 Hello Drew. Good morning. How are things going? So far so good. Yeah, nice. Let's wait a couple of minutes to see if... Yeah, let's all Lucina join and then drop. Maybe she's having issues. Yeah, with the audio. Oh, now she's back. Hey, Lucina. Good morning. Morning. Let's go a couple of minutes. I don't know if someone else is also going to join. Last week, Oliver and Taylor, we were thinking about like having a working session and trying to complete the resume principle. One concern per container principle. So I don't know if we are going to continue on that or we'll have a regular meeting. Hey, good morning. Hello, Taylor. Good morning. Should we start the meeting? Yeah, Victor. You're late in today? Yeah. I guess let's get started. If we have a short one, that's fine or we can turn it into a working session, whatever we want to do today. Makes sense. So, well, I don't know if I should just go through the through all the things that we have in our agenda, like, I suppose, aware of the incoming events. I don't think that we haven't added anyone else, right? Yeah, probably the same CFPs have closed for Tokyo Day. So they'll be going through those and we should hear an announcement on the, I guess, people get notified and then announcement on the agenda for Tokyo Day. And we have several KCDs, at least in Texas and UK. Yeah. The other one that I haven't heard anything is about the NAFRIO, TSE and Developer Summit. They were proposing two different weeks, but one of the things that I noticed is one of these weeks are we're conflicting with the open source summit, especially the second one. So I don't know if at the end of the day, they're going to choose only the first option. It's the only one who is not making conflict, like the September level. It's not a little conflict with the September 19 week. Again, they haven't confirmed that. So let's see what happened. Other than that, I think that we are okay in terms of events. So regarding PRs, I guess we are still empty and we have the same number of issues. I think that it's better to focus on the one single-purpose container issue. So what do you think? That is a great idea. Let's take a look at the one concern per container, I guess. Unless you want to... Well, I don't think there's enough of us to look at what's the next one. Let's try to knock this one out and then maybe we can look at what the next one would be. And I know you and I had talked about stuff related to NFIO, so it's probably something to dig into next. Okay. So that one is in the issue section or in the discussion? I don't remember if we put anything in an issue yet. I think we were going to look through and try to highlight. So we'd need to... I mean, we could have like a brainstorming or something where we do an initial highlight of the NFIO stuff and then try to dig into individual ones or we could do it more in depth. Hi, Aldeca. All right. So in that case, should I just go through this one or like go to discussions or what he's suggesting? I think we should defer that rather than... Let's take a look at that one process and then if there's nothing new, then we can either work on it. I think we should work on it. That's what I was going to say. And then later we can look at the NFIO stuff. Okay. I'll make sense. Aldeca, do you have any topic agenda or anything to add for today? No. I just came to learn. All right. Nice. We need more exercise on here. So, well, what we have been trying to do is... Yeah, this is the right graph, right? So we have tried to make this particular graph completed. Oliver Taylor and I, we have been trying to... Yeah, being out here with the definition and all these things. I think there are several sections who are ready to go. Others, definitely, we need more feedback. So I will give plenty of time to look. Actually, let me share the link. And if you can also participate in adding more stuff here, it will be nice. The first thing is like the summary. The summary, we can see that it's ready. So do you think that we need to revisit the summary or all of us are agreed with this particular summary? I'm good with that summary. Aldeca, do you have any feedback for this? So this best practice about... We're not quite sure on the name. It's kind of shifted several times. It was one process per container and one process per category. I think an existing test right now that's in the test suite is checking for its multiple process types in a single container is what it's trying to look for, as a not a good practice in general. So the idea of this related microservices and trying to separate concerns and then helping to break down services into their component parts. For size, so you're not going to keep expanding. You're not trying to take too many things. You know, all the different reasons, benefits about microservices is one of the main drivers. It makes sense to me. Okay. Yeah, this is our paragraph is basically taken from Docker as a best practice. So yeah, this is something that we are referring. The next section is motivation. When we were reviewing motivation and goals, we were a little bit confused. That's why we tried to put it like a small reminder what we consider like a motivation. And in this particular case, we split it into main audience. The first one is the consumer or the operator, person who are like using those CNFs in the infrastructure. So that's why we have a few points here and probably we're going to add others here as well. For us, motivation is especially is experienced from the past. So, so the or things that they want to avoid it. For that particular case, in this case, for example, the operators have been faced with the security issues or like issues during the troubleshooting process. And they are trying to avoid those particular problems. So I'm using this particular best practice, how we're using this particular best practice can avoid to go through these particular issues. For example, scaling. When we talk about the way that they have to scale, it's pretty hard to scale resources independently. So that's why we try to the motivation here is like how they can do it in a good cost manner or way. I also give a few minutes to take a look and maybe think about it. What do you think about in the area of one single container, one process per container? I think that those from the operator perspective are the only things necessary or like the thing that is the only motivation that they can face it. Are you good with accepting those as a start for this? I think that's part of what we need to do is just either close or move stuff out of the way. I mean, the top part is obviously this is just the definition description, but we shouldn't have that in later. But the rest of it? Well, if we don't add something here for developers, maybe we should also remove it. I think we were talking about moving the benefits down, but that's to the goals section. Go ahead and accept the top part and then we'll adjust this. The other just comment. All right, so we'll move it later. Maybe for the developers, I guess one of the motivations could be that they don't have to deal with other things. I don't know if they have to, I'm trying to put it like a what could be the benefits for seeing the developers to put everything in a single, well, just focusing one process per category. I'm not sure the health benefits because if you split the services in two containers, you have to start thinking about how to communicate between these two things. Let's say that you have one process, like main cache. We were like using that as a primer example. So when you have main cache and the web server in the same container, you can rely on things on main cache, but when you use this particular best practice, so you have to start thinking about distributed the communication, like sending things in the cache in another container, maybe requires some extra configuration. So that's why I would not say that could be a motivation. I need to think about one benefit. Sorry, can you all hear me now? Yeah, I don't know what happened. So the different groups having different speeds, so you could end up having like a one component that's the development speed and updates are much faster and another part is much slower. You mean doing the development, right? Yeah, that's good one. So it's one, basically it's one of the major benefits to you. I have a microservice architecture. We lost Taylor again. No, I'm here. I'm just trying to think of the development for what I was saying. I mean, it's the different speeds. I'll just say components maintained by multiple groups and our developers. Probably I would just allow or like add allowed work here. Yeah, multiple groups like maintain different components like just to keep the same the same, the similar format that we had before. Yeah. Yeah. So if I'm thinking like bugs and everything, so just the service offering, if you have a component that's minimized to the single concern, then you do have, you got to make sure that the surface wherever the communication is going to happen is solid, but internal to it, you've simplified issues. So bugs, if there's anything, is it? And this goes back to the first one. If, you know, if there's multiple components and different groups and stuff, and someone introduces a bug in one area, but you're tightly coupled, then you end up with potentially having bugs in other areas that as a result. So if you're loosely coupled on those components, then someone can do their development if they introduce a bug. And ideally, it's not going to introduce a bug for the others. If you're, if you make sure that the handshake between them is strong, and you're going to make sure that works. And even if you break something else, you didn't break everything. And then that's similar for stuff like if someone introduces a new feature, and they're not as worried about breaking everything because they can keep it internal. Yeah. What I was thinking is like, is that one not the area of operator? Like, so yeah, if you talk about like, when a developer has to fix a bug, yeah, probably it's going to be easier. Like, because they have, we're talking about modifying the source code. Well, it's, you're going to have both. So you're going to have the the ops team that may find a problem themselves. And then how can they narrow it down? And if they're broken into parts, then hopefully the problem is more visible to being in a single component, and then they can report that or whatever. And then you're also going to just have bugs that developers find themselves, maybe, you know, during more extensive testing or before it ever gets out of staging and moves onto the customer or something. So we're going to have overlap on all those. I don't think we should try to keep it all separate. If they're doing the same thing, I think that's okay. Okay. Because the other thing is it could be like, in terms like a unit testing or functional testing, I guess it would be more effective if you only have to do the functional test or like a unit test when you're container only focusing one single thing. It's interesting. So motivation. So we're kind of what is the benefit versus the motivation? So a motivation, I'm just going to say increase test coverage for each component or maybe it's confidence, motivation, and test coverage of the CNF. So then that's a, that's not the, what are we doing? This is, they want it. So the developers would like to be confident that they covered everything. The, like, so the goal, so this would be, hold on, getting it mixed up again. The goal is something you wish. So I might, I might be getting it right. The goal is to give them confidence, motivation. Exactly. Because in the past, they didn't have that confidence. Yeah, I think that's fine. Right. So the goal is confidence and test coverage. Motivation is, yeah, so the motivation. Test coverage over a complex CNF is difficult to see because of tightly, test coverage of a complex CNF. That is tightly complex and tightly coupled CNF is difficult to achieve. There we go. So now, so this is the motivation. That's their motivation. So then our goal, I'm going to just put it here. We don't really have it split. But so now the goal is confidence and test coverage of CNF, reducing the complexity. Something like that. Does that make sense? So that's our goal with this best practice. And the goal is related to this challenge. Should I just do like an increase here? Yeah. That's our target, right? Yeah. Multiple, multiple. So the motivation here. So this is multiple maintainers, which can be groups and organization can be different groups and organization. And I'm just going to write it out. Then maybe we just can block each other. Each other's development process when a CNF is tightly coupled. Does that make sense? I was thinking like a story when the goal should be accelerating the development process. So we're not close. This is motivation. Okay. So that, I mean, maybe we should actually flip this. What are the goals or what we're going to try to do? The motivation is why we're trying to do that. In a lot of ways, this could be merged and we might, should say, motivation and goal. And we just merge to these two sections. I think it causes confusion what we currently have. Yeah. That's how it's split up between them. But yeah, hold on one second. I got something to do here. Here we go. I'm going to do this in here and then show you what I was doing. I was adding the definition to the, let's do it like this. All right. This is in the proposal document. Let's see if I can just show me the actual file. Review and code space. Is that what I want? Oh my gosh. What is this weird thing? Okay. I'm not going to do it that way. Anyways, we can go look at that pull request in a minute. I'm going to pull this up. So this is the, this is telling us what each one of these sections are. But it doesn't actually give the definition, which is why I added in that pull request. But you can see the example here. Motivation. You put something in that chat? No, I'm, I just realized I'm not sharing my screen. Sorry. I can give you, I mean, I can stop sharing. Yeah, let me, let me share. Sorry. All right. So I created a pull request to add to this. This is more of how do you propose a best practice? This document, the first one, and going through all the areas. So on motivation, this is why would we, the example is why would you do a best practice proposal? So, and they're putting them in the motivation. It's saying why you would do that. Without standard mechanism for describing important cloud and best practices, wider telecom communities will struggle to understand the importance. So this is the motivation behind. So then the goal is provide a standard way for the community to propose and discuss cloud native definitions. That's what there's this in this document, like why it's here. So why are we doing this at all? So they're giving, here's the problem. The motivation is the problem that we're trying to solve. And then what are we trying to solve with what we're doing? That's, that's what we're meeting between goal and motivation. And then on goal is like just out of the scope of things. Yeah, and that could be out of scope. And this is trying to be more formal with, is the main reason to split these up into totally different sections. So motivation and then goals versus having it intermixed where you would say, let me give you one and then let me show what we're trying to do. If you were reading like a white paper or, you know, some more technical blog or something, then it may start by putting forward, here's the issue that we're trying to deal with and more extensively and then start breaking it down. This didn't explicitly point back. It's just expecting that you actually read it. So that's why we're following that whenever we're going here. So we have the motivation. I mean, I put motivation and goal, but motivation here. So this is the problem. And then the goal is how we're trying to solve it. So we really could kind of call this challenge and then, you know, solution and then non out of scope or objective or whatever. Yeah. I don't want to change all that right now. No, that's fine. Let's keep it as it is. May putting this into the actual PR. I mean, I probably no one looks at them anymore, but the this definition, but the pull requests that I was putting, I was trying to do that. We'll see how it is. Like the templates, probably a better place whenever, however, they're copying it over. We just want to delete it out. But it would say something like motivation is the challenge. And it's that the end user has some motivations. The challenge is something that end user has from the past. And it's something you want to help overcome with your goal. Right. So then that goal is something in the future we're trying to get to the objective. I think we need a little work on the how we introduce it. And maybe maybe this would be a good thing. Lucina, you've put forward in the past and we've done it. And it sounds like it's time to do it again as introduction to the working group and how people can get involved. So the fact that we're struggling to talk about this means that other people are going to have a harder time. So I think we could go in and update this at a minimum, have it like a high level and saying what what are these best practices and go, okay, the telecom community has challenges. And we're listing those as the motivation for what the best practices are trying to solve, the objective. Each best practice has an objective, its goal and what it's trying to solve related to those challenges and motivations. And then we try to list any out of scope items and the non goals. So I think we could write that up, Victor, and probably have it like as the updated intro, there's the contributor document or something that we could update and maybe link from the read me how to get involved, something, and then maybe have a work on something for like an introduction, like a, I don't know, 15 minute presentation or something for a future conference or whatever. Good for now, when we move on. Yeah, I think that, yeah, let's move on. Does it make more sense? Yildica, does it make sense to you? You haven't been around for a while trying to make sure that this is clear. How it's organized, what we're doing. I think it mostly does. My brain also needs a little time to process, but as we talked it through here, it did make sense or does make sense. Any questions? Right now, any questions? I don't have for now. All right. So Victor, thinking about it from motivation as challenges and goals are the objectives to help solve those challenges. I think we could reword some of these. I've started. So this one is the tightly coupled. This one is actually the goals. So we should move that down. We want to reduce the complexity of individual components. And yeah, and that covers basically things from from the developers and operator. Yes. So this leads to I'm going to put it right below just for a moment so that we can kind of see the relationship and then I'll move it down. So this is a goal for this one. So we were talking about test coverage. So this is another goal. So this is a goal. We'll move it down. So multiple maintainers can block each other's development. I feel like that's a motivation, right? That's a problem. Yeah, this is a motivation. And then this is a goal. Allow multiple groups to maintain the independent development life cycle and their own development maintain. Sorry, their own independent development life cycles. So motivation and then the goal. A high degree of test coverage with complex difficulty and then go with the complexity of individual components, which leads to and this one feels like multiple. So oh, we have the increased confidence. I think you put it increased confidence test coverage. So I kind of feel like that's actually part of this reduced complexity. I'm going to temporarily move this just so I can read it. We're going to also start like putting like non goals here the same way. Yeah. Probably this isn't a non goal is to specify how the house and and where is what? Well, what I was saying is like maybe like like like specifying the details like implementation details. I don't think that we have to specify how to to to split the things or like the logic that they have to follow. I think in the second bullet, if you add fine and fix box individual components, you are covering basically two audience like the developers and the operators. Well, this is really a goal. CNF challenges with complex and tightly capital CNF. So I need to do it like that so that we can see what are we trying to solve? All right, large surfacer. The other thing would be like the first bullet is like scaling issues. I want to say locate and close in there's something difficult to it's not just fine, but like reduce the area difficult to find where the bug is. So you may go, oh, the bug is in the CNF. That's different from going. Oh, it's in this one CNF that has like 15 different services that it's providing. It might be a lot harder to find like which part of this is actually causing the problem. What about the work identifying? Yeah, but I'm talking about like identifying. It's fine, but it's like reducing you're trying to reduce the area. If it's a complex CNF, but you have it broken into microservices and it's more likely that you're going to see the problem on one component. That component may actually not be the problem, but they can at least go and immediately investigate one component, then go, oh, this is caused because another component is given it the wrong information. So you're reducing the area when you report it instead of going, you're CNF's broken. You go, this component within the CNF is responding incorrectly and then they can go and investigate that. My brain is thinking of more idioms, difficult find bugs and home in on the area that the real problem is, but that's not how I'm trying to use words that aren't idioms. Anyone have anything? Luciana, somebody? Yeah, because my brain only brings idioms like identify, discover, things like that. Yeah, difficult to detect, identify. I'm going to use this potential component where bug resides, something like that. Okay. So these are the motivations and then this area is actually goals, reduce the attack surface area, simplify troubleshooting real, so we don't really have that. So something about it's instead of making something visible, you're obscuring it, it's opaque. So when they're tightly coupled, then you don't really know what's going on. But if they're broken into microservices, then you're going to have the communication as well defined between them. They could potentially have log output where it's, you could have like CNF name and component name. So then if there's a problem, then that'll help find troubleshoot. You're talking about the logs output like that? Yeah, it doesn't mean that they're not going to have it. They could have an internal, they could have it all tightly coupled, but then log stuff separately. But if you are already intentionally separating them into microservices and that's visible, then I think you're more likely going to have your logging and stuff separated because you're already having to run them independently. So on the development side, you're going to do that. I'm going to move these goals down into the goal section. I think that was here. I'm going to do this. All right. This was before. I think this general goal is true. I'm going to make this a different color so we know to remove it later. All right. If they're broken into microservices, then you're going to be taking advantage of the Kubernetes orchestration more. I think you brought this up, maybe, the architectural practices. So if you're microservice architectural practices, if you're breaking it down, then you can start taking advantage of more of those. So I think that's kind of a developer view as well. But it does, if there's ops team know that you're following that, then they can probably train and build their environment. That's not complete, right? Like simple file troubleshooting? No, it's not. I'm going to do it like that. Reducing text service, simplified troubleshooting readout. I think that's that one. So we can remove that. Upgrading process types independently. So we didn't talk about that as a motivation, but that's a motivation really for, I mean, developers may or may not. I mean, I think they do when we talk about like this one tightly coupled right here. Then they care about it, but we can also put it up here. So the motivation, a challenge for the CSPs, they on upgrades. So upgrading, so I upgrades with something about multiple services, multiple services, and maybe it's like interruptions. One word that Oliver and I were talking about in this particular case, I guess the key word here is independently. So because you want to create just one single process without affecting the other ones. All right. Something there. I was trying to catch what you're saying. And then we have it on the goal. Upgrading process types independently. Reducing risk, fine grains, monitoring. So you could actually roll back independent paces or call it off. You know, I mean, of course, this all requires people to do that work. But all right, I'm going to move this. Oops, I went too far. I'm going to remove the goals down from the developer. And then I think we saw it because we're four minutes over. So under the developers talking about different organizations or groups that are all doing maintaining and we want to make it independent so that they can work by themselves. So I'm going to move that goal down. Where'd that go? We've got to totally fix all this bottom area. It's hard. All right. And then we're saying test coverage is difficult whenever they're totally coupled. So again, developer. Oh, so we have reduced the complex. So we want to have loosely coupled, reduce the complex of individual components and move towards loosely coupled components. By the way, Taylor, we are top of the hour? Yep. We'll stop here. This is a good chunk. I think we can, if we can move, a lot of this was out of the document that Oliver copied over. There's a blog post that's linked. Yeah. I think if we can clear this section out and move it into either goals or turn them into challenges, then that'll cover most of it. The out of scope will be pretty straightforward. And the proposal is based on content that is right out of the document. I think the challenges and objectives were a more difficult thing to cover on this. Yeah, great. All right. We have progress. Yeah, definitely good progress. And I'm up for working with you again this week if you're available. Good session. Thanks, everyone. See you next time. Okay. See you. Bye. Bye.