Hi Henrik. I've probably read that text a hundred times
over the years.
And probably have discussed it with other folks in the
hallway.
If I recall correctly, my very initial interpretation was
similar to
Avi's. Then after spending some more time digesting the
text, I had
concluded with Kuntal's observation. That's why adding a
few more words
for clarification won't hurt and likely helpful to get the
right
understanding for folks reading the specification their
first time.
Kent
> -----Original Message-----
> From: Henrik Levkowetz [mailto:henrik levkowetz.com]
> Sent: Friday, June 02, 2006 11:10 AM
> To: Avi Lior
> Cc: Kent Leung (kleung); Narayanan, Vidya; Vijay
Devarapalli;
> Mobile IPv4 Mailing List
> Subject: Re: [Mip4] Issue with 3344bis
>
> Hi Avi,
>
> on 2006-06-02 15:33 Avi Lior said the following:
> > Hi Henrik,
> >
> > I am not disputing what you are saying at all.
>
> Ok.
>
> > It is just that when we read that sentence it
gives the impression
> > that the HA must be statically provisioned with a
netmask and an SA.
>
> Ok. I accept that you read it that way, but disagree
that it
> actually gives that impression, as a general statement
>
> > Rewording would help....I like Kent's verbage.
>
> Ok.
>
> > And I am not the only one mis-reading this.
>
> I'd like to hear from others, then...
>
> Regards,
>
> Henrik
>
> >> -----Original Message-----
> >> From: Henrik Levkowetz [mailto:henriklevkowetz.com]
> >> Sent: Friday, June 02, 2006 3:24 AM
> >> To: Kent Leung (kleung)
> >> Cc: Narayanan, Vidya; Vijay Devarapalli;
Mobile IPv4 Mailing List;
> >> Avi Lior
> >> Subject: Re: [Mip4] Issue with 3344bis
> >>
> >> Hi Kent,
> >>
> >> The proposed text is ok, but the change is
unnecessary.
> The current
> >> text doesn't say that you have to have MSA
and netmask before you
> >> have an assigned HA, even if Avi happened to
mis-read it that way.
> >>
> >>
> >> Henrik
> >>
> >>
> >> on 2006-06-02 08:13 Kent Leung (kleung) said
the following:
> >> > Here's a proposal:
> >> >
> >> > Change:
> >> >
> >> > "A mobile node MUST be configured
with a netmask and a mobility
> >> > security association for each of its
home agents."
> >> >
> >> > To:
> >> >
> >> > "A mobile node MUST have a netmask
and a mobility
> >> > security association for each of its
home agents. The
> >> information
> >> > can be obtained dynamically by or
configured on the
> >> mobile node."
> >> >
> >> > Kent
> >> >
> >> >> -----Original Message-----
> >> >> From: Avi Lior [mailto:avibridgewatersystems.com]
> >> >> Sent: Thursday, June 01, 2006 10:18
PM
> >> >> To: Narayanan, Vidya; Vijay
Devarapalli
> >> >> Cc: Mobile IPv4 Mailing List
> >> >> Subject: RE: [Mip4] Issue with
3344bis
> >> >>
> >> >> ViJay,
> >> >>
> >> >> My original emails suggested that
perhaps we say:
> >> >>
> >> >> Instead it should say that it MAY be
configured with .....
> >> >>
> >> >> or MUST be configurable with....
> >> >>
> >> >> Either of those cases support the
manual or static
> >> configuration and
> >> >> also the Dynamic configuration.
> >> >>
> >> >> The current wording only allows for
static configuration
> >> of the MN-HA.
> >> >>
> >> >> I agree with Vidya about the comments
regarding your
> >> additional text.
> >> >> It is not necessary.
> >> >>
> >> >> > -----Original Message-----
> >> >> > From: Narayanan, Vidya
[mailto:vidyanqualcomm.com]
> >> >> > Sent: Friday, June 02, 2006
12:15 AM
> >> >> > To: Vijay Devarapalli; Avi Lior
> >> >> > Cc: Mobile IPv4 Mailing List
> >> >> > Subject: RE: [Mip4] Issue with
3344bis
> >> >> >
> >> >> >
> >> >> > >
> >> >> > > Avi Lior wrote:
> >> >> > > > Hi ViJay,
> >> >> > > >
> >> >> > > > Sorry but I don't
think I agree with the MUST.
> If I have a
> >> >> > > > method for
bootstrapping MN-HA why MUST the MN be
> >> configured with an MN-HA?
> >> >> > > >
> >> >> > > > I think we can come up
with some text without
> specifically
> >> >> > > > making a reference to
4433 and 3957.
> >> >> > > >
> >> >> > > > The problem is that
the "MUST" requires that I
> configure at
> >> >> > > > least one MN-HA to be
compliant with 3344.
> >> >> > >
> >> >> > > I don't see it that way.
3344bis says there MUST be
> a security
> >> >> > > association between the MN
and the HA for the
> protocol to work.
> >> >> > > Since it does not specify a
way to dynamically configure a
> >> >> > > security association it
*has* to say that a security
> >> association
> >> >> > > MUST be configured on the
MN. otherwise it cannot be
> >> standardized.
> >> >> > >
> >> >> >
> >> >> > I agree with Avi on this. I
don't see why there needs to
> >> be a MUST
> >> >> > on this one. If you say that an
MN-HA SA MUST be
> available to
> >> >> > secure the RRQ/RRP, that makes
sense - it does not
> make sense to
> >> >> > say that this MUST be
"configured" per say.
> >> >> >
> >> >> > > I think the MUST is needed
in the spec.
> >> >> > >
> >> >> > > perhaps we could add the
following at the end of
> the paragraph.
> >> >> > >
> >> >> > > Future specifications
may define mechanisms to
> dynamically
> >> >> > > assign a home agent to a
mobile node and setup a mobility
> >> >> > > security association
between them. In this case,
> the mobile
> >> >> > > node need not be
manually configured with a netmask or a
> >> >> > > mobility security
association with its home agent.
> >> >> > >
> >> >> >
> >> >> > But, we already have
specifications that define this 
> >> So, to keep
> >> >> > our eyes closed to the fact that
we already have ways of
> >> getting
> >> >> > away from manual configuration
does not seem to make
> sense to me.
> >> >> >
> >> >> > If you want to say that in the
absence of a dynamic
> mechanism to
> >> >> > acquire an MN-HA SA, it MUST be
configured on the MN,
> >> that would be
> >> >> > fine.
> >> >> >
> >> >> > Vidya
> >> >> >
> >> >> > > Vijay
> >> >> > >
> >> >> > > >
> >> >> > > >> -----Original
Message-----
> >> >> > > >> From: Vijay
Devarapalli
> >> >> > > >>
[mailto:vijay.devarapalliAzaireNet.com]
> >> >> > > >> Sent: Thursday,
June 01, 2006 6:06 PM
> >> >> > > >> To: Avi Lior
> >> >> > > >> Cc: Mobile IPv4
Mailing List
> >> >> > > >> Subject: Re:
[Mip4] Issue with 3344bis
> >> >> > > >>
> >> >> > > >> Avi Lior wrote:
> >> >> > > >>>
> >> >> > > >>> Sorry for the
lateness of this comment...
> >> >> > > >>>
> >> >> > > >>> Section 3.6
> >> >> > > >>>
> >> >> > > >>> "A
mobile node MUST be configured with a netmask
> >> and a mobility
> >> >> > > >>> security
association for each of its home agents."
> >> >> > > >>>
> >> >> > > >>> This seems to
be a bad thing if we are talking
> >> about roaming
> >> >> > > >>> and dynamic HA
assignement.
> >> >> > > >>>
> >> >> > > >> I think this is
ok. as far as 3344bis is concerned,
> >> at least
> >> >> > > >> the netmask for
the home link and a mobility security
> >> >> > > >> association is
required for the protocol to work. later
> >> >> > > >> specifications
talk about how the protocol in
> 3344bis would
> >> >> > > >> work without
having to pre-configere a netmask
> >> (RFC4433), or
> >> >> > > >> without having to
pre-configure a mobility security
> >> association (for e.g. RFC 3957).
> >> >> > > >>
> >> >> > > >> since there is a
plan to advance 3344bis to draft
> >> standard we
> >> >> > > >> don't want
normative references to RFC 4433 or RFC 3957.
> >> >> > > >>
> >> >> > > >> Vijay
> >> >> > > >>
> >> >> > >
> >> >> > >
> >> >> > > --
> >> >> > > Mip4 mailing list: Mip4ietf.org
> >> >> > > Web interface:
> https://w
ww1.ietf.org/mailman/listinfo/mip4
> >> >> > > Charter page:
> >> >> > > h
ttp://www.ietf.org/html.charters/mip4-charter.html
> >> >> > > Supplemental site: http://www.mip4.org/
> >> >> > >
> >> >> >
> >> >>
> >> >> --
> >> >> Mip4 mailing list: Mip4ietf.org
> >> >> Web interface: https://w
ww1.ietf.org/mailman/listinfo/mip4
> >> >> Charter page:
> >> >> h
ttp://www.ietf.org/html.charters/mip4-charter.html
> >> >> Supplemental site: http://www.mip4.org/
> >> >>
> >> >
> >>
> >>
> >
>
>
--
Mip4 mailing list: Mip4ietf.org
Web interface: https://w
ww1.ietf.org/mailman/listinfo/mip4
Charter page: h
ttp://www.ietf.org/html.charters/mip4-charter.html
Supplemental site: http://www.mip4.org/
|