Re: [dhcwg] Fwd: New Version Notification for draft-ietf-dhc-sedhcpv6-20.txt

[Lishan Li <[e-mail]>]

2017-02-16 3:27 GMT+08:00 神明達哉 <[e-mail]>:

> At Wed, 15 Feb 2017 19:46:27 +0800,
> Lishan Li <[e-mail]> wrote:
>
> > > Encapsulated in an Encrypted-Response message?  If so, with setting
> > > transaction-id to 0 (like Reconfigure itself)?  And is the client
> > > supposed to try to decrypt it when transaction-id is 0 even if it
> > > doesn't match any outstanding Encrypted-Query it has sent?  And in
> > > that case which private key should the client use?  (For the last
> > > question, see also the comment on the assumption of the number of
> > > client certificate described in Section 6).
> > >
> > > I think these questions and issues can be addressed relatively easily,
> > > but I believe these should be clearly and explicitly described.
> > >
> > [LS]: How about the following method:
> > The transaction-id of the Encrypted-Reponse message MUST be equal with
> the
> > transaction-id of the received Encrypted-Query message and cannot be
> zero.
> > In this way, the client will try to decrypt it.
>
> Assuming the Reconfigure is encapsulated in an Encrypted-Response
> message: it won't work since in case of Reconfigure there's no
> corresponding message from the client to the server that would be
> encapsulated in the Encrypted-Query message in the first place.
>
[LS]: Yes. It is also the reason why the transaction-id of Reconfigure
message is set to zero. Right?
In this way, if the transaction-id of the Encrypted-Response is set to
zero.
Then in the client behavior part, we should states that: If the
transaction-id is 0, the client also try to decrypt it.
And in the server behavior part, we add the states that: If the DHCPv6
message is Reconfigure message, then the server set the transaction-id of
the Encrypted-Response message to 0.

>
> > > > >   The description of zero hash algorithm field seems to suggest the
> > > > >   signature and hash algorithms are tightly related in practice.
> So
> > > > >   it may be better to define the algorithm option so that
> > > > >   signature/hash algorithms are listed as pairs.
> > > > >
> > > > [LS]: Yes. In some cases, the signature and hash algorithm are
> tightly
> > > > related and cannot be separate. Sorry that I don't understand that
> > > > "it may be better to define the algorithm option so that
> signature/hash
> > > > algorithms are listed as pairs", Could you please explain it more
> > > clearly?
> > >
> > > For example, instead of listing SA-id List and HA-id List separately,
> > > we might have unified SA-id List:
> [...]
> > > and each pair of (SA-id[i], HA-id[i]) will be considered to specify a
> > > specific signature method.
> > >
> > [LS]: So, for every SA-id, there is a corresponding HA-id or multiple
> > HA-id(s).
> > SA-id and HA-id are set to a pair. Right?
>
> In this example, yes.  But I'm not necessarily insisting on this
> particular example itself.  My concern with the currently described
> way is how to address cases like:
>
> - the client includes SA0, SA1, HA0, HA1 in the algorithm option.  It
>   supports the combinations of (SA0, HA0) and (SA1, HA1), but not (SA0,
>   HA1) or (SA1, HA0)
> - the server returns a certificate with the combination of (SA0, HA1)
>   or (SA1, HA0)
>
> So the higher level question is whether and how to address this
> concern.  The above example approach could be one solution to it, but
> it can make the algorithm option unnecessarily larger.
>
[LS]: Agree.

>
> > > > > - Section 6
> > > > >
> > > > >    [...] In this document, it is assumed
> > > > >    that the client uses only one certificate for the encrypted
> DHCPv6
> > > > >    configuration.  So, the corresponding private key is used for
> > > > >    decryption.
> > > > >
> > > > >   I'm not sure if this is okay.  And, in fact, Section 11 talks
> about
> > > > >   a case where multiple key pairs (so multiple certificates) are
> used:
> > >
> > [LS]: We have discussed this issues before.
> > In most cases, a client will not uses multiple certificate(s) for the
> > encrypted DHCPv6 configuration process in the same time. Please see the
> > following reasons:
> > After the client authentication, the client will use the public key in
> the
> > certificate option for encryption. The server only knows one client's
> > public key for encryption. And the client will also only uses one private
> > key for decryption. So there is no cases that the client will use one key
> > pair for the encryption of some DHCPv6 messages and use another key pair
> > for the encryption of another DHCPv6 messages.
>
> This is fine, but I was wondering what if the client has multiple
> DHCPv6 sessions with different DHCPv6 servers using different key
> pairs.  It's probably very rare in practice, so I can live with saying
> such an operation is outside the scope of this spec.  But I'd like to
> see wg consensus on this, and if we reach it I'd like it to be
> explicitly documented.
>
[LS]: Agree. It is necessary to be explicitly documented.