigwig.baldwin.cx (Postfix) with ESMTPSA id 94D598A03C; Tue, 3 Aug 2010 09:57:07 -0400 (EDT) From: John Baldwin To: freebsd-chat@freebsd.org Date: Tue, 3 Aug 2010 09:30:24 -0400 User-Agent: KMail/1.13.5 (FreeBSD/7.3-CBSD-20100217; KDE/4.4.5; amd64; ; ) References: <201008021824.MAA22246@lariat.net> In-Reply-To: <201008021824.MAA22246@lariat.net> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <201008030930.24070.jhb@freebsd.org> X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.0.1 (bigwig.baldwin.cx); Tue, 03 Aug 2010 09:57:07 -0400 (EDT) X-Virus-Scanned: clamav-milter 0.95.1 at bigwig.baldwin.cx X-Virus-Status: Clean X-Spam-Status: No, score=-2.6 required=4.2 tests=AWL,BAYES_00 autolearn=ham version=3.2.5 X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on bigwig.baldwin.cx Cc: Brett Glass Subject: Re: Packet steering/SMP X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Aug 2010 13:57:09 -0000 On Monday, August 02, 2010 2:23:57 pm Brett Glass wrote: > The article at > > http://www.computerworld.com/s/article/9180022/Latest_Linux_kernel_uses_Google_made_protocols > > describes SMP optimizations to the Linux kernel (the article > mistakenly calls them "protocols," but they're not) which steer the > processing of incoming network packets to the CPU core that is > running the process for which they're destined. (Doing this > requires code which straddles network layers in interesting ways.) > The article claims that these optimizations are Google's invention, > though they simply seem like a common sense way to make the best > use of CPU cache. > > The article claims dramatic performance improvements due to this > optimization. Anything like this in the works for FreeBSD? You should talk to Robert Watson, he is working on something similar. -- John Baldwin From owner-freebsd-chat@FreeBSD.ORG Tue Aug 3 14:22:09 2010 Return-Path: Delivered-To: chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id B6B361065670 for ; Tue, 3 Aug 2010 14:22:09 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [65.122.17.42]) by mx1.freebsd.org (Postfix) with ESMTP id 911A48FC17 for ; Tue, 3 Aug 2010 14:22:09 +0000 (UTC) Received: from fledge.watson.org (fledge.watson.org [65.122.17.41]) by cyrus.watson.org (Postfix) with ESMTPS id 290B846B89; Tue, 3 Aug 2010 10:22:09 -0400 (EDT) Date: Tue, 3 Aug 2010 15:22:09 +0100 (BST) From: Robert Watson X-X-Sender: robert@fledge.watson.org To: Brett Glass In-Reply-To: <201008021824.MAA22246@lariat.net> Message-ID: References: <201008021824.MAA22246@lariat.net> User-Agent: Alpine 2.00 (BSF 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: chat@freebsd.org Subject: Re: Packet steering/SMP X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Aug 2010 14:22:09 -0000 On Mon, 2 Aug 2010, Brett Glass wrote: > http://www.computerworld.com/s/article/9180022/Latest_Linux_kernel_uses_Google_made_protocols > > describes SMP optimizations to the Linux kernel (the article mistakenly > calls them "protocols," but they're not) which steer the processing of > incoming network packets to the CPU core that is running the process for > which they're destined. (Doing this requires code which straddles network > layers in interesting ways.) The article claims that these optimizations are > Google's invention, though they simply seem like a common sense way to make > the best use of CPU cache. > > The article claims dramatic performance improvements due to this > optimization. Anything like this in the works for FreeBSD? Quite a few systems do things like this, although perhaps not the exact formula that Google has. For example, Solarflare's TCP onload engine programms their hardware to direct5-tuples to specific queues for use by specific processes. Likewise, Chelsio's recenetly committed TCAM programming code allows work to be similarly directed to specific queues (and generally CPUs), although not in a way tightly integrated with the network stack. I'm currently doing some work for Juniper to add affinity features up and down the stack. Right now my prototype does this with RSS but doesn't attempt to expose specific flow affinity to userspace, or allow userspace to direct affinity. I have some early hacks at socket options to do that, although my goal was to perform flow direction in hardware (i.e., have the network stack program the TCAM on the T3 cards) rather than do the redirection in software. However, some recent experiments I ran that did work distribution to the per-CPU netisr workers I added in FreeBSD 8 were surprisingly effective -- not as good as distribution in hardware, but still significantly more throughput on an 8-core system (in this case I used RSS hashes generated by the hardware). Adding some sort of software redirection affinity table wouldn't be all that difficult, but I'll continue to focus on hardware distribution for the time being -- several cards out there will support the model pretty nicely. The only real limitations there are (a) which cards support it -- not Intel NICs, I'm afraid and (b) the sizes of the hardware flow direction tables -- usually in the thousands to tends of thousands range. Robert From owner-freebsd-chat@FreeBSD.ORG Tue Aug 3 14:38:58 2010 Return-Path: Delivered-To: freebsd-chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 560AF106566C for ; Tue, 3 Aug 2010 14:38:58 +0000 (UTC) (envelope-from vince@unsane.co.uk) Received: from unsane.co.uk (unsane-pt.tunnel.tserv5.lon1.ipv6.he.net [IPv6:2001:470:1f08:110::2]) by mx1.freebsd.org (Postfix) with ESMTP id BA4438FC1A for ; Tue, 3 Aug 2010 14:38:57 +0000 (UTC) Received: from vhoffman.lon.namesco.net (150.117-84-212.staticip.namesco.net [212.84.117.150]) (authenticated bits=0) by unsane.co.uk (8.14.4/8.14.4) with ESMTP id o73EcsRn044457 (version=TLSv1/SSLv3 cipher=DHE-RSA-CAMELLIA256-SHA bits=256 verify=NO) for ; Tue, 3 Aug 2010 15:38:55 +0100 (BST) (envelope-from vince@unsane.co.uk) Message-ID: <4C5829FD.2050600@unsane.co.uk> Date: Tue, 03 Aug 2010 15:38:53 +0100 From: Vincent Hoffman User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.1.9) Gecko/20100317 Thunderbird/3.0.4 MIME-Version: 1.0 To: freebsd-chat@freebsd.org References: <201008021824.MAA22246@lariat.net> In-Reply-To: X-Enigmail-Version: 1.0.1 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: Packet steering/SMP X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Aug 2010 14:38:58 -0000 On 03/08/2010 15:22, Robert Watson wrote: > > On Mon, 2 Aug 2010, Brett Glass wrote: > >> http://www.computerworld.com/s/article/9180022/Latest_Linux_kernel_uses_Google_made_protocols >> >> >> describes SMP optimizations to the Linux kernel (the article >> mistakenly calls them "protocols," but they're not) which steer the >> processing of incoming network packets to the CPU core that is >> running the process for which they're destined. (Doing this requires >> code which straddles network layers in interesting ways.) The article >> claims that these optimizations are Google's invention, though they >> simply seem like a common sense way to make the best use of CPU cache. >> >> The article claims dramatic performance improvements due to this >> optimization. Anything like this in the works for FreeBSD? > > Quite a few systems do things like this, although perhaps not the > exact formula that Google has. For example, Solarflare's TCP onload > engine programms their hardware to direct5-tuples to specific queues > for use by specific processes. Likewise, Chelsio's recenetly > committed TCAM programming code allows work to be similarly directed > to specific queues (and generally CPUs), although not in a way tightly > integrated with the network stack. > > I'm currently doing some work for Juniper to add affinity features up > and down the stack. Right now my prototype does this with RSS but > doesn't attempt to expose specific flow affinity to userspace, or > allow userspace to direct affinity. I have some early hacks at socket > options to do that, although my goal was to perform flow direction in > hardware (i.e., have the network stack program the TCAM on the T3 > cards) rather than do the redirection in software. However, some > recent experiments I ran that did work distribution to the per-CPU > netisr workers I added in FreeBSD 8 were surprisingly effective -- not > as good as distribution in hardware, but still significantly more > throughput on an 8-core system (in this case I used RSS hashes > generated by the hardware). > > Adding some sort of software redirection affinity table wouldn't be > all that difficult, but I'll continue to focus on hardware > distribution for the time being -- several cards out there will > support the model pretty nicely. The only real limitations there are > (a) which cards support it -- not Intel NICs, I'm afraid and (b) the > sizes of the hardware flow direction tables -- usually in the > thousands to tends of thousands range. Fasinating stuff. I read http://www.ntop.org/blog/?p=86 to mean that the newer intel NICs had similar capabilities to the Chelsio T3, is it just a lack of driver and control support in FreeBSD or did I misunderstand? Vince > > Robert > _______________________________________________ > freebsd-chat@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-chat > To unsubscribe, send any mail to "freebsd-chat-unsubscribe@freebsd.org" From owner-freebsd-chat@FreeBSD.ORG Tue Aug 3 15:47:20 2010 Return-Path: Delivered-To: freebsd-chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 2EF0E106568B for ; Tue, 3 Aug 2010 15:47:20 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [65.122.17.42]) by mx1.freebsd.org (Postfix) with ESMTP id 0A8EF8FC08 for ; Tue, 3 Aug 2010 15:47:20 +0000 (UTC) Received: from fledge.watson.org (fledge.watson.org [65.122.17.41]) by cyrus.watson.org (Postfix) with ESMTPS id 7A9F546B06; Tue, 3 Aug 2010 11:47:19 -0400 (EDT) Date: Tue, 3 Aug 2010 16:47:19 +0100 (BST) From: Robert Watson X-X-Sender: robert@fledge.watson.org To: Vincent Hoffman In-Reply-To: <4C5829FD.2050600@unsane.co.uk> Message-ID: References: <201008021824.MAA22246@lariat.net> <4C5829FD.2050600@unsane.co.uk> User-Agent: Alpine 2.00 (BSF 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: freebsd-chat@freebsd.org Subject: Re: Packet steering/SMP X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Aug 2010 15:47:20 -0000 On Tue, 3 Aug 2010, Vincent Hoffman wrote: > Fasinating stuff. I read http://www.ntop.org/blog/?p=86 to mean that the > newer intel NICs had similar capabilities to the Chelsio T3, is it just a > lack of driver and control support in FreeBSD or did I misunderstand? The Intel cards I've looked at often have quite small tables -- appropriate for directing packets for virtualization purposes (i.e., a few MACs to specific CPUs), and fine for distributing using RSS, but not able to direct tens of thousands of connections individually by 4-tuple to specific queues. It wouldn't surprise me if newer generations of Intel cards support similar features. If you're interested in high-performance packet capture to userspace, this GSoC project might be interesting though: http://wiki.freebsd.org/AlexandreFiveg/SOC2010AlexandreFive Robert From owner-freebsd-chat@FreeBSD.ORG Tue Aug 3 16:23:09 2010 Return-Path: Delivered-To: freebsd-chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 26CB71065679 for ; Tue, 3 Aug 2010 16:23:09 +0000 (UTC) (envelope-from vince@unsane.co.uk) Received: from unsane.co.uk (unsane-pt.tunnel.tserv5.lon1.ipv6.he.net [IPv6:2001:470:1f08:110::2]) by mx1.freebsd.org (Postfix) with ESMTP id A217E8FC1A for ; Tue, 3 Aug 2010 16:23:08 +0000 (UTC) Received: from vhoffman.lon.namesco.net (150.117-84-212.staticip.namesco.net [212.84.117.150]) (authenticated bits=0) by unsane.co.uk (8.14.4/8.14.4) with ESMTP id o73GN6oN048991 (version=TLSv1/SSLv3 cipher=DHE-RSA-CAMELLIA256-SHA bits=256 verify=NO) for ; Tue, 3 Aug 2010 17:23:07 +0100 (BST) (envelope-from vince@unsane.co.uk) Message-ID: <4C58426A.9020905@unsane.co.uk> Date: Tue, 03 Aug 2010 17:23:06 +0100 From: Vincent Hoffman User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.1.9) Gecko/20100317 Thunderbird/3.0.4 MIME-Version: 1.0 To: freebsd-chat@freebsd.org References: <201008021824.MAA22246@lariat.net> <4C5829FD.2050600@unsane.co.uk> In-Reply-To: X-Enigmail-Version: 1.0.1 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: Packet steering/SMP X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Aug 2010 16:23:09 -0000 On 03/08/2010 16:47, Robert Watson wrote: > On Tue, 3 Aug 2010, Vincent Hoffman wrote: > >> Fasinating stuff. I read http://www.ntop.org/blog/?p=86 to mean that >> the newer intel NICs had similar capabilities to the Chelsio T3, is >> it just a lack of driver and control support in FreeBSD or did I >> misunderstand? > > The Intel cards I've looked at often have quite small tables -- > appropriate for directing packets for virtualization purposes (i.e., a > few MACs to specific CPUs), and fine for distributing using RSS, but > not able to direct tens of thousands of connections individually by > 4-tuple to specific queues. It wouldn't surprise me if newer > generations of Intel cards support similar features. > > If you're interested in high-performance packet capture to userspace, > this GSoC project might be interesting though: > > http://wiki.freebsd.org/AlexandreFiveg/SOC2010AlexandreFive I did notice this, I'll be paying attention to see what the results are. Vince > > Robert > _______________________________________________ > freebsd-chat@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-chat > To unsubscribe, send any mail to "freebsd-chat-unsubscribe@freebsd.org" From owner-freebsd-chat@FreeBSD.ORG Tue Aug 3 23:39:43 2010 Return-Path: Delivered-To: freebsd-chat@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A6ADB1065678 for ; Tue, 3 Aug 2010 23:39:43 +0000 (UTC) (envelope-from vvelox@vvelox.net) Received: from vulpes.vvelox.net (sula-ki.vvelox.net [99.69.115.46]) by mx1.freebsd.org (Postfix) with ESMTP id 6AA498FC19 for ; Tue, 3 Aug 2010 23:39:42 +0000 (UTC) Received: from vixen42.vulpes.vvelox.net (unknown [192.168.14.2]) (Authenticated sender: vvelox) by vulpes.vvelox.net (Postfix) with ESMTP id A0DF2B84D; Tue, 3 Aug 2010 18:23:02 -0500 (CDT) Date: Tue, 3 Aug 2010 18:21:43 -0500 From: "Zane C.B." To: Alexander Best Message-ID: <20100803182143.0986b62d@vixen42.vulpes.vvelox.net> In-Reply-To: <20100727094151.GA68226@freebsd.org> References: <20100727094151.GA68226@freebsd.org> X-Mailer: Claws Mail 3.7.6 (GTK+ 2.20.1; amd64-portbld-freebsd8.0) Mime-Version: 1.0 Content-Type: multipart/signed; micalg=PGP-SHA1; boundary="Sig_/ZaeGB8.kWJhdnsiK_tf1CwC"; protocol="application/pgp-signature" Cc: freebsd-chat@freebsd.org Subject: Re: ZFS ATA vs. ZFS CAM ATA performance on 8.1 X-BeenThere: freebsd-chat@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Non technical items related to the community List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Aug 2010 23:39:43 -0000 --Sig_/ZaeGB8.kWJhdnsiK_tf1CwC Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable On Tue, 27 Jul 2010 09:41:51 +0000 Alexander Best wrote: > hi there, >=20 > i just stumpled upon this article over at phoronix which benchmarks > ZFS ATA vs. ZFS CAM ATA on freebsd 8.1 [1]. it seems read > performance is really low when CAM ATA is enabled. i remember > phoronix being famous for posting stupid benchmarks (RELASE vs. > HEAD and such). however their benchmark results in this example > seem to be valid. >=20 > any comments on that? has performance of ATA CAM increased in HEAD? > would a UFS2+S+SUJ ATA vs. UFS2+S+SUJ CAM ATA also show equal > results? Not sure about any speed increases with ATA_CAM, but I for one controller I have in this system I did get a massive boost and a end to annoying disk issues with AHCI. --Sig_/ZaeGB8.kWJhdnsiK_tf1CwC Content-Type: application/pgp-signature; name=signature.asc Content-Disposition: attachment; filename=signature.asc -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (FreeBSD) iEYEARECAAYFAkxYpIoACgkQC1tfcMGJid6qJACgqtHdeABJaYRdzfx5hW3L9xFk 9J4An36x90+Y1tNoLBJF+3l7NokkJ4DW =U6be -----END PGP SIGNATURE----- --Sig_/ZaeGB8.kWJhdnsiK_tf1CwC--