Re: [AVTCORE] [xrblock] Open issue- identity info repetition indraft-ietf-avtcore-monarch

Qin,

To my mind, the most important part of the identity block was the correlation tag used to relate RTCP statistics to non-RTCP measurements (draft-clark-avt-rtcphr-02 had something along these lines, although more structured than needed). The tools to save space in RTCP XR packets by removing duplicate identity information when stacking reports are less compelling. Given this, I don't see much value in Section 5.2 of draft-ietf-avtcore-monarch as currently written, and would prefer to see the draft focus on ways to correlate RTP-layer identify information with other identify data.

I'd be interested to hear other opinions though.

Colin

On 10 May 2011, at 03:22, Qin Wu wrote:
> Just make correction. My colleague reminds me that the example figure 2  does not conflict with report block format defined in draft-ietf-avt-rtcp-xr-meas-identity-02.
> Since the detailed measurement identity information following the field SSRC of stream source in the figure 2 is omitted.
> However as I mentioned, it is not clear how much information is included in identity block until identity block is clearly defined in XRBlock WG.
> That is why I propose to remove figure 2 and related text from this document.
>  
> Regards!
> -Qin
> ----- Original Message -----
> From: Qin Wu
> To: avt@ietf.org ; xrblock@ietf.org
> Cc: magnus.westerlund@ericsson.com
> Sent: Monday, May 09, 2011 4:08 PM
> Subject: [AVTCORE] Open issue- identity info repetition indraft-ietf-avtcore-monarch
> 
> In the last IETF meeting, there was discussion on identity info repetition or dupplication. However it
>  was not clear whether new packet type for identity block is needed. Here is some clarification and
> prososal from my point of view.
>  
> In the example figure 2 of section 5.2 in drat-ietf-avtcore-monarch, it is assumed there is an RTCP
>  XR packet containing four metrics blocks, reporting on streams from two source.
> each source is associated with two meric block.
>  
> In order to reduce overhead to carry duplicated data for all the metric block contained in
> the same RTCP XR packet, it was proposed by draft-ietf-avtcore-monarch that measurement
> identity information should be separated from metric block and form independent block.
>  
> However this proposal save zero octets in figure 2. So the question  is
> whether an identity block is actually needed if identity information
> only contain SSRC of stream souce, or if just stacking
> multiple XR Blocks is acceptable.
>  
> However I re-check with RTCP XR Report Block for Measurement Identity defined in
> draft-ietf-avt-rtcp-xr-meas-identity-02, identity information included is more than
> SSRC of stream source. So it should be useful to define new packet type for identity block.
>  
> However the example figure 2(i.e., only contain SSRC of stream souce) is actually not consistent with report block format defined in
> draft-ietf-avt-rtcp-xr-meas-identity-02(i.e., contain more than SSRC of stream source).
>  
> Also as we know, identity block work has just started from XRBlock WG,
> Therefore  I suggest that the example (Figure 2, and the related discussion) is removed until the identity block is clearly
> defined in XRBLock.
>  
> Any Comments or suggestion?
>  
> Regards!
> -Qin
>  
> 
> 
> _______________________________________________
> Audio/Video Transport Core Maintenance
> avt@ietf.org
> https://www.ietf.org/mailman/listinfo/avt
> _______________________________________________
> xrblock mailing list
> xrblock@ietf.org
> https://www.ietf.org/mailman/listinfo/xrblock

-- 
Colin Perkins
http://csperkins.org/