Re: [AVTCORE] WGLC on draft-ietf-avtcore-ecn-for-rtp-04 - more

[Roni Even <[e-mail]>]

Magnus,
I am OK with your answers
Roni

> -----Original Message-----
> From: Magnus Westerlund [mailto:[e-mail]]
> Sent: Thursday, October 27, 2011 3:36 PM
> To: Roni Even
> Cc: [e-mail]
> Subject: Re: [AVTCORE] WGLC on draft-ietf-avtcore-ecn-for-rtp-04 - more
> 
> Hi Roni,
> 
> Thanks for the feedback.
> 
> 
> On 2011-10-22 14:05, Roni Even wrote:
> >
> > 1.      In section 7.2  "All session participants connected over the
> > same transport   will need to use the same initiation method." maybe
> say
> > MUST use instead of will need to use.
> 
> sure.
> 
> >
> > 2.       In section 7.2 " If support for both in-band and out-of-band
> > mechanisms is signalled, the sender SHOULD try ECN negotiation using
> > STUN with ICE first, and if it fails, fallback to negotiation using
> RTP
> > and RTCP ECN feedback" I think that it should say that the sender
> should
> > offer ice initialization method before rtp.
> 
> I would suggest to formulate this as:
> 
> If support for both in-band and out-of-band mechanisms are signalled,
> the sender when negotiating SHOULD offer detection of ECT using STUN
> with ICE with higher priority than detection of ECT using RTP and RTCP.
> 
> 
> >
> > 3.       In section 7.2.1 y ou refer not reference the keep-alive RFC
> 
> Why do you think we should reference RFC 6263? Is is because we discuss
> which packets would have least impact on the media flow when used as
> ECT
> probes? IF that is the reasons I don't think RFC6263 has much to offer.
> The outcome of this document is that nothing short of real RTP or RTCP
> packets work well for keep-alive. And in this context of ECN we can't
> use the RTCP packets.
> 
> >
> > 4.       In section 7.2.1 Generating RTCP ECN Feedback it says
> > "Reception of subsequent ECN-CE marked packets MUST  result in
> > additional early or immediate ECN". Why use MUST and not SHOULD since
> > there is an exception case following the sentence.
> 
> Because if we use SHOULD then one might consider that additional
> exceptions than just that one exists. And that is not the case. If you
> use a method that requires timely feedback, then you better send that
> feedback.
> 
> >
> > 5.       In section 7.2.1 "ECN initiation is considered to have
> failed
> > at the instant when a ny RTP s an RTCP packet that doesn't  contain
> an
> > RTCP ECN feedback report or ECN summary report " The RTCP ECN
> feedback
> > report may be missing for other reasons like no timely response or
> using
> > AVP profile.
> 
> I think you are misunderstanding what was written, maybe because of the
> bad formulation in the sentence. I have rewritten this to:
> 
> ECN initiation is considered to have failed at the instant the
> initiating RTP sender received an RTCP packet that doesn't contain an
> RTCP ECN feedback report or ECN summary report from any RTP session
> participant that has an RTCP RR with an extended RTP sequence number
> field that indicates that it should have received multiple (>3) ECT
> marked RTP packets.
> 
> Is that clearer?
> 
> 
> >
> > 6.       In section 7.3.3 receiver driven congestion control you
> mention
> > layered codecs as an option I think that using ccm tmbbr is more
> useful
> > when there are no layered codecs in use (more common today) and
> should
> > be mentioned in this section.
> 
> I am very split about discussing this. I know there exist
> implementations that uses TMMBR as primary congestion control. However,
> RFC 5104 hasn't specified it in that way.
> 
> I also see an issue with bringing it up in this context, namely that we
> bless this without any specification or requirements on the solution.
> For example how responsive to congestion is the usage of TMMBR. One
> parameter is the  AVPF operations mode and due to this RTCP bandwidths.
> another is the algorithm determining the appropriate bit-rate.
> 
> I know that the multicast scalable encoding share similar concerns but
> there we have even less alternatives.
> 
> I think reformulating the introduction of the section to make clear
> that
> this is just an example of Receiver driven congestion control is most
> appropriate.
> 
> 
> In a receiver driven congestion control mechanism, the receivers can
> react to the ECN-CE marks themselves without providing ECN-CE feedback
> to the sender. This may allow faster response than sender-driven
> congestion control in some circumstances and also scale to large number
> of receivers and multicast usage. One example of receiver-driven
> congestion control is implemented by providing the content in a layered
> way, with each layer providing improved media quality but also
> increased
> bandwidth usage.
> 
> >
> > 7.&nbs ;  In section 10.6 you use the wrong range definition "From
> RFC
> > 5389: STUN Attribute types in the first half of the
> > comprehension-requiredrange (0x0000 - 0x3FFF) and in the first half
> of
> > the comprehension- optional range (0x8000 - 0xBFFF) are assigned by
> IETF
> > Review"
> >
> 
> Good catch.
> 
> Thanks
> 
> Magnus Westerlund
> 
> ----------------------------------------------------------------------
> Multimedia Technologies, Ericsson Research EAB/TVM
> ----------------------------------------------------------------------
> Ericsson AB                | Phone  [phone]
> Färögatan 6                | Mobile [phone]
> SE-164 80 Stockholm, Sweden| mailto: [e-mail]
> ----------------------------------------------------------------------