Re: [dhcwg] [v6ops] Combining IPv6 ND and DHCPv6 into a single, unified function

["Templin, Fred L" <[e-mail]>]

Hi Michael,

This isn't an experiment; this is a unified stateless/stateful autoconfiguration
service for IPv6 for all time.

We all want prefix delegation, and that will necessarily be a stateful function.
And, if the prefix delegation service would not be DHCPv6 PD then we would
need to develop a new PD architecture. Either way, though, it should be
implemented as a unified function with IPv6ND.

Thanks - Fred

> -----Original Message-----
> From: ipv6 [mailto:[e-mail]] On Behalf Of Michael Richardson
> Sent: Wednesday, November 22, 2017 8:54 AM
> To: [e-mail]; [e-mail]; [e-mail]
> Subject: Re: [v6ops] Combining IPv6 ND and DHCPv6 into a single, unified function
> 
> 
> My understanding is that Fred Templin essentially wants to carry DHCP
> (configuration) options in ND/RA.
> 
> I think that they are most often stateless (i.e. not addresses at all or
> other resources).
> 
> I've heard the IPv4 PPP folks lament that they should have just had an IPCP
> option to carry DHCPv4 options, and indeed that's why we have almost no IP6CP
> options, as it's all done by ND and DHCPv6.
> 
> {Having implemented a DHCPv6 server just to serve PPP links, it really feels
> dumb.  Yes, the management is all centralized, but it's centralized via
> RADIUS, not via DHCP relays. }
> 
> So I suggest that Fred should go ahead with his document as Experimental,
> with our blessing, and an obligation to report on how well it works in
> his situation.   Given the timescales that I think his industry works at, he
> might need more time than an average Experimental code allocation.
> He would have to deal how/if/when all experimental systems would be withdrawn
> should the experiment be unsuccessful, I think.
> 
> I guess he could use ND codes 253/254, but I think an assignment would not
> kill here.
> 
> --
> Michael Richardson <[e-mail]>, Sandelman Software Works
>  -= IPv6 IoT consulting =-
> 
>