Received: with ECARTIS (v1.0.0; list netdev); Wed, 07 Jul 2004 12:31:38 -0700 (PDT) Received: from mail.shareable.org (mail.shareable.org [81.29.64.88]) by oss.sgi.com (8.12.10/8.12.9) with SMTP id i67JVZgi004592 for ; Wed, 7 Jul 2004 12:31:36 -0700 Received: from mail.shareable.org (localhost [127.0.0.1]) by mail.shareable.org (8.12.8/8.12.8) with ESMTP id i67JVP81019101; Wed, 7 Jul 2004 20:31:25 +0100 Received: (from jamie@localhost) by mail.shareable.org (8.12.8/8.12.8/Submit) id i67JVPg4019099; Wed, 7 Jul 2004 20:31:25 +0100 Date: Wed, 7 Jul 2004 20:31:25 +0100 From: Jamie Lokier To: Stephen Hemminger Cc: "David S. Miller" , bert hubert , netdev@oss.sgi.com, linux-net@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] fix tcp_default_win_scale. Message-ID: <20040707193125.GA17266@mail.shareable.org> References: <20040701133738.301b9e46@dell_ss3.pdx.osdl.net> <20040701140406.62dfbc2a.davem@redhat.com> <20040702013225.GA24707@conectiva.com.br> <20040706093503.GA8147@outpost.ds9a.nl> <20040706114741.1bf98bbe@dell_ss3.pdx.osdl.net> <20040706194034.GA11021@mail.shareable.org> <20040706131235.10b5afa8.davem@redhat.com> <20040706224453.GA6694@outpost.ds9a.nl> <20040706154907.422a6b73.davem@redhat.com> <20040707110653.7c49bef1@dell_ss3.pdx.osdl.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040707110653.7c49bef1@dell_ss3.pdx.osdl.net> User-Agent: Mutt/1.4.1i X-archive-position: 6761 X-ecartis-version: Ecartis v1.0.0 Sender: netdev-bounce@oss.sgi.com Errors-to: netdev-bounce@oss.sgi.com X-original-sender: jamie@shareable.org Precedence: bulk X-list: netdev Content-Length: 698 Lines: 19 Stephen Hemminger wrote: > But: isn't it better to have just one sysctl parameter set > (tcp_rmem) and set the window scale as needed rather than increasing > the already bewildering array of dials and knobs? I can't see why > it would be advantageous to set a window scale of 7 if the largest > possible window ever offered is limited to a smaller value? That's a fair question. It seems to me the only effects of a larger scale than necessary are (a) the buffer size can be increased after the connection is established, and (b) coarser granularity which can only degrade performance over low mss links. So why do we set a larger window scale than necessary? Is it to support (a)? -- Jamie