On Sat, Feb 21, 2015 at 07:02:44AM +0100, Mikael Abrahamsson wrote: > >NO, just between the first-hop (homenet) routers. Should work with > >unchanged > >of the shelf crap-APs as long as they're attached to a homenet router. > > Could someone please explain to me how this is supposed to work? How do > the first-hop routers figure out where the client is? Do we do /128 route > injection into the homenet for active IPv6 addresses in that /64, and > announce the same /64 everywhere, and then we use proxy-nd for all off-L2 > active IPs? I guess we need to handle IPv4 as well, so we use proxy-arp > for that?
I was referring to eg: https://tools.ietf.org/html/draft-hertoghs-lisp-mobility-use-cases-01 and available commercial implementations, eg: from cisco for this (google lisp host mobility and try to ignore all the lisp details ;-)). I haven't tried to figure out all details and whether/how we would want to adopt a similar scheme to homenet. Something like: -> a first-hop-router would need to be able to detect presence of a roaming client. Probably based on MAC-address. And then generate the /128 (and /32) prefix for the clients address. Thats effectivly what LISP does, but they do not need to only create a route but feed into their database. -> for seamless mobility, you'd need to have on all subnets for the default-router the same virtual MAC address (eg: VRRP). Thats also someting set up with LISP. This ensures that traffic sent by the client will continue to be forwarded after roamin. -> What i couldn't figure out quickly yet is how DHCP updates to the client are dealt with especially default-router IP address. Cheers Toerless > Mikael Abrahamsson email: [email protected] -- --- Toerless Eckert, [email protected] _______________________________________________ homenet mailing list [email protected] https://www.ietf.org/mailman/listinfo/homenet
