Re: [ippm] [lmap] I-D Action: draft-ietf-lmap-framework-05.txt

Dear Authors, et. al,

Please see my comments for the latest version of LMAP framework document below. I also include IPPM WG in the thread since I believe its valuable.

1. Active Measurement Method
I like the way Greg define the Active Measurement Method which I quote below. To me, it is more intuitive to define the method as the process of measuring on traffic than referring to the measurement task.
I also agree that the Active Measurement Traffic could also be updated accordingly. 

Active Measurement Method - The process of measuring some performance or reliability parameter associated with the transfer of Traffic by generating and/or receiving Active Measurement Traffic.

2, Passive Measurement Method
The current definition says: Passive Measurement Method (Task)- A Measurement Method (Task) in which a Measurement Agent observes existing traffic but does not inject Active Measurement Traffic.

The word "observe" exclude many existing measurement schemes, which are currently discussed in IPPM, as passive measurement method. My suggestion is to define the Passive Measurement Method as following:

Passive Measurement Method- The process of measuring some performance or reliability parameter associated with the existing traffic on the network.

This definition may need to be honed, but it makes a clear line between active and passive. For active, you are measuring on the injected traffic. For passive, on the other hand, you are measuring on the existing traffic on the network.
The key point for passive measurement, imho, is you are measuring on the REAL traffic on the network. For specific method, you may do something more than just observing. Examples can be found http://tools.ietf.org/html/draft-elkins-ippm-pdm-metrics-04 and http://tools.ietf.org/html/draft-chen-ippm-coloring-based-ipfpm-framework-01. But as long as you are measuring on the existing traffic, your method should be consider as passive.

Another benefit of this definition is it avoids the concept of "pure passive", which has never be properly defined and vague.
Someone may consider these so called not "pure passive" measurement method as hybrid. But in that case the Hybrid Measurement Method itself need to be properly defined first. 
The discussion about hybrid happened in IPPM mainly about how to combining active and passive. It mentions some form of hybrid, such as "combined hybrid" and "concurrent hybrid", which I don't see these example methods fall in any of the category.

People in IPPM is working on the Framework of Passive Measurements, we will submit the document soon. The definition of passive measurements is very important and the coordination between two working group is essential. I believe now is the key time to have both working group looking on it.
I would like the authors consider my suggestion on the definition and encourage discussion in the working group.

BR,
Vero

> -----Original Message-----
> From: lmap [mailto:lmap-bounces@ietf.org] On Behalf Of
> philip.eardley@bt.com
> Sent: Thursday, May 22, 2014 6:10 AM
> To: gregimirsky@gmail.com; lmap@ietf.org;
> draft-ietf-lmap-framework@tools.ietf.org
> Subject: Re: [lmap] I-D Action: draft-ietf-lmap-framework-05.txt
> 
> Greg,
> thanks for the reading!
> 
> must vs MUST
> initial LMAP work vs for future study
> 
> suggest leave both of these issues for the AD /IESG / RFC editor to see what
> they prefer.
> 
> configuration - thanks for spotting, sorry this is my error. it should be
> something that the Control protocol does.
> 
> definition of active measurement method & task. perhaps this could be honed.
> would like to understand your comment about "coordination is optional" a bit
> better.
>  an active measurement task involves the MA measuring traffic that's been
> sent by (*) another MA, or a measurement peer. (*) or received from.  there
> needs to be some kind of coordination to arrange for this traffic to be sent. is
> the problem that the current words could be read to imply that the
> coordination must be directly between the MAs, whereas it could be that the
> controller sends a task/schedule instruction to one MA and a report instruction
> to another MA (as in example A3 in the appendix)?
> 
> thanks
> phil
> 
> ________________________________________
> From: lmap [lmap-bounces@ietf.org] On Behalf Of Greg Mirsky
> [gregimirsky@gmail.com]
> Sent: 16 May 2014 01:16
> To: lmap@ietf.org; draft-ietf-lmap-framework@tools.ietf.org
> Subject: Re: [lmap] I-D Action: draft-ietf-lmap-framework-05.txt
> 
> Dear Authors, et. al,
> please find my comments to the latest updates of the LMAP Framework
> document below:
> 
>  *   Use of RFC 2119. Even though the Framework document is on
> Informational track RFC 2119 may be used. Hence my question: Do we want to
> turn 'must' into MUST and so on in the document?
>  *   Section 3. I think that definition of an Active Measurement Method only
> as "A generalization of an Active Measurement Task" is not sufficiently clear.
> Perhaps the following may be used instead by re-using or referring to
> explanation of Measurement Method:
> 
> Active Measurement Method - The process of measuring some performance or
> reliability parameter associated with the transfer of Traffic by generating
> and/or receiving Active Measurement Traffic.
> 
>  *   If definition of the Active Measurement Method references use of Active
> Measurement Traffic, then the current definition can be updated to explain
> that Active Measurement Task is realization of Active Measurement Method
> among participating Measurement Agents and Measurement Peers with
> specific parameters, e.g. profile of the Active Measurement Traffic.
> 
> And I believe that coordination of Active Measurement Task among
> participating MAs and/or MPs is optional. The current definition reads as it is
> mandatory. I suggest to change it by splitting the sentence and making it to say
> "Coordination of an Active Measurement Task among participating
> Measurement Agents and/or Measurement Peers may be achieved by using
> protocols. Definition of such protocols by LMAP WG is for further study."
> 
> 
>  *   I don't see much difference between Configuration Protocol and Control
> Protocol. I think that the former supports subset of functionality of the latter.
> Besides, it is not clear whether Configuration Protocol uses Control Channel
> between Controller and a MA or not. I think this differentiation between
> Configuration and Control is unnecessary.
>  *   Reference to "initial LMAP work" may be too vague as charter of the
> LMAP WG will likely to change with time. Can it be changed to "NNN is for
> future study"?
>  *   Section 5.2 been introduced in this version. Firstly, I don't recall that the
> WG discussed Configuration Protocol and agreed that it is essential part of the
> LMAP Framework. Adding new concepts, IMO, doesn't help to pass WG LC.
> 
> Regards,
> Greg
> 
> 
> On Tue, May 13, 2014 at 10:21 AM,
> <internet-drafts@ietf.org<mailto:internet-drafts@ietf.org>> wrote:
> 
> A New Internet-Draft is available from the on-line Internet-Drafts directories.
>  This draft is a work item of the Large-Scale Measurement of Broadband
> Performance Working Group of the IETF.
> 
>         Title           : A framework for large-scale measurement
> platforms (LMAP)
>         Authors         : Philip Eardley
>                           Al Morton
>                           Marcelo Bagnulo
>                           Trevor Burbridge
>                           Paul Aitken
>                           Aamer Akhter
>         Filename        : draft-ietf-lmap-framework-05.txt
>         Pages           : 54
>         Date            : 2014-05-13
> 
> Abstract:
>    Measuring broadband service on a large scale requires a description
>    of the logical architecture and standardisation of the key protocols
>    that coordinate interactions between the components.  The document
>    presents an overall framework for large-scale measurements.  It also
>    defines terminology for LMAP (large-scale measurement platforms).
> 
> 
> The IETF datatracker status page for this draft is:
> https://datatracker.ietf.org/doc/draft-ietf-lmap-framework/
> 
> There's also a htmlized version available at:
> http://tools.ietf.org/html/draft-ietf-lmap-framework-05
> 
> A diff from the previous version is available at:
> http://www.ietf.org/rfcdiff?url2=draft-ietf-lmap-framework-05
> 
> 
> Please note that it may take a couple of minutes from the time of submission
> until the htmlized version and diff are available at
> tools.ietf.org<http://tools.ietf.org>.
> 
> Internet-Drafts are also available by anonymous FTP at:
> ftp://ftp.ietf.org/internet-drafts/
> 
> _______________________________________________
> lmap mailing list
> lmap@ietf.org<mailto:lmap@ietf.org>
> https://www.ietf.org/mailman/listinfo/lmap
> 
> _______________________________________________
> lmap mailing list
> lmap@ietf.org
> https://www.ietf.org/mailman/listinfo/lmap