From mboxrd@z Thu Jan 1 00:00:00 1970 From: Henrik Petander Subject: Re: [patch]: CONFIG_IPV6_SUBTREES fix for MIPv6 Date: Wed, 11 Jun 2003 11:48:19 +0300 Sender: netdev-bounce@oss.sgi.com Message-ID: <3EE6ECD3.6050103@tml.hut.fi> References: <20030609203659.089b241b.nakam@linux-ipv6.org> <3EE5F85E.9080006@tml.hut.fi> <20030610.095135.28806569.davem@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Cc: nakam@linux-ipv6.org, lpetande@morphine.tml.hut.fi, yoshfuji@linux-ipv6.org, vnuorval@tcs.hut.fi, kuznet@ms2.inr.ac.ru, netdev@oss.sgi.com, ajtuomin@morphine.tml.hut.fi, jagana@us.ibm.com, kumarkr@us.ibm.com, usagi-core@linux-ipv6.org Return-path: To: "David S. Miller" In-Reply-To: <20030610.095135.28806569.davem@redhat.com> Errors-to: netdev-bounce@oss.sgi.com List-Id: netdev.vger.kernel.org David S. Miller wrote: > > If you want to do these things using routes or xfrm rules, you must > integrate the creation of them into either zebra or racoon. You > cannot have a setup where mipv6d and racoon/zebra fight each other > flushing each other's settings. It doesn't work. > In the routing based approach there should not be any conflicts between mipv6 and zebra: We would create cached host routes based on the existing routes. Thus if zebra was running, the mipv6 daemon would not change the routes created by zebra, but only cached host routes. If zebra changed any routes, it would cause the deletion of any invalid cached routes. The mipv6 daemon would listen to netlink messages for route deletion and would then reinsert the mipv6 state into a new cached route. Does this make sense to you? Thanks, Henrik