Re: [AVT] WGLC comments on draft-ietf-avt-ports-for-ucast-mcast-rtp-11: Part 3 Binding of C2->P4 RTCP to correct RTP session

"Ali C. Begen (abegen)" <abegen@cisco.com> Fri, 14 January 2011 18:31 UTC

Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable
Date: Fri, 14 Jan 2011 10:33:03 -0800
Message-ID: <04CAD96D4C5A3D48B1919248A8FE0D540E16DA3A@xmb-sjc-215.amer.cisco.com>
In-Reply-To: <4D307CF7.6040508@ericsson.com>
Thread-Topic: WGLC comments on draft-ietf-avt-ports-for-ucast-mcast-rtp-11: Part 3 Binding of C2->P4 RTCP to correct RTP session
Thread-Index: Acu0Cg4HuFZP+NotRq+eEK6A1Px+YgAC7uDw
References: <4D307CF7.6040508@ericsson.com>
From: "Ali C. Begen (abegen)" <abegen@cisco.com>
To: Magnus Westerlund <magnus.westerlund@ericsson.com>, IETF AVT WG <avt@ietf.org>, draft-ietf-avt-ports-for-ucast-mcast-rtp@tools.ietf.org
Subject: Re: [AVT] WGLC comments on draft-ietf-avt-ports-for-ucast-mcast-rtp-11: Part 3 Binding of C2->P4 RTCP to correct RTP session
Precedence: list

> -----Original Message-----
> From: Magnus Westerlund [mailto:magnus.westerlund@ericsson.com]
> Sent: Friday, January 14, 2011 5:43 PM
> To: IETF AVT WG; draft-ietf-avt-ports-for-ucast-mcast-rtp@tools.ietf.org
> Subject: WGLC comments on draft-ietf-avt-ports-for-ucast-mcast-rtp-11: Part 3 Binding of C2->P4 RTCP to correct RTP
> session
> 
> Hi,
> 
> This third issue was also brought up before and also Tom Van Caenegem
> comment on it.
> 
> The RTCP reports going from C2->P4 from a receiver is not obvious on how
> to correlate with the Unicast RTP session it is reporting on that are
> sent P3->C1.
> 
> The issues that may arise are the following. First there might be a NAT
> with port and address dependent mapping between them. Thus, even if
> C2=C1, the externally seen address+port can be different between traffic
> going to P3 and P4. In addition we shall not forget that there might be
> multiple clients behind the same NAT that participate in the same
> multicast session. So several RTCP flow might go to P4 from this IP
> address.

Correct, there could exist multiple clients sending rtcp to p4.
 
> I think some details on how to correlate this RTCP flow needs to be
> included in the document. What are the determining factors? SSRC can't
> be it unless we couple the multicast and unicast session. Which I think
> is a bad idea. CNAME is likely the only method that is directly
> applicable and would match from the C1->P3 request to the C2->P4 one.

Yes, via a CNAME. And since we now have a good CNAME guideline document, this should work just fine.

Report correlation is about the individual unicast sessions, but I agree it is a good idea to state this explicitly in this draft, too and also require the new CNAME guideline document so that cnames are unique.

> But this needs to be stated as a requirement.

+1.
-acbegen
 
> 
> Cheers
> 
> Magnus Westerlund
> 
> ----------------------------------------------------------------------
> Multimedia Technologies, Ericsson Research EAB/TVM
> ----------------------------------------------------------------------
> Ericsson AB                | Phone  +46 10 7148287
> Färögatan 6                | Mobile +46 73 0949079
> SE-164 80 Stockholm, Sweden| mailto: magnus.westerlund@ericsson.com
> ----------------------------------------------------------------------

[AVT] WGLC comments on draft-ietf-avt-ports-for-u… Magnus Westerlund
Re: [AVT] WGLC comments on draft-ietf-avt-ports-f… Ali C. Begen (abegen)
Re: [AVT] WGLC comments on draft-ietf-avt-ports-f… Magnus Westerlund