 |
 |
 |
 |
|
 |
|
 |
|
 |
|
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
Humble apologies for not reading this lowpan doc, but I have
listened to its core ideas during the past IETFs prezos and understand the link
model being used for it. I still have some general comments that
are worth discussing when bringing such work to 6man. Note that if the multi-link, multi-hop network has all
client nodes as off-link to each other, then there is only one type of regular
ND (RFC4861) RA that can signal off-link. This is an RA with no PIO
(Prefix Information Option). So if the RA has no PIO, how are the
lowpann client nodes acquiring their IPv6 address global address without SLAAC
getting no prefix from? In 6man, we would like to know how are the lowpan
clients acquiring their address and do they even have a global v6
address? I think I heard some mention of ULA, so that gives me a hint of
global address use because ULA has same scope as a Globally Unique Address
(GUA). Therefore it would be interesting to see the IPv6 ND RA
config on the lowpan edge routers. Anyone has working models of this work to demo to us during
any future IETF? Thanks, Hemant |