 Okay, so the first one we have is cluster net. This is a resubmission. And the last time we we had given them a message saying we want to see for the recording. This is our sandbox meeting from February 28. Period. Cool. Hi, everyone. Today is February 28th. This is a sandbox meeting. And we are going to start with our board. So the upcoming meeting today, we are going to talk about at least four of them. The first one is cluster net. And for cluster net. It's a resubmission and last time we told folks that we needed the more activity. Let's go check that. But just to remind everybody. What we're talking about here, maybe we should go through it quickly. Repo URL. That's the GitHub org. So managing your Kubernetes clusters as easily as visiting the internet. Okay. Okay, it supports multiple clouds, edge and stuff. Let's help you manage million thousands of millions of Kubernetes clusters as easily. Okay. Okay, so it is a multi cluster management. It has a scheduling framework too. And CLI. So this is the traditional architecture we have for multi tendency. There's a parent cluster and there are several child clusters. Has anybody been able to try this out? And so, so things I talked about taking a look at this, I have to go through this. Yeah, go ahead. Also this before. Yeah, I think it's a good project. It has, you know, the provider features and the information about the design is, you know, is good enough has enough information. Um, as far as I know, though, um, it's also last time's comment was, I think to say it's a contribution, not much contribution is that the comment. That's what we told them last time. Yeah, it looks like. I think it spoke with some well, you know, as well. Well documented, and also has quite some activity. Um, I looked at the number of contributors that were at least who made comments after the last year C meeting, I think that was April 22. It looks like there are new contributors that have been added, but the original author. I think the shoe. They definitely have the most number of comments. So that's like 400 comments and then others are like 20 30 and so on. But it still shows some progress. That is a good sign for sure. Yeah, it looks to me kind of fine for the sandbox. It's got, you know, I mean, it's not a huge number of people, but it's not, you know, there seem to be a number of a number of people contributing. Yeah, looks like they're on the right track. I think that they've been doing better. I don't think that they're ideal. Like when we ask for more participation, I don't think that we they've had the expectation that we had said at that time, but I think they're close enough. If we accept them, I would just add a caveat that their cluster net can contributing guide lacks the guidelines that they said that they would put in it. Um, it's actually says the following is a set of guidelines for contributing to cluster net. They're just guidelines, not rules. Use your best judgment and feel free to propose changes to this document and a pull request. And then it just has a link to the code of conduct. Do you want to add that there and then we'll start the work. Yep. I can do that now. So voting voting is going to be a little different this time. So what is going to happen is I'm going to come in here at the top of the hour. Um, that's 9 a.m. Pacific and kick off a vote from here. You get to come in and basically use GitHub and it will tell you everything in the particular issue about how to be able to vote. The reason that I'm doing it at 9 a.m. Pacific is because we have a week to vote. And the vote will automatically close at 9 a.m. Pacific on March 7th. So. Sounds good. So, yeah, let's leave. Keep your votes in your head. All of that. But we're not going to do a direct vote like in this particular meeting. You got a little bit to think about it. Sounds good. So I mean, you're going to send out the vote link, right? Yes. It'll, it'll be a vote that will be opened over in GitHub. I will collate all of them and send them to you. So, you know what's open. Okay, sounds good. Okay, I'm closing this one. Then the next one is inspector gadget. This is from our friends at Microsoft. I'm going to talk about the deployment and execution of ebpf programs, including many based on BCC tools. Didn't we do one for solo as well? You're talking about bumblebee. Yeah, bumblebee. At the time we made the decision that because they were only packaging ebpf programs into containers that wasn't anything special for cloud native. Yeah. Okay, so here that is addressed right like there is debugging troubleshooting inspecting various aspects. And then there is a continuity cryo. Have you seen this just in work with continuity cryo? I haven't seen any with Kubernetes. But I presume it's there. So we do have one question here that raised up. Do you remember this one Nikita. Yeah, I think we're still logged on the license review request. I think it's with the legal committee over on our end. Okay. So no need to take that into account. Okay. So going and looking at their website. This one I have played with and it's definitely more than just packaging a few programs. They're continuing to innovate with this ruling. And it's actually pretty good. You like it? Yeah. Good. Did you use it for containery stuff or Kubernetes or both? Kubernetes. Okay. Yeah, I think it provides very useful information debugging, right? The communities, I think it's a very important topic. I think this contributes to that. Let's go look how many people are there. I also know that they have a work with like head workshops and what are that integrated with like Qubes CDL plugins, like Qubes CDL trades. Basically, so that plugin allows you to schedule. Trades programs in your cluster. So they have integrations with other tools out there too. Okay. So like demon sets that go around putting things in. Yeah. So it kind of sits in the middle of like, so you have the ability to target lead to target a specific workload, as opposed to like a lot of things. A lot of the other tooling that's out there that will just target everything. And then you just dealt with a whole lot of data. This is more like point. This is more like a point solution. You're looking at a, you're trying to understand what's happening with a specific pod or a specific process. And this tooling lets you really focus on that problem. And it feels like they are, they are talking to the APS server to then you get that. Okay. Yeah, there is enough number of issues, enough number of pull requests. I definitely see this project growing in adoption, but I think most of the contributors seem to be from default slash Microsoft. Right. So maybe in the sandbox will help them get. Contributors from different companies. They seem to have like kicked off in July for a significant amount. And in, since then it's been what one, two, three, four, five, six, seven, eight people. I mean, Turin, Berlin, at least there is geographical Berlin to Berlin. Okay. Anything else you want me to look at folks. I'm satisfied here to put this to an old word. I think we can add that, you know, they, if they can extend their contributor, the more diverse paid like from other companies, that would be good. Yeah. You know, that is on the mandate for sandbox itself. So, but, you know, never hurts to remind people. And that's what they are here for. So, plus one. They also mentioned it in the issue body itself that they would like to play for sandbox so that they can draw their contributors. Okay, four contributors, they are calling it out. No mind. Okay. So, can I go to the next one. Any other things you want to look here twice. Okay, let's go to the next one. The next one is. I did have a chat with one person here. Yeah, so contributing 99% of the code. Coda black responded to that. And then Emily, you responded them to go talk to the tag. I don't see a notation whether they've actually gone and talk to the tag yet. I haven't either. Yeah. Similar to help file. So I was looking at this, I haven't tried it out, but I was looking at the project and like it seems to do two things. So one is sort of custom combined customize and help in a way. So, for example, it can add templating on top of customization manifest and so on. So in my opinion, I think if you want to use helmet customized together today that is already supported in projects like flux already so you can already add substitutions for customized patches for health charts today. And like it's called out there and file is also another existing tool that you can use to manage health charts in one go. The second thing that it does claims to do is like mighty environment support so you can deploy to different environments like that fraud and so on. So this can also be done today using customized overlays and it's already supported in flux in our go CD. But I think the main technical method that this tool has is that it's simple enough. So it can do configuration management like without any extra cluster side dependency so no controllers or CRD is and so on. But I personally not convinced that this is enough to warrant being in the sandbox. The one thing that could be a very big blocker is that it seems to be a one person project. I think the author is also called it out in the shoe. So, I don't really see this as a good fit for sandbox as of now. Right. I would agree I'd like to see them go have a discussion with app delivery to get more feedback from them as well as potentially expose them to more contributors. I think I agree with what the two of you said. Yeah, and I think the more contributor feedback was what we had given to the first project we'd like to say was a cluster net. Yep. And I think it's that same guidance they went out to go tackle that needs to be tackled here. Okay. Sounds good Matt. So Matt, can you cut and paste that or in writing your own words here. And we can move right and come back to this issue and make a comment on it. Do we want to vote or do we want to just say that explicitly. Like, I don't want them to feel that they're hitting a minus one bunch of minus ones. Amy, what do you think I don't know. I think that seems accurate. I think this is one of the ones where we get to be able to say no vote at this time reapply later. Yeah, I'll Matt I'll work with you on how to be able to like frame that in the issue here. We would end up applying the tag assigned label and put them in postponed. So the issue would remain open. We can do that. Yeah. Yeah. I can provide that as a comment so we don't lose track of this discussion. Yeah, I will totally take that. Thank you. Thank you. Anyone else wanted to chime in here. Yeah, I think this one person, you know, contributor is. Yeah, we see other one right so I have a party I think we will let them to have more contributors right or more diversity, but this one is just one person. Okay, let's go to the next one then. The next one is loxy LB. This is one of our projects that comes from the East Asia region. And most of the people I saw where from there. I mean, EBPF is hot. So, they're going with that and it's a load balancer. And it looks like they are in the edge space a lot. Yeah, it's a load balancer that that all balance traffic to the age sites. The age clusters to the edge clusters not from, or maybe both I guess. It's a service, you know, low balance to the age. So it is a Cuban. Okay, so it is bound to cuban it is scenarios for each on-prem cloud provider so I guess that's a good thing. Okay. Going to their repository, South Korea. So EBPF, did anyone get a chance to look at this in depth. So I did find one, when I last look, there was one repo in the Github org which was missing so I asked them to add it full compliance for load balancers. At the time to like look at this deeply but from a quick glance, a couple of things stood out was there was some Github discussions that were, the maintenance was really proactive in answering feedback and I saw a blog post from someone who used this project and they had some feedback on bug reports and feature requests and the maintainers actually responded to them and also implemented this feature request and that was nice to see. One other thing that stood out to me was they, it doesn't really have to be a pocket but they have a custom fork of the IP root 2 package because they need some patches to properly load and unload EBPF modules. I am personally not a huge fan of having the folks, I think it would be good to see them either upstream these patches or work with the community to find a solution. And again, most of the contributors seem to be from a single company that created this project. But again, maybe being in the sandbox sense of community structure. So that's the netlocks.io. So locks is common here locks LB and netlocks LB contributors. This looks bad to people. And that looks like a water count that they use. I've actually seen the human responses come from that account so I've been very confused if that's a water content. And maybe it's both creates an issue as a less PR. If it's a bot that's helping the project, I think that's fine. But given that there is what appears to be some human level of interaction coming from it, I'd prefer that it have a better identity associated with it if it is a human. They have only four packs, four releases. Yeah, I mean, it's still relatively new. Yeah. Oh, I might have been an older cat by so something is essentially called a cat. The first external was 080, which has all the things. And then after that it's been like, not too many changes. Yeah, it's like four, three other tags between them have five PRs. Can we ask them to present a tag network or to also increase the contributor base and to get. Yeah, let's use the same thing that we used for. Clue cuttle go to the tag and do it as pending and come back when you have a few more external folks, not just from the netlocks seems acceptable to everyone. Who's going to comment on the issue to let them know. Nikita. Nikita and Amy, right? Happy to work with to be able to make sure that all of the pieces come together. Okay, so all right. Looks like we will be opening two votes both cluster net and expect your gadget. Correct. Yeah. Perfect. All right. Anything else for this session of sandbox review. I think we are out of time for this session because we have to use the other half for some other discussions that we need to have. So, please go ahead and thank you. Thank you all very much. Thank you all very much.