From owner-freebsd-questions@FreeBSD.ORG Wed Jun 17 18:55:48 2009 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id F061E106564A for ; Wed, 17 Jun 2009 18:55:48 +0000 (UTC) (envelope-from nvass9573@gmx.com) Received: from mail.gmx.com (unknown [213.165.64.42]) by mx1.freebsd.org (Postfix) with SMTP id DDA1D8FC17 for ; Wed, 17 Jun 2009 18:55:47 +0000 (UTC) (envelope-from nvass9573@gmx.com) Received: (qmail invoked by alias); 17 Jun 2009 18:55:46 -0000 Received: from ipa255.25.107.79.tellas.gr (EHLO [192.168.254.1]) [79.107.25.255] by mail.gmx.com (mp-eu002) with SMTP; 17 Jun 2009 20:55:46 +0200 X-Authenticated: #46156728 X-Provags-ID: V01U2FsdGVkX18aviw2OW8aji7y5e8wLtPdltmqfzyiGVXVFRE7ot 6ZZoCDLLtUxnjL Message-ID: <4A393C21.8040800@gmx.com> Date: Wed, 17 Jun 2009 21:55:29 +0300 From: Nikos Vassiliadis User-Agent: Thunderbird 2.0.0.21 (Windows/20090302) MIME-Version: 1.0 To: Gary Gatten References: <70C0964126D66F458E688618E1CD008A0793EFF4@WADPEXV0.waddell.com> In-Reply-To: <70C0964126D66F458E688618E1CD008A0793EFF4@WADPEXV0.waddell.com> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 X-FuHaFi: 0.62 Cc: wojtek@wojtek.tensor.gdynia.pl, freebsd-questions@freebsd.org, raulbecette@unp.edu.ar Subject: Re: Gateway load balance X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 17 Jun 2009 18:55:49 -0000 Gary Gatten wrote: > I meant, the original source flows are nat'd to whatever outbound interface ip they are RRLB to. That way any internal flow can use any isp connection. Most of my interesting routing, NAT, etc I do with Ci$co so not all that up on BSD's options. Yes, works with ci$co, doesn't with FreeBSD ;) You cannot add more than one next hop for any given prefix. Well, you can achieve almost the same result, but only with the help of a packet filter which can bypass the routing table lookup and forward to the next hop of preference. But that's hardly the same with ECMP which many vendors support... Yet, that's about to change. There is ECMP support in 8.0:)