Re: [Dots] WGLC on draft-ietf-dots-signal-call-home-06

Hi Jon, 

Thank you for sharing this cool news. 

Please see inline. 

Cheers,
Med

> -----Message d'origine-----
> De : Dots [mailto:dots-bounces@ietf.org] De la part de Jon Shallow
> Envoyé : mercredi 6 novembre 2019 18:13
> À : 'Valery Smyslov'; dots@ietf.org; dots-chairs@ietf.org
> Objet : Re: [Dots] WGLC on draft-ietf-dots-signal-call-home-06
> 
> Hi All,
> 
> I have done a preliminary working implementation of DOTS Call-Home  - there
> is still some little used functionality to finish off.
> 
> *** Issue # 1 - TCP TLS sessions
> 
> https://tools.ietf.org/html/draft-ietf-dots-signal-call-home-06#section-3.1
> 
> " If UDP transport is used, the Call Home DOTS server begins by
> initiating a DTLS connection to the Call Home DOTS client."
> 
> This works as expected where the UDP + DTLS layer roles are switched.
> 
> " If TCP is used, the Call Home DOTS server begins by initiating a
> TCP connection to the Call Home DOTS client.  Using this TCP
> connection, the Call Home DOTS server initiates a TLS connection
> to the Call Home DOTS client."
> 
> With the current libcoap implementation, I am unable to create a CoAP
> session using TLS on an existing TCP connection - so the DOTS Client is
> unable to accept an incoming TCP session and then initiate TLS integrated
> with the CoAP session. 

[Med] Why the client would initiate the TLS session? Do you want to test a RFC8071-like approach?   

 I agree that RFC8071 only switches the TCP layer
> roles.  

[Med] Yes. 

However all works if the TCP + TLS layer roles are switched - in
> the
> same way that UDP + DTLS work.

[Med] Great. We went for this approach because we preserve the same role for both DTLS and TLS:

              +-----------+                        +-----------+
              | Call Home |                        | Call Home |
              |    DOTS   |                        |    DOTS   |
              |   server  |                        |   client  |
              +-----+-----+                        +-----+-----+
              (D)TLS client                        (D)TLS server  

> 
> Enhancing the libcoap code to support only the TCP layer being switched is
> doable - but have no idea as to whether other CoAP implementations can
> handle this specific requirement.

[Med] Do we really need to do that? 

> 
> *** Not (yet) Implemented #1 - Redirected Signalling
> 
> https://tools.ietf.org/html/draft-ietf-dots-signal-call-home-06#section-
> 3.2.
> 2
> 
> I do not see any issues here.
> 
> *** Not (yet) implemented #2 - New Conflict Cause 4
> 
> I do not see any issues here.
> 
> *** Not (yet) Implemented #3 - Address sharing considerations
> 
> https://tools.ietf.org/html/draft-ietf-dots-signal-call-home-06#section-
> 3.3.
> 2
> 
> I do not see any issues here - just need to interface with the translators
> to get the appropriate information.
> 
> Regards
> 
> Jon
> 
> > -----Original Message-----
> > From: Dots [mailto:ietf-supjps-dots-bounces@ietf.org] On Behalf Of Valery
> > Smyslov
> > Sent: 23 October 2019 08:36
> > To: dots@ietf.org
> > Cc: dots-chairs@ietf.org
> > Subject: [Dots] WGLC on draft-ietf-dots-signal-call-home-06
> >
> > Hi,
> >
> > this message starts a Work Group Last Call (WGLC) for
> draft-ietf-dots-signal-
> > call-home-06.
> > The version to be reviewed is here:
> https://www.ietf.org/id/draft-ietf-dots-
> > signal-call-home-06.txt
> >
> > The WGLC will last for two weeks and will end on November the 7th.
> > Please send your comments to the list before this date.
> >
> > Regards,
> > Frank & Valery.
> >
> >
> > _______________________________________________
> > Dots mailing list
> > Dots@ietf.org
> > https://www.ietf.org/mailman/listinfo/dots
> 
> _______________________________________________
> Dots mailing list
> Dots@ietf.org
> https://www.ietf.org/mailman/listinfo/dots