From owner-freebsd-arch@FreeBSD.ORG Sat Mar 19 15:15:51 2011 Return-Path: Delivered-To: freebsd-arch@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 05CD210656A7 for ; Sat, 19 Mar 2011 15:15:51 +0000 (UTC) (envelope-from imp@bsdimp.com) Received: from harmony.bsdimp.com (bsdimp.com [199.45.160.85]) by mx1.freebsd.org (Postfix) with ESMTP id C06838FC08 for ; Sat, 19 Mar 2011 15:15:50 +0000 (UTC) Received: from [127.0.0.1] (localhost [127.0.0.1]) by harmony.bsdimp.com (8.14.3/8.14.1) with ESMTP id p2JF9jv1055338 for ; Sat, 19 Mar 2011 09:09:47 -0600 (MDT) (envelope-from imp@bsdimp.com) Message-ID: <4D84C72A.3030905@bsdimp.com> Date: Sat, 19 Mar 2011 09:09:30 -0600 From: Warner Losh User-Agent: Mozilla/5.0 (X11; U; FreeBSD amd64; en-US; rv:1.9.2.13) Gecko/20101211 Thunderbird/3.1.7 MIME-Version: 1.0 To: freebsd-arch@FreeBSD.org References: <132388F1-44D9-45C9-AE05-1799A7A2DCD9@neville-neil.com> In-Reply-To: <132388F1-44D9-45C9-AE05-1799A7A2DCD9@neville-neil.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Subject: Re: Updating our TCP and socket sysctl values... X-BeenThere: freebsd-arch@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussion related to FreeBSD architecture List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 19 Mar 2011 15:15:51 -0000 On 03/19/2011 00:37, George Neville-Neil wrote: > I believe it's time for us to upgrade our sysctl values for TCP sockets so that > they are more in line with the modern world. At the moment we have these limits on > our buffering: > > kern.ipc.maxsockbuf: 262144 > net.inet.tcp.recvbuf_max: 262144 > net.inet.tcp.sendbuf_max: 262144 > > I believe it's time to up these values to something that's in line with higher speed > local networks, such as 10G. Perhaps it's time to move these to 2MB instead of 256K. > > Thoughts? So long as these can be tuned down to the lower end for memory constrained environments, that's fine. In general, I believe that the project has aimed to having the tunables in the system optimized for the fastest hardware, while being usable on the last few generations of hardware and tunable, where possible, for the older hardware. Warner