Qi,

>>> MAP is incompatible with dynamic IPv4 address assignment using DHCPv4.
>>> 
>>> Yes, dynamic IPv4 address assignment is only used by lw4o6. But 
>>> OPTION_S46_IPV4ADDRESS in map-dhcp-05 is also proposed for only lw4o6,
>>> so I think it's not a problem to have a discussion of dynamic IPv4 
>>> provisioning.
>>> 
>>> either you can consider MAP DHCP as the building block used by a separate 
>>> document describing the
>>> DHCPv4 address assignment case. or we can split out LW46 provisioning from 
>>> MAP DHCP altogether.
>>> 
>>> I think the key problem here is the scope of the document: whether 
>>> "softwire-dhcp" discussing both static and dynamic IPv4 provisioning,
>> 
>> to nitpick, MAP DHCP is also dynamic by the way, just that the IPv4 address 
>> lifetime is equal to the IPv6 prefix lifetime.
> 
> [Qi] IMO, map-dhcp has an assumption that the mapping between IPv4 address 
> and IPv6 prefix is pre-determined, which is a static mapping. But that 
> assumption is not always true. We should allow flexibility for the users to 
> use a dynamic mapping.

yes, MAP has that assumption.

>>> or "stateless-softwire-dhcp" discussing only static provisioning. I'm just 
>>> suggesting "softwire-dhcp" solution.
>> 
>> I would be in favour of having separate drafts, that can be combined to 
>> create more complex solutions.
> 
> [Qi] If one draft can make it clear, I don't think we need a combination.

I think the difference is that MAP-DHCP is really there to provision the 
link-layer (tunnel).
if we follow the principle (that I also stated to Ted), that DHCPv6 is used to 
provision the link-layer,
and DHCPv4 is used to provision the IPv4 layer. then I think the "fallout" of 
that should be
that we'd remove the OPTION_S46_IPV4ADDRESS from the MAP DHCP draft.
are you fine with that approach?

>> (and please don't read from this that I support doing DHCPv4 address 
>> assignment for these mechanisms, I am far from convinced that it is needed).
> 
> [Qi] There are requirements for it. And it follows DHC's consensus. 

I don't think it is appropriate for you to declare DHC WG consensus. in any 
case this would have to be consensus in softwires, wouldn't it?

cheers,
Ole

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires

Reply via email to