From mboxrd@z Thu Jan 1 00:00:00 1970 From: roopa Subject: Re: Per-flow IPv4 ECMP Date: Mon, 28 Sep 2015 06:48:09 -0700 Message-ID: <56094519.8030801@cumulusnetworks.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "netdev@vger.kernel.org" , pch@ordbogen.com To: Matthew Dupre Return-path: Received: from mail-pa0-f48.google.com ([209.85.220.48]:36483 "EHLO mail-pa0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755163AbbI1NsL (ORCPT ); Mon, 28 Sep 2015 09:48:11 -0400 Received: by pablk4 with SMTP id lk4so78429522pab.3 for ; Mon, 28 Sep 2015 06:48:10 -0700 (PDT) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On 9/28/15, 1:57 AM, Matthew Dupre wrote: > Hi, > > I'm interested in the Linux kernel's support for per-flow IPv4 ECMP (= i.e. consistent path selection based on a hash of the connection tuple)= =2E I'd been led to believe[1] that this depended on the route cache, = which was removed in 3.6. > > However, I tested a route with multiple next hops on a 3.10 and 3.13 = kernel, and ECMP was per-flow! Obviously I'm pleased that this was the= case, but I'd like to understand why this is supported, and whether I = can rely on it in future. > > Could anyone give me a little clarification on whether this is now su= pported by some means other than the route cache, and whether that supp= ort is intended to be continued? > This is being worked on currently by Peter N=F8rlund https://lwn.net/Articles/657431/