Re: [ippm] Call for adoption: milestone on coloring-based loss and delay methodologies

[Brian Trammell <[e-mail]>]

hi Mach, all,

A few suggestions, as an individual contributor:

> On 18 Apr 2016, at 11:10, Mach Chen <[e-mail]> wrote:
> 
> Hi Brian,
> 
> Firstly, I support the adoption of draft-tempia-ippm-p3m.
> 
> BTW, I would argue that draft-chen-ippm-coloring should be in the scope of the current charter.
> 
>> Additional methods will be defined for the composition and calibration
>> of IPPM-defined metrics, as well as active, passive and hybrid
>> measurement methods for these metrics. In addition, the WG encourages
>> work which describes the applicability of metrics and measurement
>> methods, especially to improve understanding of the tradeoffs involved
>> among active, passive, and hybrid methods.
> 
> This is the charter text you quoted, IMHO, the second part is just what draft-chen-ippm-coloring is doing, which defines "the applicability of metrics and measurement methods".

Parts of the document do indeed discuss measurement method applicability; however, draft-chen-ippm-coloring-based-ipfpm-framework goes much further than that. It prescribes an architecture of centralized coordination of a set of measurement agents (which may or may not be Measurement Agents in the LMAP sense; my understanding of the current state of the LMAP work indicates that the information model there might not be applicable to hybrid measurements as described in these documents at this point in time).

> It applies the marking-based idea to P2P, P2MP, MP2P scenarios, it also defines how to correlate the metrics from ingress MAs and egress MAs, how to handle packet re-ordering, synchronization, etc., these are fundamental to the marking based loss and delay measurement.

Agreed. It seems to me that these sections of the document could be merged into draft-tempia after its adoption as a working group item (should it be adopted, which looking at the mailing list at this point seems likely).

> As for the "operational collection and use of metrics produced by such methodologies", this actually has been separated to another draft(the IPFIX extension draft).

Perhaps the architectural bits of this draft belong in the separate draft as well? (Again, such a draft would  in my personal opinion need a new charter to be adoptable in IPPM)

Cheers,

Brian (as an individual)

> 
> So, I think that daft-chen is in the charter and would hope that the WG would consider to adopt draft-chen as well. I remember that there were quite a lot supports on site during IETF 95th meeting.
> 
> Best regards,
> Mach
> 
>> -----Original Message-----
>> From: ippm [mailto:[e-mail]] On Behalf Of Brian Trammell
>> Sent: Friday, April 15, 2016 9:21 PM
>> To: IETF IPPM WG
>> Subject: [ippm] Call for adoption: milestone on coloring-based loss and delay
>> methodologies
>> 
>> Greetings, all,
>> 
>> We're slowly working our way through the queue of things to do after the
>> Buenos Aires meeting; we'll be making a few confirmations of calls for adoption
>> in the coming days. One of the items for the chairs to consider following our
>> face-to-face discussions is what, if anything, the working group wants to do
>> about "coloring"-based hybrid measurement.
>> 
>> On a review of our charter, we find that a milestone on measurement
>> methodologies, largely as described by draft-tempia-ippm-p3m, is covered by
>> Paragraph 4:
>> 
>>> Additional methods will be defined for the composition and calibration
>>> of IPPM-defined metrics, as well as active, passive and hybrid
>>> measurement methods for these metrics. In addition, the WG encourages
>>> work which describes the applicability of metrics and measurement
>>> methods, especially to improve understanding of the tradeoffs involved
>>> among active, passive, and hybrid methods.
>> 
>> We would therefore like to make a call for adoption of the following milestone:
>> 
>> (date TBD): Submit an Experimental draft on coloring-based hybrid
>> measurement methodologies for loss and delay to the IESG
>> 
>> and for the adoption of draft-tempia-ippm-p3m as the basis for a document
>> meeting this milestone. Please indicate support for or concerns about adoption
>> to the [e-mail] mailing list by Friday 29 April 2016.
>> 
>> Please note the IPR disclosure on draft-tempia-ippm-p3m
>> (https://datatracker.ietf.org/ipr/2557/) when considering this adoption call.
>> 
>> Larger frameworks for the operational collection and use of metrics produced
>> by such methodologies, however, such as described by
>> draft-chen-ippm-coloring-based-ipfpm-framework, do not appear to be covered
>> under the present charter. Therefore, we can make no call for adoption for an
>> appropriate milestone at this time.
>> 
>> Thanks, cheers,
>> 
>> Brian and Bill, as chairs