[{"author": "Shukri Abdallah", "text": "

are slides uploaded?

", "time": "2023-11-07T10:51:55Z"}, {"author": "Cindy Morgan", "text": "

Meeting materials for this session: https://datatracker.ietf.org/meeting/118/session/eimpact

", "time": "2023-11-07T10:57:32Z"}, {"author": "Jari Arkko", "text": "

All slides we have received are uploaded (6 decks)

", "time": "2023-11-07T11:01:03Z"}, {"author": "Tony Li", "text": "

Not showing up in meetecho...

", "time": "2023-11-07T11:01:33Z"}, {"author": "George Michaelson", "text": "

Maybe its me, but I still believe for a large class of links the energy budget is length, loss and BER informed more than data rate at a higher level

", "time": "2023-11-07T11:06:33Z"}, {"author": "George Michaelson", "text": "

so it feels like attributing to flows, thats fine: but its accounting, not actual

", "time": "2023-11-07T11:06:46Z"}, {"author": "George Michaelson", "text": "

I stress I do NOT mean to imply no work to do here

", "time": "2023-11-07T11:07:01Z"}, {"author": "George Michaelson", "text": "

ah. the slides have kind of gone there

", "time": "2023-11-07T11:07:31Z"}, {"author": "Jim Reid", "text": "

Are there metrics, processes and methodologies from other sectors - say manufacturing, transport, agriculture, construction, etc - which the IETF could plagiarise?

", "time": "2023-11-07T11:16:54Z"}, {"author": "George Michaelson", "text": "

you would think they'd be in information processing more than physical processes

", "time": "2023-11-07T11:17:57Z"}, {"author": "Jim Reid", "text": "

I was thinking these sectors might be further down the road on how they develop and use metrics and how to act on them.

", "time": "2023-11-07T11:19:40Z"}, {"author": "Warren Kumari", "text": "

I suspect my comment was unclear; what I was trying to say was: \"If you can demonstrate a finanncial incentive for people to collect/deploy this, you will have a much easier time getting people to do so...\"

", "time": "2023-11-07T11:24:03Z"}, {"author": "George Michaelson", "text": "

Actually, ITU, ETSI &c feel like good forum to do this work in. I think this might be one of those times when liaison to SDO is the key thing for us.

", "time": "2023-11-07T11:24:15Z"}, {"author": "George Michaelson", "text": "

if we kept this work as \"keep us informed\" we might be light for drafts, mainly adopting other work

", "time": "2023-11-07T11:24:36Z"}, {"author": "Warren Kumari", "text": "

The metrics will also be more accurate/reliable than \"You will do this because complaince...\"

", "time": "2023-11-07T11:25:09Z"}, {"author": "George Michaelson", "text": "

I even think it might address Jim's comments about prior art

", "time": "2023-11-07T11:25:19Z"}, {"author": "Jim Reid", "text": "

+1 George. If $SDO has already figured this out, the IETF could just adopt that and decide how to apply it to our YANG models, protocol bloat or whatever.

", "time": "2023-11-07T11:29:19Z"}, {"author": "Alexandr Railean", "text": "

Internet infrastructure energy consumption is a small fraction of what our civilization consumes in general. At best, if we optimize it - the small fraction will get even smaller, would that make a difference?

\n

Shouldn't we focus on matters that would yield a better return of investment?

", "time": "2023-11-07T11:30:11Z"}, {"author": "Jim Reid", "text": "

IMO it's not a binary choice Alexandr. Focusing on the biggest areas of concern shouldn't mean ignoring the others.

", "time": "2023-11-07T11:33:38Z"}, {"author": "Arnaud Taddei", "text": "

What is the mailing list for this work?

", "time": "2023-11-07T11:33:56Z"}, {"author": "Tony Li", "text": "

e-impact@ietf.org

", "time": "2023-11-07T11:34:27Z"}, {"author": "Arnaud Taddei", "text": "

Tx

", "time": "2023-11-07T11:34:33Z"}, {"author": "Tony Li", "text": "

@Alexandr Railean Networking consumes non-trivial amounts of power.

", "time": "2023-11-07T11:35:13Z"}, {"author": "Alexandr Railean", "text": "

Are there some accurate statistics about it? What source do you recommend?

", "time": "2023-11-07T11:35:54Z"}, {"author": "Tony Li", "text": "

https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks

", "time": "2023-11-07T11:38:07Z"}, {"author": "Tony Li", "text": "

https://www.techtarget.com/searchdatacenter/tip/How-much-energy-do-data-centers-consume

", "time": "2023-11-07T11:38:52Z"}, {"author": "Tony Li", "text": "

There are lots of such articles and they point to DC consumption being around 1%. Networking is only a small chunk of that, to be sure.

", "time": "2023-11-07T11:39:48Z"}, {"author": "Jari Arkko", "text": "

Schedule: We will run a bit over -- there's a few minutes buffer between this and the next session. We try to finish 12.50-12.55. For the current and upcoming lightning talks, comments are welcome but please make them on the list.

", "time": "2023-11-07T11:40:36Z"}, {"author": "Jim Reid", "text": "

Then there's the networking kit outside DCs: corporate server farms, CPE, etc.

", "time": "2023-11-07T11:41:49Z"}, {"author": "Tony Li", "text": "

Under the rubric of thinking globally and acting locally, it's our responsibility to act in our area. Certainly no one else will be able to.

", "time": "2023-11-07T11:42:51Z"}, {"author": "Jari Arkko", "text": "

A comment on the API topic: there's also a lot of interesting questions beyond the API formation itself. Can we trust the information, incentives for providing correct information, whether there are dynamic situations that make answers provided at one instant in time different from what the situation is a bit later, etc. More work needed, but very interesting topic!

", "time": "2023-11-07T11:44:13Z"}, {"author": "Tony Li", "text": "

Obligatory plug: https://datatracker.ietf.org/doc/html/draft-li-ivy-power-01
\nThis is work towards a YANG model for power consumption and management.

", "time": "2023-11-07T11:45:50Z"}, {"author": "Jari Arkko", "text": "

Yes, Tony, seemed very useful!

", "time": "2023-11-07T11:46:40Z"}, {"author": "Markus Stenberg", "text": "

While the small part of 1% number may sound demoralizing, ultimately entities want to act 'green' (or at least appear to do so), and measuring how 'green' their operations is in and of itself valuable to them; their scope is only their operations and not random coal power plants in country X.

\n

As cloudy SaaS datapoint, we've had customer requests for information of footprint of what they're doing in the cloud in terms of emissions. Due to that some standards somewhere would be definitely nice on how to produce and report and analyze the data.

", "time": "2023-11-07T11:48:11Z"}, {"author": "Jari Arkko", "text": "

Comment on Jan's talk: I think we should design products so that idle power is minimized :-) I think it is possible in many cases.

", "time": "2023-11-07T11:50:40Z"}, {"author": "Suresh Krishnan", "text": "

@meetecho can we get a meeting materials refresh

", "time": "2023-11-07T11:50:50Z"}, {"author": "Dhruv Dhody", "text": "

https://datatracker.ietf.org/meeting/118/materials/slides-118-eimpact-lightening-talk-beyond-carbon

", "time": "2023-11-07T11:52:35Z"}]