Re: [v6ops] How do you solve 3GPP issue if neither operator nor handset supports PD?

[Alexandre Petrescu <[e-mail]>]

Le 24/11/2020 à 13:34, Philip Homburg a écrit :
[...]
> The reason is that in theory you should do neighbor discovery even if a link
> has no L2 addresses. I found that many implementations just don't react
> to an NS or send any themselves.

Indeed NS/NA are also sent on some ptp links.  They are there just to 
respect standard, in some cases.

The use of ND is relevant on ptp links, however.

The distinction is in what is that ND used for.  Is NS/NA used to 
resolve an address that is a nexthop in the rt table, or is it used to 
resolve an address that is just in the dst of the packet.

If the Gateway sends an NS for an address that is in the nexthop field 
in an entry in its rt table, then we are in presence of delegation.

If the Gateway sends an NS for an address that is in the dst field of a 
packet that it needs to forward, then we are in presence of assignment. 
  In this case, the NS/NA is not of any use, because the NA is not used.

Alex

> 
> So all traffic is just sent to the other side of the link. I.e., the route
> could actually point to the link, but you can't really tell the difference.
> 
> In theory you could try to see if the router has an address on the link, but
> I never tried that experiment.
> 
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> [e-mail]
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
>