Re: [nmrg] call for adoption as RG document : draft-cx-green-ps-02

[Kiran Makhijani <[e-mail]>]

Hello Chairs,

I support the adoption of this work. It does comprehensive study of energy-efficiency in networks. Especially, it has covered protocol level part which I find as most relevant to the IETF and can help build requirements/solutions. Others are probably informational and are important for completeness. I hope as the document is adopted; authors would consider the following comments.


  1.  It will be nice to see some text on separation between the terms "green" and energy efficiency. Particularly usage like "greener' needs a sentence of explanation (relative to what). Some other terms that need reference or definition are: 'pollution metrics', energy-intensity, carbon intensity, etc. As an group, we can also discuss if something like ‘pollution metrics’ be used or can be come up with a different term with similar significance.
  2.  Related Question: If my entire network is only consuming green/renewable energy sources then are pollution metrics be 0?
  3.  The document is an idealized problem statement (until the conclusion). It assumes that the required ‘network functionality and operations’ with all the energy optimizations will not be affected. That may not always the case. For example, "in instrumentation equipment..." section, placement of NFs in DC-CPUs may not possible for low-latency cases, so it will be worthwhile mentioning trade-offs where energy-efficiency gains are not feasible.
  4.  Just a thought:- one more challenge/opportunity would be to evaluate existing protocols with some kind of low-power mode, with a goal to reduce the control-message distribution with some cost to performance (besides energy-aware path).
  5.  Section 3.1 Power is ultimately drawn from devices.  -- did you mean power drawn ‘by’ devices?
  6.  section 5.3 could also reference SCHC (8723). It might be worth highlighting the work done in constrained IoT, improvements made in those maybe applied to greener networks . However, it will be longer discussion but worthwhile pointing out that some of that work could be applicable and other may not meet large-scale or feature-intensive networks. Also, SCHC once the set of rules are configured reduces the header to one-byte, energy efficiency gains due to this could out-weigh initial rules setup and rule-lookup costs.
  7.  7. Intro: What are examples of different levels of granularity? Do you consider incorporating metrics to measure CPU, mem, transmission, port level granularity? Or per feature based, i.e., energy consumed by routing protocol vs SDN approach.
Cheers,
Kiran

From: nmrg <[e-mail]> on behalf of Jérôme François <[e-mail]>
Date: Friday, May 26, 2023 at 6:47 AM
To: [e-mail] <[e-mail]>
Cc: [e-mail] <[e-mail]>
Subject: [nmrg] call for adoption as RG document : draft-cx-green-ps-02

(Replace previous email with wrong subject, sorry... )



Dear NMRG,



This message starts the call for adoption as RG document for:



Challenges and Opportunities in Management for Green Networking - draft-cx-green-ps-02

https://datatracker.ietf.org/doc/html/draft-cx-green-ps

Please let us know if you support or not the work becoming a RG document.



In all cases, provide detailed comments to support your opinion and send them on the mailing list.



We are particularly interested in assessing the following aspects:



1.    NMRG being a *research* group it is important to articulate well the problem(s), challenge(s) and research questions.



2.    Is the scientific soundness of the draft satisfying? (quality of the scientific proposition, rigor of the method, impact) ; "the measure of success is the quality and impact of the research results" [RFC2014]



The procedure for RG document adoption and important criteria are detailed here: https://mailarchive.ietf.org/arch/msg/nmrg/CVEyLUvfxJk1Ud5WdM9Y5LGvQmU



The call for adoption is open for two weeks and ends on June 10th, 2023.



Best regards, NMRG chairs

Jérôme & Laurent