 Cool. So we're five out of the hour now. You want to take it away, Friedrich? Sure. So welcome to the network service match meeting. So to get started, let's go ahead and do a agenda bashing. So is there anything that is not on the agenda that people would like to talk about? I have more or less copy pasted some selected points from last week, so there's not really like an updated agenda. I don't know, maybe we need to. I also went through and tidied that up a little bit. Okay, good. It's not the right word. I went through it and I think I added one or two things as I went. Perfect. But I think probably the right focus is kind of where we arrive out, which is the release when we get there. And we do have Alex, who's been kind enough to join us to talk through progress on the logos as well when we get to that point in the agenda. So, but I'll leave it to Friedrich to go walk us through. Okay. And so those that's been added to the agenda already. So we're good. Okay, so we have so events wise we have this particular meeting which occurs every every Tuesday, we have a document sweetening, which occurs every Wednesday at 8am. And we have a use case meeting which occurs every other week on Monday at 8am. So the next this coming Monday is the CNC of telco working group birds of a feather. So, sorry, that's been renamed to the telco user group. So it's not Monday. Yeah. Sorry. It's the first Monday of the month. First and third. That will be May 6, May 6. I believe you're right. So I believe that all the other Monday Mondays are belonging to us. Yeah. So this brings me to the point because we have an extra Monday. Do we do we use the Monday or do we keep it free? Yeah, I mean, I'm bonus Monday. Yeah, already I think Ramki has scheduled it for you. No free Mondays for you. And that scenario, it occurs every, every other week, except for the last Monday, if there's an extra bonus Monday. Yeah. For the use case document. So, so we will be having we will be having a Edison use case meeting this this coming Monday. I think it is technically the second and fourth week of Monday and possibly the sixth on really odd months. Yeah, second, second, fourth and fifth. In fact, yeah, it'll be great if you can join. I think we made some very good progress last week. And you also identified some solid open source implementation of EPC to sort of, you know, kick off something real right I mean, for Mac project so it'd be really good if you guys can join. I'm glad yesterday I had something really important that popped up. I was not able to avoid, but I should have no problems with future Mondays. Awesome. Thank you. Oh, and also, on the, the, our webpage, is there a way to sort of point out that the use case calls are on every the second and fourth Mondays. I mean, at least Prem was saying there's no easy way, not logical because right now it's like, you know, every week is the indicator right. So we changed it to seven fourth and fifth Mondays on the document. Probably on the calendar. Is that what you're looking at? Because I can do that. Yeah, that'll be cool. Basically, yeah, just want to, because there's this confusion is that at use cases that every Monday or other than it is right. People get very confused. Yep. Great. So we've had a container world in last week that Prem gave a talk I saw lots of great pictures come out of it. Great comments. And so anything else that you want to share, Prem, before we continue on. Yes. Actually, the container world, this is basically focusing on Kubernetes and containers. And this year they have added the service mesh as a separate track. And basically, the I gave under the service mesh track and people are really excited. I was requesting people to probably join our calls and then start contributing to it. But the talk was well received. What I did was I brought in the concept of 5G and edge and also explained about Kubernetes CRD and then how it all connects and how network service mesh is going to probably play an important role in the network transformation is what I highlighted. Yeah, I can. Great. Yes, the presentation I'll probably put it sometime in the wiki page so that people can access it. Cool. So we have to con EU coming up in Barcelona in May 21st to 23rd. We have to book your hotels now if you have not done so. If you forget if the prices get too high around the era. They do have good public transportation. So you can book along the rail. We have intro and deep dive talks over a coupon EU. So feel free to come and join us on those. Also giving the demo booth so Frederick would be would probably if we want to take that offer we probably needs someone to man the booth in the elephant. So is there anything we want to do about that right at the moment I mean I could spend I should be able to spend some time in the booth. Awesome. So I'll probably include you the thread with the Brandon. So I told Brandon that you had I spoke to you and then you have agreed to help it. One thing as well is so anyone who anyone who is here is also welcome to join to join in on the booth or help out as well so it's doesn't doesn't just have to be it doesn't just have to be me so if anyone wants to be wants to help out in that way like will will help you get up to speed. So this is this we can probably use the same. EnSM demo with Odl and it's simple use case I have an instruction that you can just follow and then you can have it and this for your information. This is available in the packet.net, which means you can access it from anywhere without the VPN and you should be able to show the demo. Nice. So one other thing as well for for that particular demo is how difficult would it be to set up a set of scripts or something to for people to reproduce it on their own. Okay, that's a good idea. Let me check on that I'll get back to you on that. Yeah. So, so let's get through the agenda to get to the main item as well so we have Fido many summits coming up. I have a talk that's been accepted about EnSM and Fido and what's gone well what's what's been problematic so we'll be giving the talk on that. Both Nikolai and I have submitted presentations for cloud native network service. So we will see. We'll see what comes out of that. I don't think they posted anything on what's been accepted yet. Yeah, it should have been posted by now. I think I haven't seen any emails about it. They usually post an email before they post scandals. Yeah, I mean, CFP notifications. Yesterday. Running slightly behind. Yeah, probably. Good. Okay. So we have QCon China coming up in Shanghai, where Nikolai and I should be giving an intro maintainer track in the main conference. We have ONS Europe the call for paper is open for anyone who wants to go to Antwerp and talk about network service mesh. The call for paper closes in mid June. We have QCon and MEF in November. So MEF is more of a telco focused thing. It's in Los Angeles. And QCon, which is going to be our primary focus in this time period, is going to be in November 18th through 21st. So the call for papers open up in May. So with that, we now have a active network service mesh Twitter account. Special thanks to Lucina for taking that up. And is there anything else that we want to talk about on the social media side, like we have the word team. So I assume that there may be more than one person contributing or perhaps there's a slot open for people to help contribute. Okay, that's about the number of the stars that we have, right? Okay, 114 followers to start to 369 following though. So let's see how, let's see how that changes over time. Okay. An easy way to do that is make sure to stick the end service mesh on the presentations. So with that. The other thing I wanted to make point I wanted to make is Lucina has been doing an amazing job on the Twitter account. We've had so much cool traffic. Since she took over last week. Thank you. Thank you. I do. I suspect most of those 114 followers probably came from. Likely came from her starting her interactions. Like we were, we were not that we were not active on it at all. So great. So we have a new. A local conversation. So Alex, you have the floor. By that. Can you hear me okay? Yes. Great. I will share my screen. I guess that's probably the best. Okay. Do you see now? Yep. Yep. Okay, great. So I revised some of the logos based on our, I don't know, I don't know. I don't know. I don't know. I don't know. And I got, I got rid of the spider as we discussed. For now. And I tried to stick to only doing the horizontal versions until we decide on a direction. And then obviously stacked in icon will be easy. I know. NSM will be there. Another version, but I wrote out the full word for now. So yeah, I, these are obviously rough. I don't know if that's, I'm not typing. Oh, can you hear me? No, no, no. You're fine. Oh, I'm sorry. I'm not typing. I'm not sure why it's doing that. Okay. Great. So can. Yeah. So I'll just scroll down. I guess as I did last time. And you can just let me know if you see anything or any directions. Is it possible to, to get this out there? Should I go put, put this up and we'll continue the email thread and get some comments from there as well. Yeah. The little globe mesh thing is kind of cool. What do other folks think? What are you guys seeing that you really like? I like the one with a 3D effect. It just went past. Yeah. Yeah. That, that one I think is really cool. Yeah. One more. Yeah. This one. Exactly. Yeah. Yeah. No, I, I, I kind of like that one too. Yes. There are, there are a lot of good things. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. There are, there are a lot of good things here actually. So I do kind of like that. Yeah. And then I'm kind of fond of the, the swirl that's in the, the sort of round swirly one that's near the top now that's two over. That's quite nice. Yeah. Great. Yeah. So obviously these are rough. So I feel like once I get some. I'm going to be able to review them. Like a couple of them, I'm going to do a few things. And then I'll be able to review them. I'll be able to do a few things like a couple of things, like a couple of things I like or a couple of things that I like or a couple of like or whatever. I'll do, I'll do the same thing again. And I'll be able to revise. And make them look better. That seems wise. Cause there's a lot going on here. Yeah. So would you rather me just send you the PDF and then we can work. it's a chance to like the comment on it. Because one of the things that I found super useful about the conversation that we had with the last round was there were so many perspectives that would never have occurred to me that turned out to be really good. And so I think, you know, that ended up being very, very positive. Great. Okay, sounds good. I'll do that right now. Yeah, it makes me feel like it's a global project with some of the logos, which is a great thing. Great. Okay, so I'll do that right now. Thanks so much. So the new conversation is to continue on the mailing list. So if you are in this meeting and are not aware of the mailing list, you should be able to find that mailing list on the website. If you don't find it, get ahold of me or Ed on Slack and we'll help you get there. So speaking of IRC versus Slack. So we had a discussion before just a reminder that we are starting conversations on Slack. So there is a link in the agenda. It is slack.cncf.io that will get you an invite into Slack. And the channel is Pound NSM. The Pound is implicit. And there's also a Pound NSM Dev, which you should probably get on the website in other places. I think Nikolai correctly observed that Dev's hatching back and forth on some things can be a little dense. Yeah. Would you please check my PR? Okay, so I'm sharing, I hope my screen is in. So this is on our main site on community. We have the, in the communication channels, we have both mailing lists and the Slack and the invite link also. Cool. Yeah. And also the IRC channel has been updated with that information as well. So if people don't get it all, if they eventually notice the topic, they should know where to go. Great. So continuing on with that, we now have the Andromeda, we now have the Andromeda release that is coming up. So there is an Andromeda backlog and Nikolai, you have the floor to talk about everything with regards to that. Okay. Thank you, Fred. So first, a little bit of an update that I need to actually create a PR for this to update the specs also. But the original idea was that we code the branches NSM, Dash, V, whatever, whatever. But I got inspired by how Kubernetes are doing their branching and tagging so essentially the tags are done this way with V and then same bear. And then the branches are release dash and then the manager and minor name of the tags to be released actually, the version has to be released off that branch. So that's the, I don't know, proposal to be done. I don't know. I mean, I don't think that we have time to vote here, et cetera, et cetera. So if anyone has any objections regarding this, then just say it now and then. I think we did kind of previously agree on this part of the release in terms of dates and whatnot, which I, you know, several weeks ago, which I think is good. And so like the major thought that occurs to me is, you know, when do we want to pull the trigger today on pulling the branch? I want to go check and see if we have any major bug fixes out before we pull the branch. But then the other thing we're going to have to be careful about once we do pull the branch is making sure that any bug fixes to master, get sure you pick the release and vice versa. Exactly. So one of the major things that I wanted to have in and thank you, Ed, for actually merging these guys, the CircleCI was leaking resources. So with that in, I'm more or less comfortable in, you know, just branching and then monitoring which fixes should get in both branches, et cetera. But that's probably, probably just me. So we need to make sure that we're really diligent on this because what's going to end up happening if we're not careful is this is going to turn into a very manual process. So my recommendation is we start looking at what others in the community who are doing something similar do because there's bound to be some type of bot or other similar thing we could use to test against various branches. And when we commit, we'll then commit into the other branches as well or merge into the other branches. So I think that we, I think we should be a bit careful with this because worst case scenario is we end up having to manually cherry pick every single thing that's that's every little patch that's out there. Yeah. I mean, that's, that's of course, yeah. Yeah. And as the person who's had to have, who's owned that role in the past or other projects, these are a lot of times. Yeah, I know. I know. I mean, I have also been in this, but I mean, my main motivation behind branching is that once we have the branch because, I mean, once we have the branch that there's even more kind of, you know, motivation about moving, moving towards the release. Otherwise, we can see just having everything against master forever. And we don't have much time. Yeah. And in the meantime, let's also add something to the, to the worksheet that when you, when you commit something that's to, for the person to indicate whether this is something that they believe should be cherry pick back to the stable release or not. And conveniently, we have a label. Yeah. I think that this one is. Oh, I mean, I've mostly been using that label on issues. But you could also use the ours. The labels work too. I'm happy with labels. I like labels. They're label-y. Yeah. But what Fred said is that, yeah, I mean, the main testing would happen against master, I guess. And then cherry picking this to back to stable could be a bit, you know, manual testing. And then, I don't know, maybe we can invent an integration branch, but then, yeah. Well, we have an integration branch. It's called, it's called release. So I think if we, we just need to, what we need to do is just see what's out there, just to see if, if we can find something that'll, like, when you commit to master and if it's labeled with v01 or it's labeled with release 01, then it also runs the CI against those as well. So maybe it's possible we circle CI may even do it. So we just need to look to see what's out there. I see. Okay. So we can do this and just automate the process. So I think the tagging and so on that will still be manual. Like that's, that's on us to do it. But, but we don't want people having to rerun the, the automation manually. Okay. Okay. Let's, let's quickly go through what's, what's going on with the issue. So we have a couple of IPv6 related PRs that have been, that are being worked on. I think that we have a good progress here with IPv6 payloads. The problem actually is that this PR treats mainly the testing, not really the implementation. So probably for the time being, we're just going to integrate it the way that it just works for IPv6. And I mean, it just tests it and then it fails. And then we have to figure out how actually it will work. So most of the others are kind of ongoing. I know that the Asia, isn't that already in, that should be in already Asia. Yeah. Yeah. It's already in. Yeah. Okay. This one is almost been being completed, I believe. I'm not sure what this one is. I mean, these are the, the PRs that are on working. I believe that we are kind of having some hold on this. And then we have the backlog is a little bit longer than what I would like to see. But I think that we still have some time to figure this out. And maybe if we just focus all our efforts on the, only on the labeled ones, then probably, probably we'll be able to fix this for the stable, which is actually supposed to happen in three weeks. So that's, that's more or less the status. I don't know if anyone has any specific questions, wants to outline something. I know that this one is also being worked on. Right. AWS. Yeah. Yeah. Yeah. We have a little bit of probably take depth around occasional restarts of network service manager. I mean, like in different restarts scenarios, we have something, I know that Ivana is working on something. And also Andre and Ilya are also contributing to hold this. We're having a stable restart and how to heal mechanism. But apparently there are some corner cases that needs to be cleared. I would say that these are the main. Actually, we have a great news for the last week. We have integrated 202 from the VPP. And this apparently is also a lot of issues and close a lot of PRs, which was actually really great. So thanks for all the guys that actually took this work. And that's more or less it from me. Guys, can I just add one thing? Because I'm trying to get in this and test it on all the public cloud. But before that, there's some details I would like to learn. Because I know, notice that some of the folks have already done jobs on GKE or Azure. So basically what I noticed that is that what we want to do is to do like say, to build and test and to try to pass those CI tests. So one thing is that we'll kind of like say, will this bring any changes to the clouds like say bottom like say network infrastructures? Because we'll really like say touch those data planes like. So because that definitely are going to arise some issues on that. So I'm not quite sure about that part. No, I mean it doesn't touch off for the time being. We're just running NSM as an overlay networking clear on top of the existing infrastructure. I guess that once become kind of mainline and more mainstreamer than what we are today. Maybe graduate from CNCF, I don't know. Then I guess that there will be some kind of integration with their existing networking solution. So that you should be able to just run network service manager on top. And then it will talk straightly maybe through some agent, I don't know, to the infrastructure of the clouds. But I don't know if there's someone that knows the internals of some of these clouds or if you know something about Alibaba or Huawei or whatever you mean is there. Maybe we can work through some solution that actually integrates with the real networking infrastructure of the public cloud, which would be even even better. Okay, so if I'm not mistaken like saying, for currently we can trade like what we are planning to say to like, for example, get a Kubernetes cluster on public cloud and just try to run NSM somehow like an app, something like that without touching, without really touching the lower level, like say, yeah. Is that true? That's exactly it. So we just run the, the existing integration test basically, they just run the NSM, the NS Manager, Damon Set, and the BBB agent data playing data set, Damon Set. And then that's the infrastructure bits of NSM, but they run on any public cloud. And then you can run through, put it whatever network service endpoints you'd like on top of that. Gotcha, great. So go ahead then. There's an easy path to extend out the current existing stuff as well would be someone could treat a VPC or some other similar thing as a network service. And so it becomes the subnet or the bridge domain that you, that you, or we'll say subnet is there because cloud services is being more layer three. So it could be the subnet that you connect to, and then have VPNs connected through as part of that cross-connect. So there's other possibilities as well that would be possible without having to actually manipulate the underlying infrastructure at the fabric or data center level. And of course, if the clouds feel that they can gain benefit from adding NSM into their infrastructure, then of course we'll help them out to work out what would be a way to make that look like. But there's plenty of use pieces that require zero cooperation from them, other than they don't destroy their current APIs that people currently use for production. Cool, thanks. This is very helpful and I'll let you guess now if there's any update. Okay, great. So we have 20 more minutes and an empty agenda. I mean not empty but nothing. Almost likely have an upcoming release. Yeah. You know, this is really great because usually when releases come for most projects, it's like you're running around, everything's on fire. That was last week. I mean, the Ardham cover is super bugged and there's the fix for it. So we'll say that was in VPP. Well, what's super funny is most of the things that we've had to fix in VPP aren't actually VPP bugs. Most of them are kernel bugs that we have to hack around the VPP because the kernel is not going to fix them. This one I think legitimately was a VPP bug. So it does happen. Yeah, okay. Well, we'll say it was a kernel bug. Well, great. Well, is there anything that anyone would like to talk about before we wrap it up and give people some time back? Yeah, I have one important agenda. Sorry for writing it late. But I believe we would need to update our documentation so that we reflect the exact state of NSM. There's a bit of old models spread across different documents. So thoughts? So that is on the agenda for the Wednesday stuff for the Wednesday calls. If there is a specific document that you think needs to be updated that you want to call out, definitely call it out. Sure, okay. So I think this is also like, for example, the concept like channel. If you recollect when we started, we had the concept of channels and other definition around that. Of course, what I'm trying to say is the what is NSM doc that's Jeffrey pointed that out as well. And that's that is on the short term agenda for the NSM documentation goal. Right, yeah. Either we can delete it or we can just create an absolute folder and then put a disclaimer saying that the concepts have changed. Yeah, I'm okay with we had set it to draft and it didn't do anything. So I'm okay with if someone wants to delete that document. It's in CI so we can always or rather it's in Git so we can always pull it back up. But just just to get it off the website for now. So if you want to submit a pull request for that then once it passes I'll merge it. If you want to be a little bit more proactive in this or rather another option would be actually just delete it. We'll just deal with it. Is there anything else? Great. So I will be expecting a pull request from you, Prem. And thank you everyone for making sure you sign off on it because it'll actually Yeah, only challenges it has to be done. And there are like lots of other documents it's spread across. So I think I need some time. I'll probably get a consolidated list of documents. We can have a quick sync up in our doc call and then conclude it. Perfect. Yeah, that'll work too. And then I, okay. Did I explain multiple pull requests? Anyways, thank you everyone. And we will see you next, next week at the same time. And take care. Sounds good. Bye. Thank you. Bye. Bye. Yeah. Bye.