From mboxrd@z Thu Jan 1 00:00:00 1970 From: Brian Haley Subject: Re: [RFC] bonding: add better ipv6 failover support Date: Thu, 25 Sep 2008 11:42:57 -0400 Message-ID: <48DBB181.9050205@hp.com> References: <200809151335.16817.asid@hp.com> <20080915180015.GB1078@havoc.gtf.org> <200809151416.49447.alexandre.sidorenko@hp.com> <48DA71A8.5050900@hp.com> <7958.1222288188@death.nxdomain.ibm.com> <48DAFB92.7040904@hp.com> <17219.1222355242@death.nxdomain.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Vlad Yasevich , Alex Sidorenko , Jeff Garzik , "netdev@vger.kernel.org" To: Jay Vosburgh Return-path: Received: from g4t0017.houston.hp.com ([15.201.24.20]:32571 "EHLO g4t0017.houston.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752105AbYIYPnA (ORCPT ); Thu, 25 Sep 2008 11:43:00 -0400 In-Reply-To: <17219.1222355242@death.nxdomain.ibm.com> Sender: netdev-owner@vger.kernel.org List-ID: Jay Vosburgh wrote: > Brian Haley wrote: > >> This is an RFC patch to add better IPv6 failover support for bonding >> devices, especially when in active-backup mode, as reported by Alex >> Sidorenko. >> >> What this patch does: >> >> - Creates a new Kconfig option in the IPv6 Networking section to >> compile-in the support in the bonding driver. This also forces >> IPV6=y since that's required to link everything. > > I think it's probably better to have the IPV6 dependent bits > somehow depend on CONFIG_IPV6 rather than having a Kconfig entry. I > doubt that many real-world users will say yes to IPv6 and bonding, but > no to the bonding IPv6 support. I also suspect that the IPV6=y > requirement won't fly with distros. I'm sure there's a way to do this better, for example, SCTP can be built as a module with IPv6 support and have IPV6=m. I'll try to make it work without the option when IPV6=y or m. >> - Adds a new master_ipv6 address member to the bonding struct to >> hold a copy of the primary IPv6 address on the bond. > > Do we need to issue an NS for each ipv6 address, or is one > sufficient? It didn't seem like it from my testing, that single NS was enough to wake-up the switch when pinging either the link-local or global. I'd have to add another global with a different prefix and re-test. > Do ipv6 addresses configured on VLANs need one (or more) NS per > VLAN? I didn't test with VLANs, there would probably need to be some additional work there. > I haven't tried the patch yet, so I'll comment further once > I've had a chance to test it (which may not be until tomorrow). Thanks, -Brian