 We're going to wait until about five after so another few minutes here and then get started. Hello. Hey, good morning. If you have any agenda items you can add them to the meeting notes as well as your name. Yeah, I didn't know it earlier I haven't joined for quite a few sessions of this I just wanted to discuss the dashboard and the activity on arm overall. It was really a more free for discussions with the group. Yeah, we missed the meeting up at cube corn and with bank holidays or other vacations it made things difficult to attend the last few calls. What may how we can go forward with the group. Yeah, I really wanted to get more interactions and contribution from more people. Open it up. Are there any walk in items right now there's way to that. Anyone else have anything. Yeah, I, I really put this as a placeholder as as tracking the other work happening on CNCFC I and looking at the respective project. Which are there I just wanted to understand a little bit on how much more work was happening and also in the tracking of what works or what doesn't work on the respective project highlighted into the CI dashboard. So it'd be good to to get your, your view or your status on that and and and see how we can move from there and I agree subsequent question you raise into how to get more participation into that activity is a very valid one too. Sounds like on specifically the CNCFC dashboard. Yeah. And do you have any knowledge of other arm CI stuff you can talk to you mean which are outside the scope of the dashboard or yeah, outside of the scope of the dashboard. Yeah, so the arm team is is still progressing work around the Kubernetes conformance testing as you know so that that's work which is progressing and there are other components of cloud native elements that our engineering team is contributing architecture support say but not so much focused on to the CI elements. So I think this page on the CI dashboard still captures the main core projects as a downstream CI for for the validation exercise of the branches you're calling from and still value in actually highlighting what passes or what doesn't pass in terms of build test and deploy on to the arm platforms that packet. All right. The tricky part is I think how to monitor when things fail whether the failure is is what the failure is attributed to whether it's particular commit so if it's an architectural issues that would need to be fixed or whether it is something else which is elsewhere in a code and we just resource after a few additional changes in a string code base right. Talk a little bit to CI dashboard initiatives that's this and I'm having some browser issues I can see is shown a little weird but you can go to Sancia.ci and do a look at it. And the current iteration is hardest talking about the test environment that projects are running in. And then you have all the different projects that are running on it. Overall right now the initiative is in a maintenance mode. We have. We had talked about the net and as part of. Maybe one of the comments that you had, what, how do we determine whether it's a. Maybe I build issue with it, maybe the environment or the upstream project is having a build problem or whatever it may be. So one of the things that we started moving towards and was the long term goal to shift everything over was integrations with the upstream project so linker D2 might be an example. So the linker D2. Is actually doing an integration. The first iteration of this. While it shows get lab what's actually happening is it's not doing a build. There's a an integration to a proxy over to whatever the projects. CI system is so linker D2 is on their own. So it's going to be here. It's actually Travis the I bring this up. And then it goes out and looks at whatever the project is doing, and then gets the status information. So rather than having everything internal, which say core DNS and a lot of them, these were built builds that are actually running the builds and creating the binaries in get lab, doing that it's pushing like external which is already done. And then what happens when you look at different architectures like if we go over and look at say arm. Then if you see here. Yeager is showing in a essentially what this could mean is Yeager doesn't have public CI builds on arm. So ideally what would happen would be to work with the projects to get public builds of all the architectures and then the different versions. So over here, those aren't familiar and Texas overriding this but it says Kubernetes underneath. So this is the stable and head of Kubernetes and testing do these work. So maybe they do public builds of arm on stable but not on head. Rather than having it all internally and redoing the build systems that are already happening, try to help projects to have more coverage on at least the stable versions of the projects and stable versions of Kubernetes that are running daily, which would include arm. That's the goal there. And then the next part which you can see all of the deploy is great but if I go over here to x86 stable, I think was. So some of these are filled. Actually, it looks like they're all filled for whatever reason. But if these pass versus the in a what this means is we have artifacts or the the binaries or containers and everything for the project are created. What we would want for the second step to work with projects would be to get them to publish the artifacts, at least on every release, ideally on when they do commit some other things. If you have the artifacts from the CI builds publicly available, then we would like to pull those down and use those artifacts for deploy the test column is for integration. So this started, as I was saying, and some of these tests and the tests, these are actually integrations into their CI systems. And towards that, and then a lot of the focus got shifted away. So, I think the next step would be to see overall interest from the projects and the value that it can bring to them and then contributions for supporting the effort to go forward. So essentially, we have a prototype and the idea of, okay, this looks like we can do it. We've started some integrations. We have that configuration, maintainable, so tough to do that. So we have the prototype for that. And if people are interested in us, and that's where we go, which would help with those false positives, believe that you're referring to. So if we're integrating externally, we'd be able to do that. And then you have this more of a health status dashboard for all the projects as a focus and integrate with all the different, whether you've published a Docker hub or somewhere else. That could be wherever and taking all the different places and aggregate and show the health of the projects. And from the arms standpoint, what that would mean is finding, encouraging and working with the projects to publish those run those, and then using some type of standards for showing the status showing the artifacts and making those as available as x86. Okay, and that's for graduated and incubating projects as you have on to this list. Graduated and incubating the only other one that we have as a subset of own app about having other dashboards to show other things, whether that's maybe sandbox projects or other projects besides the CNCF projects. And we had some initial work and there's a lot of, I guess, design around additional screens that potentially could show more projects. I take on too much though we want to keep it focused which is where we thought maybe doing the integration and then helping to help with projects to publish and use some type of standard that makes it easier to integrate and show the status. Okay, now that that sounds great actually and in your discussion with some of the projects like Diego and similar which you've shown a proof of concept on x86. Have you also discussed the arm angle with them or and if you have planning to we can we can participate in that discussion and see how to stimulate interest or how best we can we can support that. We have had a few discussions. I, before we dive in I think we need to figure out how how to get more contributing besides a small group, which is part of the things that we can take on more, but I'm happy to talk with you about some of those plans and ideas and then we can see how that could go. Okay, I would be happy to do that. I mean the ultimate goal I agree with you is to reuse or plug into more of what the upstream project is doing because that's where most value is right so the more we can go to that direction the better long term that is so if the there is a two words having a kind of status reflecting the how the upstream is and rather than rebuilding things downstream I think it's much better. Right. Ideally status, what we were envisioning, which goes back kind of to the start of this, but it would be a little bit different on the timeline, but you could see the status of all the different stages for whatever people publish whatever the projects are willing to publish. And then beyond that, if someone was interested in a certain build. You'd be able to click there and get the artifacts for that so if you said you wanted Flint a tested version of Flint day on the latest stable Kubernetes that built on arm. You could go through and have a direct link to the published artifacts and you could go through and find all of those pieces quickly and see how they fit together, including being able to retest yourself. If we're saying you're looking at something in production or here. You want to build a prototype and you want to match match those. So the long term would be a matrix of the different compatibility testing that you're seeing. And what's behind the dashboard is a API server that provides the stats via JSON. So the one potential that you could get out of this would be some type of subscription feed or whatever you want. Where you can get the compatibility testing or track the test results from common form. So it's again, what type of interest from the community and point more of the efforts have been shifted to some other initiatives. And I guess I'll mention that it's just they do have CI but where things are. The CNF conformance and then CNF test bed. So these are all telecom related CI capabilities in this and there is actually a CI system. These examples and potentially what could happen on this, which would be a relation I can relate it back to the dashboard would be having status updates for different use cases. So right now what we're looking at is individual projects and deployment. This would be more like an integration test or testing with full examples. And then the other initiative that's related would be the CNF conformance of cloud native network function conformance. So this is testing how network applications for telecom are following cloud native principles. This actually does run all of the this is is a test suite and it's running in Travis. So what's not happening or what's not here on either of these initiatives as a dashboard but theoretically you could do something similar and say we what is the status of a group of tests that you care about that may be testing a certain set. So if people are interested in dashboards outside of say that the CI system that's running a specific project, then that would be a possibility to take something like this and and build something for another project or take the idea and have it refocused. Maybe we don't care as much about build that you want to test some type of integration between projects. Okay, yeah, I have a new colleague who is on I'll try to connect him with you is going to look more into the CNF aspects where I will try to do that in the coming few weeks. Sounds good. Because shy who I had introduced before I moved company so it's, you have somebody else who is now looking into that. All right. Okay. And in terms of community do you have beyond arm do you have other interests whether from developers or companies who are willing to join contribute or how is how have you discussed that. I'm interested in the on the telecom side. So, and way, looking at contributors and testers for the conformance CNF conformance. The test bed has some folks from places like Intel and some telecoms that are trying this out which is it. It actually deploys a environment packet to run the different tests and builds out everything automatically the tool chain. And there's people that are looking at testing this on and testing this themselves with their own packet accounts and also testing it on self hosted machines within labs. There is the CNCS dashboard, whether this dashboard itself the one that CNCS CI has or the software, there's been. There's not really been much traffic and request to update anything on the dashboard lately. There was, I guess, over the last couple of months with, I think with COVID-19 and kind of turn everything else up in there. But in the last couple of months, there's been a few mentions on different slack channels or maybe someone mentioned something in a Twitter issue related to a build problem. And there will be some someone pointed something out and it's noted. So it I guess it's somewhat still seen, but we haven't been pushing to market it to see if we want to switch and go some other path, maybe to help with the false positive stuff. I think before continuing to push and say, let's use this and getting more projects, it's kind of, we need to decide if it's going to go forward. So they're going to be enough interest to help contribute or help make that happen so that we can bring it to that next stage, because right now it's in this in between with, you know, core DNS is doing builds, running the test and then building the artifacts and doing the deploys on boy, etc. And then you have some of them that are doing integrations. But they're only, we're only partway there. So ideally we'd get shifted fully over. And if we're going to put in that effort, and we need to make sure that community wise it's going to bring value and there's have at least some other interest of party, maybe to help in some fashion. Yeah. I agree. It makes sense. Thank you. Yeah. Does anyone else have any comments or questions are totally different. This brings up other ideas. It was on here with every talk about the play. Well, if there's no other topic. Okay. No, I don't have any additional topics there so we can, we can follow up offline. Yeah, sounds good. Native related CI would be good, but we got to get enough other groups of people interested. Yeah. Okay, thank you. Have a good day. Bye.