Re: [dhcwg] RE: [Mip6] [Updated] DHCP Option forHomeInformationDiscovery inMIPv6

[Julien Bournelle <[e-mail]>]

Hi all,

On Wed, Jul 05, 2006 at 05:58:42PM -0700, Vijay Devarapalli wrote:
> Alper,
> 
> we don't define architectures in the IETF. 
> draft-ietf-mip6-bootstrapping-integrated-dhc is supposed to be the only 
> bootstrapping *solution* for the integrated scenario. thats the reason 
> it is currently a WG document and draft-jang isnt.

 I'm not sure that the intent was to be the only bootstrapping solution
 for the integrated scenario. We have added the suffix -dhc to specify
 that the delivery of HA information to the MN was based on DHCP but
 others solutions may be possible (cf.
 draft-bournelle-pana-mip6-01.txt).

 My 2 cents,

 Julien

> 
> lets continue this discussion at the IETF meeting. an offline meeting 
> with the working group chairs before the MIP6 WG meeting would help, I 
> think.
> 
> Vijay
> 
> Alper Yegin wrote:
> >>there is one issue.
> >>
> >>draft-jang just does not define the DHCP options. it goes further.
> >>abstract of draft-jang says
> >>
> >>    This draft defines a DHCP-based scheme to enable dynamic discovery of
> >>    Mobile IPv6 home agent address, home address, and home subnet.  New
> >>    DHCP options are defined to carry the information from a DHCP server
> >>    to the DHCP client running on the mobile node.
> >
> >What is "further" in this? The words "dynamic discovery?" This is what DHCP
> >gives you for anything that you use it for.
> >
> >
> >>section 4.2 further goes and ahead and specifies mobile node
> >>behavior. the draft just does not define the options. 
> >
> >These are all within the scope of defining a new DHCP option.
> >
> >>it is also
> >>incomplete because it neither describes the scope of the problem
> >
> >This is all defined in the Introduction section. If you have specific
> >comments, let us know and we'll fix it up.
> >
> >>it is trying to address nor gives a complete solution.
> >
> >Complete solution, I suspect, is the architecture you are referring to.
> >Again, this is a "protocol" document, not an "architecture" document.
> >
> >>we can't have two documents that both describe how to use DHCP
> >>for MIP6 bootstrapping.
> >
> >That's not happening. draft-jang-mip6-hiopt-00 defines a protocol
> >specification. draft-ietf-mip6-bootstrapping-integrated-dhc-01 defines the
> >architecture and refers to the former for the specific protocol details.
> >
> >From your other e-mail:
> >
> >>>If the use cases are not limited to only bootstrapping,
> >
> >>what other use cases? are those use cases specific to Mobile IPv6?
> >>do they fit into the MIP6 WG charter?
> >
> >I had explained the use of statically configured DHCP servers earlier,
> >remember?
> >
> >....
> >
> >Vijay, we have circled around your concerns many times by now. I'm going to
> >stop now and see you in Montreal.
> >
> >Alper
> >
> >
> >
> >
> >>Vijay
> >>
> >>Kent Leung (kleung) wrote:
> >>>Hi Kuntal and Vijay.
> >>>
> >>>See my posting, "Generalizing draft-jang-dhc-haopt for MIPv6 support in
> >>>DHCP", back on Jan 4, 2006 on this topic.  I proposed to keep the 2
> >>>drafts separate, one for defining DHCPv6 opcodes for MIPv6 parameters
> >>>and a bootstrap draft that references the DHCP options needed to achieve
> >>>its function.
> >>>
> >>>I agree there's no need to drag in references other to say that there's
> >>>value for the DHCP MIPv6 options.  3GPP2/WiMAX specs can reference the
> >>>DHCPv6 opcodes for MIPv6 draft/RFC, much like what MIPv6 bootstrap would
> >>>have done.
> >>>
> >>>As for what Vijay said about the WiMAX scheme, definitely true that it
> >>>doesn't conform to DHCP RFC 3315.  But we should keep the issues
> >>>separate.  I haven't read all the emails on the thread.  Just jumping in
> >>>to say the following:
> >>>
> >>> - Keep 2 separate self contained drafts
> >>> - IETF MIPv6 bootstrap draft references DHCP for MIPv6 draft
> >>> - WiMAX and 3GPP2 can also reference DHCP for MIPv6 draft for their
> >>>bootstrapping schemes
> >>>
> >>>Just some explanatory comments:
> >>> - 3GPP2 conforms to DHCPv6 in its bootstrapping scheme
> >>> - WiMAX does not conform to DHCPv6 when using proxy DHCP server scheme
> >>>
> >>>But *ALL* of the above bootstrapping schemes use the same DHCP options
> >>>for MIPv6 parameters.  That's what justifies an independent DHCP for
> >>>MIPv6 draft, IMHO.
> >>>
> >>>Kent
> >>>
> >>>
> >>>
> >>>
> >>>>-----Original Message-----
> >>>>From: Chowdhury, Kuntal [mailto:[e-mail]]
> >>>>Sent: Thursday, June 22, 2006 7:19 PM
> >>>>To: Vijay Devarapalli; Behcet Sarikaya
> >>>>Cc: [e-mail]; Heejin Jang; [e-mail]
> >>>>Subject: RE: [dhcwg] RE: [Mip6] [Updated] DHCP Option for
> >>>>HomeInformationDiscovery inMIPv6
> >>>>
> >>>>Sorry, I am not getting this. Why are we dragging WiMAX and
> >>>>3GPP2 references to justify the need for basic DHCP op-codes
> >>>>for HoA/HL/HA info?
> >>>>
> >>>>draft-jang-mip6-hiopt-00.txt defines the DHCPv6 op-codes for
> >>>>HA, HoA and HL info.
> >>>>
> >>>>draft-ietf-mip6-bootstrapping-integrated-dhc-01.txt is the
> >>>>MIP6 bootstrapping solution that leverages DHCPv6 for HA/HL
> >>>>bootstrapping.
> >>>>
> >>>>There are tons of I-Ds and solutions in IETF that leverage other I-Ds.
> >>>>So, what is so special here? The decision to NOT merge these
> >>>>I-Ds was taken in the bootstrap DT. Why are we wasting
> >>>>people's time here to revisit basic decisions?
> >>>>
> >>>>-Kuntal
> >>>>
> >>>>
> >>>>
> >>>>>-----Original Message-----
> >>>>>From: Vijay Devarapalli [mailto:[e-mail]]
> >>>>>Sent: Thursday, June 22, 2006 12:55 PM
> >>>>>To: Behcet Sarikaya
> >>>>>Cc: [e-mail]; Heejin Jang; [e-mail]
> >>>>>Subject: Re: [dhcwg] RE: [Mip6] [Updated] DHCP Option for Home
> >>>>>InformationDiscovery inMIPv6
> >>>>>
> >>>>>Behcet Sarikaya wrote:
> >>>>>
> >>>>>>Hi Vijay,
> >>>>>>The idea in the present WiMAX document is quite simple:
> >>>>>>it uses EAP-PKMv2 (which is like 802.11i) authentication
> >>>>and during
> >>>>EAP
> >>>>
> >>>>>>transaction, they assume that AAA server determines that
> >>>>the MS is a
> >>>>
> >>>>>>CMIPv6 authorized user based on the AAA profile and it
> >>>>will include
> >>>>all
> >>>>
> >>>>>>the so-called bootstrapping info in the Access-Accept. Then MS can
> >>>>use
> >>>>
> >>>>>>RFC 3736 to make a DHCP Information Request to DHCP Relay
> >>>>colocated
> >>>>with
> >>>>
> >>>>>>NAS and use draft-jang (extended now with HoA) and receive all the
> >>>>info
> >>>>
> >>>>>>to configure MIP6.
> >>>>>the DHCP Relay replies to the MN without relaying the DHCP request
> >>>>>from the MN to a DHCP server? this is not allowed with the current
> >>>>>IETF specs.
> >>>>>
> >>>>>
> >>>>>> In this setup can you make a case for
> >>>>>>draft-ietf-mip6-bootstrapping-integrated-dhc-01.txt?
> >>>>>as far as I am concerned,
> >>>>>draft-ietf-mip6-bootstrapping-integrated-dhc-01.txt
> >>>>>is the document that describes how to use DHCPv6 for MIPv6
> >>>>>bootstrapping. as for the setup in the WiMAX NWG, it is
> >>>>something that
> >>>>
> >>>>>doesn't conform to the IETF DHCP specs.
> >>>>>
> >>>>>Vijay
> >>>>><snip>
> >>
> >>_______________________________________________
> >>Mip6 mailing list
> >>[e-mail]
> >>https://www1.ietf.org/mailman/listinfo/mip6
> >
> 
> 
> _______________________________________________
> Mip6 mailing list
> [e-mail]
> https://www1.ietf.org/mailman/listinfo/mip6

-- 
julien.bournelle at int-evry.fr

_______________________________________________
Mip6 mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/mip6