From mboxrd@z Thu Jan 1 00:00:00 1970 From: Evgeniy Polyakov Subject: Re: Is 64k bind patch making bad assumption? Date: Mon, 2 Feb 2009 21:46:25 +0300 Message-ID: <20090202184625.GA1252@ioremap.net> References: <20090202103515.60788d0f@extreme> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Evgeniy Polyakov , David Miller , netdev@vger.kernel.org To: Stephen Hemminger Return-path: Received: from tservice.ru ([195.178.208.66]:33583 "EHLO tservice.net.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752596AbZBBSqj (ORCPT ); Mon, 2 Feb 2009 13:46:39 -0500 Content-Disposition: inline In-Reply-To: <20090202103515.60788d0f@extreme> Sender: netdev-owner@vger.kernel.org List-ID: Hi Stephen. On Mon, Feb 02, 2009 at 10:35:15AM -0800, Stephen Hemminger (shemminger@vyatta.com) wrote: > The patch to allow more that 64k connections is assuming that > all IP addresses on a machine have the same connectivity and policy. > I can think of several cases where this is not true: > 1. Server machine with public IP's for connections and private IP's > for backend database connection > 2. Applications where certain IP's are required for security > protocols > 3. Machines doing policy based routing > > Doesn't this patch change/break this? I do not think this is related. Patch itself just allows to do what users ask for. He can do the same with explicit port selection and reuse option, but bind(0) with the reuse option (i.e. automatic kernel port selection) was limited. -- Evgeniy Polyakov