Re: [HOKEY] Oversight in key management draft

["Hoeper Katrin-QWKN37" <khoeper at motorola.com>]

Glen, all

I just posted a new version (draft-ietf-hokey-key-mgm-08.txt) addressing
Glen's last comments about the key type.

I believe all last call comments have been addressed and the draft is
ready for IESG now.

Regards,
Katrin




> -----Original Message-----
> From: Glen Zorn [mailto:gwz at net-zen.net]
> Sent: Tuesday, July 07, 2009 9:39 AM
> To: Hoeper Katrin-QWKN37
> Cc: hokey at ietf.org
> Subject: RE: Oversight in key management draft
> 
> Hoeper Katrin-QWKN37 [mailto:khoeper at motorola.com] writes:
> 
> > > Glen Zorn [mailto:gwz at net-zen.net] writes:
> > >
> > > > Hi, Katrin.  While doing a final review as part of the document
> > > > shepherding
> > > > process, I noticed what I think is an oversight that may have
> > occurred
> > > > while
> > > > removing the RADIUS-specific stuff from the draft: section 4
says
> > "The
> > > > following key types are defined: 0 (DSRK), 1 (USRK) and 2
> > (DSUSRK)."
> > > > This
> > > > seems out of place now.
> > >
> > [KH]
> > The Key Request Token (KRT) carries (among other things) the key
type
> > of
> > the requested key. In the draft we specify three keys that can be
> > requested/delivered, i.e. DSRK, USRK and DSUSRK.
> >
> > For that reason I thought it still makes sense to assign those
numbers
> > to identify these keys, independent of which AAA protocol is used.
> 
> OK.  Why?
> 
> >
> > Do you think it should be left to the implementers to define the key
> > types?
> 
> I think that the key types have been defined in other documents,
including
> types not mentioned in this note; I'm not sure of the benefit of
having a
> single set of numbers across all AAA protocols, but that's just me.
> 
> > If so, should we provide any guidance on this in the draft?
> >
> > > As does the IANA Considerations section & BTW, remind me why this
is
> > > Standards Track instead of a BCP...
> > >
> > [KH] Sorry, I cannot remind you,
> 
> Sorry, it was more a question for the WG at large; I wasn't trying to
put
> you on the spot.
> 
> > because when I became the new editor
> > of
> > this document this decision was already made and I cannot recall any
> > discussions on this topic. What do previous contributors think about
> > this?
> >
> > I think changing it to BCP might be a good idea, given that we do
not
> > know any current implementations that use the KDE protocol defined
in
> > the draft.
> 
> Well, the protocol is (rightly, I think) quite abstract, but other
work is
> being done (e.g.,
>
http://www.ietf.org/internet-drafts/draft-wu-dime-local-keytran-02.txt).
> 
> >
> >
> > > > If it is an error, do you want to fix it now
> > > > or
> > > > wait for IETF Last Call?
> > > >
> > > > ~gwz
> > > >
> > > > Play assigns meaning to human activity--work erases it.
> > > >   -- P.L. Wilson
> > > >
> > > >
> > > >
> > >
> >
>