Hello Pascal, We still believe NSA is IP address. For details, please see response inline.
Regards, Guangpeng > -----Original Message----- > From: Pascal Thubert (pthubert) <pthubert=40cisco....@dmarc.ietf.org> > Sent: Wednesday, August 24, 2022 1:17 PM > To: Liguangpeng (Roc, Network Technology Laboratory) > <liguangp...@huawei.com> > Cc: Michael Richardson <mcr+i...@sandelman.ca>; 6lo <6lo@ietf.org> > Subject: Re: [6lo] Call for WG adoption of > draft-li-6lo-native-short-address-03 > > Hello Guangpeng > > I tried and failed so far to convince you that NSA is not by nature an IP > address > as we are used to. In my book it sits at L2 or L2.5 (with BIER and MPLS). > > Consider this: > > - if it was an IID how many devices max could there be in a subnet ? > > This is very limited not at all the traditional order of IPv6. Also it seems > to > depend on the shape of the network. IP should not be impacted by L2 > considerations. > NSA is not impacted by any L2 techniques. It works on various L2 wireline technologies. So far we designed allocation function base on the tree topology, which we believe can be applied to many scenarios. > - NSA expresses a source route path not a destination. > > In that it is akin to an MPLS stack. > NSA indeed encodes path information into the address which greatly reduce complexity of routing, hence an advantage of solution. > - what if I swap the device at the end of the path? > > Traditional IP expects that a device can retain its address even if it is > plugged on > a different switch port. > As long as the parent node stays the same, NSA can also retain the child's node address when it swaps. > - what if there are multiple paths to the same device (again the point on > redundancy)? > We have two solutions illustrated in the reliability draft, please comment more on that. > So if you told me that you are shooting for IP in IP and that NSA is for the > destination in the outer header I’d be rather easy to convince. That would > become another variation of an SRv6 compression technique. I’d then suggest > you present it at spring. > > If you told me it is a L2 or L2.5 mesh under technique I’d also listen > carefully. > The debate would become whether 6lo is home for the work. > > But sorry I cannot see that NSA matches an IP address. That would fuse the > devices and the network together. > > Regards, > > Pascal > > > Le 24 août 2022 à 06:36, Liguangpeng (Roc, Network Technology > Laboratory) <liguangpeng=40huawei....@dmarc.ietf.org> a écrit : > > > > Hi Michael, > > > > That's great. Considering NSA lives with IPv6 space, there is very low > probability for the new device cannot get an address. I think we are aligned > with each other on supporting IPv6 in the first place. > > > > Cheers, > > Guangpeng > > > >> -----Original Message----- > >> From: Michael Richardson <mcr+i...@sandelman.ca> > >> Sent: Wednesday, August 24, 2022 12:00 AM > >> To: Liguangpeng (Roc, Network Technology Laboratory) > >> <liguangp...@huawei.com> > >> Cc: 6lo <6lo@ietf.org> > >> Subject: Re: [6lo] Call for WG adoption of > >> draft-li-6lo-native-short-address-03 > >> > >> > >> "Liguangpeng (Roc, Network Technology Laboratory)" wrote: > >>>>> No, as long as there is enough address for this new device. > >>>> > >>>> And if there isn't? > >> > >>> What happens if you want add the 256th device to a subnet with /24 > >> IPv4 > >>> address block? > >> > >> I was a fool to have no used IPv6 in the first place. > >> > >> -- > >> Michael Richardson <mcr+i...@sandelman.ca> . o O ( IPv6 IøT > consulting ) > >> Sandelman Software Works Inc, Ottawa and Worldwide > >> > >> > >> > > > > _______________________________________________ > > 6lo mailing list > > 6lo@ietf.org > > https://www.ietf.org/mailman/listinfo/6lo _______________________________________________ 6lo mailing list 6lo@ietf.org https://www.ietf.org/mailman/listinfo/6lo
