WEBVTT

00:00.000 --> 00:13.560
Yeah, okay. So when I was finishing high school, I was thinking I wanted to study something

00:13.560 --> 00:25.120
at university but I wasn't sure what. But it's not very yet. So a friend of mine said

00:25.120 --> 00:32.120
to me, since you can spend hours at your laptop, that time there were no letters. So

00:32.120 --> 00:39.800
you could spend hours in front of the computer. So why don't you study computer science?

00:39.800 --> 00:47.200
That was quite a lack of self-awareness for me, but she was right. And that was the beginning

00:47.200 --> 00:54.440
of my professional career. And now I still spend many hours at the computer, just everything

00:54.440 --> 01:05.440
is much more complicated with trying to straddle in with pods and nodes and clusters.

01:05.440 --> 01:12.640
So because my first encounter with the computers was with the home computers, that

01:12.640 --> 01:16.640
was much simpler.

01:16.640 --> 01:21.640
Okay.

01:21.640 --> 01:28.640
And while I go, my daughter, who is 12, and is she's in the audience, but I will not point

01:28.640 --> 01:34.640
at her because she will hit me. She will never come and get back again. She said to me,

01:34.640 --> 01:40.680
what is it that you're doing at the computer all the time? Whatever it is, I don't want

01:40.680 --> 01:47.680
that, I don't want to do that when I grow up. It was painful, but it kept me thinking,

01:47.680 --> 01:57.680
what is it that I'm doing? Because apart from our imaginary world with all these obstructions

01:57.680 --> 02:05.680
and objects and layers, there's a real world, which I see everyday reading the newspapers.

02:05.680 --> 02:14.680
There are wars, there are effects of the climate crisis, and there's human rights being violated

02:14.680 --> 02:24.680
in the name of tech. So I said, what am I doing to make this better? And actually, there's

02:24.680 --> 02:33.680
a lot I can do from my place. And there's a lot you can do as well, I realize. So that's

02:33.680 --> 02:43.680
why we're here, to convince you that you can do something better because a question for everybody.

02:43.680 --> 02:51.680
Have you ever found that you were running instances that shouldn't be running anymore that we're not

02:51.680 --> 03:00.680
needed? Okay, that's a lot of people are nodding, so I think. And so are I, and this is very common,

03:00.680 --> 03:11.680
and much more common than you think. From a study, then last year, by cast AI, it was found that

03:11.680 --> 03:19.680
it is clusters, there was only 13% of the CPU was being utilized, and only 20% of the memory.

03:19.680 --> 03:30.680
So this leaves a lot of room for improvement. This means, this can mean that a lot of energy consumed

03:30.680 --> 03:39.680
for nothing, a lot of emissions generated, also for nothing. Imagine a car driving, all the

03:39.680 --> 03:48.680
times taking nobody nowhere. That's what we have. So we can do something. We can start looking

03:48.680 --> 03:58.680
at what we have, and we can start measuring, and we can optimize. This is something where there's

03:58.680 --> 04:07.680
a lot of tools and open source projects that you can work with to improve. So, for example, for

04:07.680 --> 04:17.680
optimizing, at the end of the slides, if you want to have more references, there are links to projects, but you can apply

04:17.680 --> 04:26.680
best practices, auto scaling, right size, your applications, even run them at the right moment when

04:26.680 --> 04:36.680
there's renewable energy. And we are going to talk about two projects where we are involved with our open source,

04:36.680 --> 04:44.680
and you could also contribute to that, which specifically help to measure your emissions.

04:44.680 --> 04:54.680
And, yeah, don't forget to celebrate your improvements. I'm going to leave you, yes,

04:54.680 --> 05:04.680
starting with the real part. Yes, I'm going to talk about the open source tool guide.

05:04.680 --> 05:17.680
Kubernetes emissions inside too. This is the dashboard. Yeah, this is the main part of the dashboard.

05:17.680 --> 05:22.680
On the left side, we have the energy consumption of the Kubernetes cluster. So, it's

05:23.680 --> 05:30.680
how much you use this is for two days. And you can drill in for all the pots or all the notes. You can completely see what's

05:30.680 --> 05:41.680
the next part is about the carbon intensity of that period. So, in which location do you get is the energy generated in this case in Ireland,

05:41.680 --> 05:49.680
and for which period for two days is the second one is the embodied emissions. So, that is the

05:50.680 --> 05:56.680
emission generating during the creation of the server, and of the transportation of the server to the data center,

05:56.680 --> 06:04.680
and after that to recycle the hardware. So, that's that's part of the third part.

06:04.680 --> 06:13.680
Sorry, this is the embodied two laptops. And the last one that's all about this equation is the

06:13.680 --> 06:20.680
so-for common intensity equation. It's an isus standard since last year. So, what you do is you multiply the energy

06:20.680 --> 06:28.680
times the carbon intensity plus the embodied, and that is your Gramsier 2 equivalent of that period.

06:28.680 --> 06:38.680
And in this case, it's 25,000 Gramsier 2 equivalent for two days for this cluster. And it's around 50 notes that it's scaling.

06:38.680 --> 06:48.680
So, it's 20, 15. What does that mean? Do you know how much it is compared for an airplane flight, like I'm telling you.

06:53.680 --> 07:00.680
For a year, yeah, it's so loving. Now, for this translates two days of 50 notes,

07:00.680 --> 07:06.680
round is 200 to 3 kms airplane, average airplane flight. So, that's from Amsterdam to Brussels,

07:07.680 --> 07:15.680
to have a slight idea. How is Skype build-up? It's a Kubernetes cluster. So, it's around some

07:15.680 --> 07:19.680
period. Kubernetes, what you saw is a graphana dashboard. You saw a little part of it,

07:19.680 --> 07:26.680
but it gets its data from from media. It gets its energy from the Kepler exporter.

07:26.680 --> 07:33.680
It gets from the kernel with EDPF, the grid intensity. So, the electricity generated is from grid

07:33.680 --> 07:40.680
the grid intensity exporter. It's from the green sofa foundation. And we use the electricity maps API.

07:40.680 --> 07:45.680
There are other APIs. There are also the watt, time API in the ember, some can choose.

07:45.680 --> 07:52.680
It also depends on the region for the Netherlands. It's very accurate, but for other countries, it can be an estimation.

07:52.680 --> 07:58.680
The bottom one is by our Fista exporter. That's running locally as a database, all kinds of sheriffs.

07:58.680 --> 08:08.680
So, it knows all about AWS sheriffs or Headstone sheriffs. You can look at them up, but the embodied is.

08:08.680 --> 08:24.680
There are other opportunities. If you're running really Kubernetes, you might have to look at the green refuse to the aim is there to estimate and track the

08:24.680 --> 08:33.680
emissions of CNCF products. It's part of the environmental sustainability tech technical advisory group.

08:33.680 --> 08:45.680
And Flapia is on the picture. So, they are both contributing there. They are always looking for new contributors.

08:45.680 --> 08:54.680
Please, we welcome. Yeah, they currently have a CSD pipeline. They run CNCF products through the pipeline.

08:54.680 --> 09:00.680
And they measure the CNCI, the sofa carbon intensity rate of this product.

09:00.680 --> 09:12.680
So, if you want to contribute, look up the slack and CNCF. But now what can you do?

09:12.680 --> 09:21.680
Yeah, being here is very, very good. So, you're interested in the subject, probably maybe you're already doing something or contributing to some of the source project. So, that's, thank you.

09:21.680 --> 09:29.680
Yeah, make it part of your system. If you design a new product or a new sofa or you're running an up sofa, look at one move.

09:29.680 --> 09:39.680
Yeah, make it part of your system is sustainability, sofa, good sofa practice. Yes, it is. Make it part of you.

09:39.680 --> 09:44.680
Start a proofing by first measuring, improve, measure again, see what it gains.

09:44.680 --> 09:51.680
Becoming a visitor at your own community company. So, talk about it. So, people more be aware of it.

09:51.680 --> 10:00.680
And please consider it. There are a lot of open source. Yeah, in its tips. So, please join.

10:00.680 --> 10:12.680
Yeah, and maybe just maybe our kids will also be active after the agriculture game.

10:12.680 --> 10:18.680
And we're going to really sustainable cloud that we are helping.

10:18.680 --> 10:28.680
Thank you. We have some slides with some useful links, but it's online. So, you can look it up.

10:28.680 --> 10:55.680
One question. Yeah, the CNCF project seems very interesting. If I have a project which is not inside of the CNCF, can we collaborate to get it reviewed?

10:56.680 --> 11:05.680
Yeah, the question is, if you have a project outside the CNCF, can you also come to the CNCF and collaborate?

11:05.680 --> 11:20.680
So, for nowadays there's a tech environmental sustainability where people in the CNCF get together to promote sustainability in the communities or in the CNCF environment.

11:20.680 --> 11:30.680
If it's anything related to Kubernetes, yes, it's very welcome. I mean, green reviews are specific, it has a specific goal. If you have other projects you can come and talk.

11:30.680 --> 11:40.680
The midies are two every two weeks. Anyone can join. Yeah, please come. Thank you.

11:50.680 --> 11:52.680
Thank you.

