Hi Xiaohu, > -----Original Message----- > From: Xu Xiaohu [mailto:[email protected]] > Sent: Tuesday, March 22, 2011 10:24 PM > To: Templin, Fred L; 'Ole Troan'; 'Rémi Després' > Cc: [email protected]; [email protected] > Subject: > re:[Softwires][dhcwg]fwd:NewVersionNotificationfordraft-guo-so > ftwire-6rd-ipv6-config-02 > > Hi Fred, > > > -----邮件原件----- > > 发件人: Templin, Fred L [mailto:[email protected]] > > 发送时间: 2011年3月22日 23:06 > > 收件人: Xu Xiaohu; 'Ole Troan'; 'Rémi Després' > > 抄送: [email protected]; [email protected] > > 主题: RE: > > > [Softwires][dhcwg]fwd:NewVersionNotificationfordraft-guo-softw > ire-6rd-ipv6 > > -config-02 > > > > Hi again Xiaohu, > > > > > > Thanks for your explanation. I understand the advantages now. > > > > > > > > In your approach, to obtain stable IPv6 prefixes for CEs, > > > some routing > > > > protocol should be run between the CE and the BR while the BR > > > > would have to > > > > maintain a lot of routes for the delegated native IPv6 > > > > prefixes. > > > > > > The routing protocol in this case would be DHCPv6 > > > prefix delegation. To spread the load, we simply > > > add more BRs, and make them communicate with each > > > other either directly in a full mesh or via a set > > > of gateways in a partial mesh. > > > > This might bear a few more words of explanation. By "the routing > > protocol in this case would be DHCPv6 prefix delegation", what I > > mean is that the BR acting as a Delegating Router should register > > a "link up" event when a CE acting as a Requesting Router asks it > > for a prefix delegation, and the BR should maintain that > link as being > > up until the CE either moves away or dies. On the backhaul network > > Without any routing protocol, there should be some keep-alive > mechanism > running between the BR and the CE at least. Otherwise, the > route for the > delegated prefix could be pointed to an invalid or even a > wrong next-hop (i. > e., CE). Is my understanding right?
Usually, routing protocols require some form of hello messaging to maintain link states. We could use DHCP lease renewals for this purpose, but I wouldn't want to set so short a lease lifetime. So, the CE could send periodic "pings" to the BR in the data plane, and the BR could respond back with ACKs to let th CE know that it is stil alive. Does this address the question? Thanks - Fred [email protected] > Best wishes, > Xiaohu > > > that connects BRs, then, I am assuming something like an iBGP > > instance where all BRs advertise their links that are up > and withdraw > > their links that have gone down. > > > > Fred > > [email protected]= > > _______________________________________________ Softwires mailing list [email protected] https://www.ietf.org/mailman/listinfo/softwires
