Re: [ippm] Adoption call for draft-olden-ippm-qoo-02

[Bjørn Ivar Teigen <[e-mail]>]

Hi Greg,

Thanks for the feedback. Much appreciated.

On Tue, 6 Feb 2024 at 14:50, Greg Mirsky <[e-mail]> wrote:

> Dear All,
> I've read the draft and found it well-written and proposing a
> pragmatic approach to a real operational issue. I support the adoption of
> this document by the IPPM WG. Below, please find my notes and questions for
> future discussions:
>
>    - In your opinion, what is the relationship among this draft,
>    draft-teigen-ippm-app-quality-metric-reqs
>    <https://datatracker.ietf.org/doc/draft-teigen-ippm-app-quality-metric-reqs/>, and
>    TR-452.1? Which of these documents lists the "full set of requirements"?
>
> In my opinion, app-quality-metric-reqs is a higher-level analysis of the
needs of different stake-holders and an overview of which needs different
metrics can deliver solutions to. TR-452.1 defines the quality attenuation
metric and how to measure it.

>
>    - It seems like the perspective of 'we' is not usually used in IETF
>    documents. It would be helpful to re-word passages like "we believe", "we
>    can", and "we propose".
>    - In your opinion, what is the relationship between "active testing
>    from applications" and "monitoring from network equipment"? Do you expect
>    that both are using the same measurement method or only produce the same
>    performance metrics?
>
> Our idea here is that the latency distributions can be measured both in an
end-to-end way from within an application, and/or in a per-segment or
per-sub-path way using either passive observations of active probing from
CPEs, switches, routers, and other network equipment. The compositionality
of the quality attenuation metric makes this possible in a very neat way
--- if latency is measured at the HTTP layer, for instance, that simply
means the measurements include HTTP processing overhead (and potentially
more than one round-trip), whereas if latency is measured on the IP layer
the results do not include the extra processing that HTTP requires. The
general way of comparing latency distributions to network requirements will
work in either case (though, of course, the requirements must be formulated
and validated for the specific measurement setup that is used).

>
>    - I think it would be helpful to add references to the measurement
>    methods listed in Section 2, e.g., STAMP, IRTT, etc.
>    - Also, a reference (or more than one) would be helpful in the
>    following:
>
>    Using Latency distributions to measure network quality is nothing new
>
>    and has been proposed by various researchers/practitioners.
>
>
>    - Do you mention a passive measurement method as defined in RFC 7799
>    in the following
>
>    Latency Distributions can be gathered via both passive monitoring and
>    active testing.
>
> What is an example of the passive measurement method that supports the
> measurement of latency distribution?
>
>
>    - It is stated in the draft that
>
> The active testing can use any type of IP traffic.
>
> It would be helpful if this statement is expanded and supported by some
> examples, e.g., TCP-based measurements, raw IP, etc.
>
>
>    - It seems that an additional information on whether the list of the
>    measurement parameters in Section 3 is sufficient to "ensure network
>    measurements can be analyzed for precision and confidence".
>    - The document introduces notions of "perfection" and "usefulness".
>    What are your thoughts about the work on the Precision Availability
>    Metrics <https://datatracker.ietf.org/doc/draft-ietf-ippm-pam/>, and
>    its metrics that express a service availability/unavailability. I am
>    interested further discussing if the terms from PAM can be used in QoO.
>
>
I've opened an issue to address the rest of your comments, here:
https://github.com/domoslabs/QoOID/issues/12

Cheers,
Bjørn

Regards,
> Greg
>
>    -
>
>
> On Tue, Jan 16, 2024 at 9:13 AM Tommy Pauly <tpauly=
> [e-mail]> wrote:
>
>> Hello IPPM,
>>
>> This email starts a working group adoption call for "Quality of Outcome”
>> (draft-olden-ippm-qoo).
>>
>> https://datatracker.ietf.org/doc/draft-olden-ippm-qoo/
>> https://www.ietf.org/archive/id/draft-olden-ippm-qoo-02.html
>>
>> The call will last for 3 weeks, and end on *Tuesday, February 6*. Please
>> reply to this email with your review comments and indicate if you support
>> adopting this work.
>>
>> Please note that we did a previous adoption call that did not receive
>> sufficient feedback. At the last meeting at IETF 118, we did have a good
>> amount of comments and questions, so please do reply to this email if you
>> have reviewed the document.
>>
>> Thanks,
>> Tommy & Marcus
>> _______________________________________________
>> ippm mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/ippm
>>
> _______________________________________________
> ippm mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/ippm
>


-- 
Bjørn Ivar Teigen, Ph.D.
Head of Research
[phone] | [e-mail] | www.domos.ai
[image: https://www.understandinglatency.com/]
<https://www.understandinglatency.com/>