Ian,

> How about:
> 
> --
> The solution described in this document is suitable for provisioning IPv4
> addressing and other configuration necessary for establishing softwire
> connectivity using DHCPv6. This means that the lifetime of the IPv4
> configuration is bound to the lifetime of the DHCPv6 lease. For MAP-E and
> MAP-T, this is necessary due to the mapping between the IPv4 and the IPv6
> address. Lightweight 4over6 allows for the de-coupling of the IPv4 and
> IPv6 lease times. If this is required, then DHCPv4 over DHCPv6
> [ietf-dhc-dhcpv4-over-dhcpv6] should be used for IPv4 address leasing.
> 
> Additional DHCPv4 options are not transported natively in DHCPv6. If these
> are required for client provisioning, then DHCPINFORM transported in
> DHCPv4 over DHCPv6 should be used.
> ‹
> 
> Does that cover it?

yes, that appears good. I'll add that to the next revision if I don't hear loud 
objections.

cheers,
Ole

> 
>> Ian,
>> 
>>> From a discussion with Bernie and Tomeck earlier: To give some clarity
>>> about what the different 4o6 provisioning mechanisms are suitable for,
>>> can we add in some text to bound the scope of map-dhcp to provisioning
>>> static v4 configuration parameters (i.e. precluding dynamic v4 leasing)
>>> with no additional DHCPv4 options and add in an informative pointer to
>>> using DHCPv4 over DHCPv6 for dynamic/additional options?
>>> 
>>> Likewise, I¹m putting a similar back pointer to MAP-DHCP in the
>>> dhc-v4-configuration draft:
>>> 
>>> For the most simple IPv4 provisioning case, where the client only needs
>>> to receive a static IPv4 address range assignment (with no dynamic
>>> address leasing or additional IPv4 configuration), DHCPv6 based
>>> approaches [ietf-softwire-map-dhcp] may provide a suitable solution.
>>> 
>>> The DHCPv4oDHCPv6 doc should have a similar pointer to map-dhcp for
>>> static as well.
>> 
>> could you propose some text?
>> I'm not quite sure what bounding of scope you'd like to see.
>> all the lifetimes of configuration information defined in MAP DHCP are
>> bounded by the lifetimes of the tunnel,
>> i.e. the lifetime of the End-user IPv6 prefix.
>> 
>> the IPv4 address assignment will be as dynamic as the underlaying IPv6
>> assignment is.
>> 
>> what using DHCPv4 address leases gets you, is separate lease times. given
>> that, this mode is incompatible with MAP-T and -E,
>> I'm not quite sure what this document can say about 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