Revolutionizing Software Delivery: Unveiling the Power of CI/CD and GitOps – The CD Pipeline Show EP 10

ai. And you are watching another episode of the CD Pipeline. The CD Pipeline is a joint production between Techstrong and the good folks at the CDF, the Continuous Delivery Foundation, which is part of the Linux Foundation.

And we've been doing, uh, CD pipeline shows now for, oh, almost a year, I would guess. And, uh, we cover obviously topics relevant to the ICD continuous delivery. We've looked at a lot of the latest tools and methodologies and people in this space, and it's usually a 45, 50 minute, great discussion around important topics, especially to our audience, which tends to be DevOps, cloud native software delivery folks.

So, with that being said, let me introduce you to this episode, which is about, uh, GI Ops, right? We're, we, are, we're talking about how, uh, the relationship between CICD and GI Ops and, and we're going to dive into, many of you out here may not necessarily, you may think you know what GI Ops is or you have, have a vague idea, but we're gonna try to nail that down for you as well. Couldn't think of a better panel to do that.

Let me introduce you to them. Now. First I'd like to introduce, and, uh, I'm gonna do my best on names, but I apologize.

Natali, Vito joins us today. Natali, welcome. And if you wouldn't mind introducing yourself to the audience.

Absolutely. Thank you, Alan, for the invite. I'm, uh, absolutely excited to be here.

Uh, my name is nato. I'm, uh, leading the developer advocates team in Red Dot. Our role is to be technical evangelist for cloud native, develop open source technology.

I am a CDF ambassador in 2023. I am, uh, I'm, uh, personally involved with the, um, open source and passionate about CICT and tecton and all the open source technology around that. So thanks for the invite.

Thank you. And it's a pleasure to have you here. Thank you, Natali.

Next up, I'd like to introduce you to Christian Hernandez. Christian, welcome. And maybe if you can introduce yourself.

Yeah. Thank you, Alan. So, uh, yeah, my name is Christian Hernandez.

I am a, uh, open GI ops maintainer. I am also a contributor to the Argo City Project. I am, um, involved in all sorts of open source contributions around GI ops and cloud native technologies.

Um, I'm currently working at Acuity as, uh, head of community there. Um, acuity is, um, was founded by the folks that started the Argo Project, so I am very deep in the weeds in Argo and GI Ops. Excellent.

Next up, and this is where the names start getting hard, I'll do my best. Philippe, I Philippe, help me with your last name. Okay.

Hold on. Night, night to be with you. So my name is Philipp sgi.

Um, I'm French, I'm working for Range. Uh, I'm the VP Software engineering, uh, and I'm in particular in charge of the cloud native, uh, networking and telco transformation. And the story with, uh, CICD and GitHub's, quite simple.

I would say that for, for ages, we use, I would say physical, uh, network appliance. Then we move to virtualized ecosystem. And now we are about to create the next generation that will be cloud native.

And when you have cloud native networking and telco, the only operating model that we can apply and we are building is totally based on GitHub. So here is, uh, basically the connection with the topic of today for my site. Excellent.

Thank you, Philippe, you pronounced your name much better than I ever would've, so I appreciate it. Next up, uh, we're gonna do this one. It's a hard one too for me, but it's ddu duii.

Uh, thank you, Alan. It's dsi NAI Tanika, and I am currently the chair of the board for the Continuous Delivery Foundation. Uh, I'm also on the Technical Oversight Committee for the committee, continuous Delivery Foundation, as well as the Technical Oversight Committee for the Spinnaker Project.

Um, spinnaker's used by a lot of folks for their good ops operations, and we continue to make improvements to it to support the, the effort. Excellent, excellent. Um, last but not least is my co-host of CD Pipeline, and I haven't had a chance to be on a show with her since the new year.

Certainly as well, the one and only Lori LaRusso. Hi, Lori. Hey, Alan.

It's great that Valentine's Day is tomorrow because there's nothing but love in this room for, uh, GI Ops and the CDF. Oh, I like how I did that so cheesy. I like, I was a guy.

Love it. Segue. So, yeah, so I'm the outreach and marketing chair for the CDF.

I'm also A-C-N-C-F ambassador, and today is a really cool topic because last year at CD Con, if y'all will remember, it was a combination between CD Con and GI UPS Con. And that's when I met Christian. And this is when the whole world of GI UPS sort of blew up in my face.

And I was like, oh my goodness. So we've been trying to do this, uh, panel for a while, and I'm very excited that we've got it going on today. Absolutely.

Um, okay, I think that's everyone on here. Lori, I'm gonna ask you if you want to kind of kick things off and, and, and then, you know, it's open to our panel to join in. Thanks.

So I think let's start back at the introduction between CDF and GI Ops Con back, um, last May. So Christian, you were the chair or one of the chairs of GI Ops Con, and why did you think it was a good parent to have the CDF and GI Ops con together? Well, it's, it's all, uh, it's all related, um, into the same processes, right?

So it's, it's having, um, you know, many people who are using, um, the CI tools from the CD Foundation, um, are also using tools, um, that are, are using GI ops, right? That are some of the, the GET ops tools. A lot of the things, um, that people are doing is all, um, it's all related and it's all, uh, like one big happy family.

So I, I thought it only made sense for, uh, for the CD Con and get Ops Con to kind of come together to exchange notes, right? Like, this is what all these conferences are about, about, uh, exchanging notes about, um, you know, talking about best practices. What are you doing?

What am I doing? Here are some great talks. And so it's, um, to me it was only a natural fit, right?

So since get ops kind of focuses on that last mile delivery of, of your applications, um, you know, why not step back and kind of just, um, embrace the whole ecosystem of CICD into it as well. Fair? Yeah, it was, it was, I thought probably one of the, and I'm gonna be super, um, biased here, but a super well-rounded, um, content on both sides so that, you know, users or attendees from either the GI op side or the city con side really could just go back and forth and get that well-rounded education on, uh, practices, reference architecture, all that good stuff.

Um, so Natalia, I see in the background, uh, underneath your beautiful Red Hat fedora, you've got the GI Ops cookbook. Why don't you give us a little information about that and then we can kind of dig into the, to the topic? Absolutely.

Thanks, Laurie for the opportunity to also show my, the book I have in the background. Here I am, I, I wrote this book with another colleague is a, is a cookbook showing this cooperation between the CI part and the CD part. And there's a, um, you know, pipeline plus giros.

0, this evolution in the books where it's a kind of, we, we put kind of a climax where we start from the scratch from Git and then we arrive into containers, and then we arrive into pipelines, and then we arrive into giros and all multi cluster giros, uh, deployments. Uh, so that's the, the book evolution and then the book's narrative. And, uh, this is a cookbook, so contains receipt to, to use it in, in this particular case, those are tecton received and Argo CD received, but the concepts are agnostic.

Those are valid for, for any, uh, tool and any tool like that. Fair. Hey, Christian was one of the reviewer of this book, so thanks again, Christian.

Yeah, No problem. Of course. And so, Philippe, I think like in your introduction, you were kind of talking about how GI ops was the only way forward.

So give us a little, a little bit more about sort of what you're doing, and then I would love to hear the panelists take on how that, how that works. Okay. So, um, what we're doing is quite simple and, and for us, the relationship between the CICD ecosystem and the GitHubs ecosystem is basically a continuum.

I would say that, um, in, in we have perhaps a very strongly opinioned way to build a platform, deploy the network and telco services on top. But basically everything that sounds CICD, I would say is pre-production and build. And every, everything that is, I would say, live and deployment is a hundred percent under the control of GitHubs.

And something that is extremely interesting into this transformation is that when you are moving from environment where you, what you are basically, um, deploying and managing or virtual machine, and when you are, um, entering into this world of continuer that have been introduced by Natalie, I think it's super important that we are having four new superpower that are totally incremental. It's about the residency, it's about the self healing topic. It's about the close reconciliation, and it's about the drift management.

And when you have so much complex infrastructure to manage when you are a telecom operator, I would say it's super important to have a super industrial and professional, I would say, way to manage your infrastructure and deploy the services at on top. And basically what we are doing today is that we are, we, we are building, I would say, um, our, our own, uh, Kubernetes distro in a Linex Foundation Europe project called Silver Projects. If you want to have a look at, uh, I strongly recommend, so with projects, we are using a cluster, PI Tora, I would say the infrastructure in whatever, uh, bare metal or virtualized environment.

And once it's intern base, the framework underlining is ity, where we are basically bootstrapping and making all the install. And then we are deploying, I would say was something that is called the network Integration Tooling factory. That is the professional way to deploy at scale the network function on top of our infrastructure.

And it's in this environment, a strongly opinionated, I would say, flexible deployment plus a tooling zone and a testing zone. And something that is super important is that the testing framework we are using at integration stage in CICD are exactly the same. We are running intent based in production for functional tests thanks to GitHub.

So we have this continuum that, for us is super important. Did you see, you wanna weigh in on that? Sure.

No, I think, uh, everything was said that could be said about that. That was wonderful. Um, I, if I do add one thing, I just wanna say that the, what's important in this process is that the tooling has to support growth and change is, as things in GI ops, you know, emerge and needs emerge as security changes happen, as, uh, networking needs change, the, the tools themselves need to be able to support that flow.

And that's what we wanna be sure that we're providing in the CDF. So when I first became aware of the term GI ops, right, it was really a cloud native term, right? And the idea was that we're gonna start doing sort of CICD within Git itself, right?

And whether you were using GitHub or GitLab or any other Git, right? It was the idea, well, the security, the testing, the, the CICD, what we're gonna do with writing Git was kind of a common theme. Became aware of, uh, Argo probably the first GIT ops kind of project, uh, we were aware of.

But we've heard several things here, right? spinnaker using GI ups, right? And Spinnaker sort of predates GI ups, certainly open GI ups.

I don't know if our audience is familiar with, with the project CNCF project, CDF project, right? Let, let's explore that. Like what are some of the, uh, Argo is another one that I think we mentioned, right?

What, what are some of the tools here that our audience, some of the names that our audience needs to be familiar with? Well, yeah, Christian. Well that was Great because so many projects have been mentioned, so yeah, Christian, go for it.

Well, yeah, well there's, um, there's a lot of, uh, GI ops tools, uh, specific to Cloud Native, and then there's a lot of GI ops adjacent tools. Um, so you mentioned, uh, a few of 'em, and I think a few of 'em were mentioned, right? Uh, the ones that come to mind immediately, especially in the cloud native, uh, community is Argo CD and Flux, right?

Those are kind of like the two ones that kind of floated to the top. Um, you know, both, you know, created, um, you know, at companies trying to solve a specific problem, decided, Hey, I think that I should share this stuff 'cause this is actually kind of, um, you know, kind of a, a, a neat thing that we're doing. And so those are kind of the, the, the two tools that, that emerge, um, from some of these, um, uh, uh, uh, ideas and, you know, need to have a way to manage cloud native infrastructure at scale, right?

And that's kind of where Open Get UPS came into play. 'cause there was like kind of these two tools that were trying to solve the same problems. Like, Hey, let's get together and, um, you know, let's discuss best practices and, you know, kind of define what GI UPS actually means from the people who are creating the, um, you know, Argo CD and the Flux Project, right?

And we, you know, um, and it wasn't just people from those communities or people from, um, other communities as well, uh, folks also from like, uh, GitHub, uh, folks from AWS from Red Hat, obviously, um, you know, pretty much all over, all over. I remember that first meeting we had had, you know, you know, over 50 people in it. It was, it was kind of a little overwhelming trying to define what GI Ops is.

And really that's the idea of Open GI Ops itself, is trying to say, okay, what are some of the principles that you need to follow in order to be, um, you know, in order basically define what GI Ops was. We didn't want what happened to GI Ops, what happened to, um, like the term cloud, right? Cloud became kind of sort of ambiguous and cloud became sort of, kind of like, it, it can, I don't know what it means, right?

It, it, it can mean, you know, a hard drive, um, plugged into my network, right? Or, you know, people think it's this or that. And so we kind of wanted to narrow down what are the principles of GI ops, right?

And kind of what, and what we found was it's, it's kind of, um, it kind of widened the, um, uh, not only what we considered what GI Ops was, but kind of trying to, um, open our minds to other kind of solutions that were kind of GI ops adjacent to kind of like bring them in to, um, uh, uh, to the, the whole GI ops ecosystem, right? Not just focused on Kubernetes, right? But also other things as well.

And so that's kind of where, you know, the, the, the, the cloud native aspect of it kind of landed. It's like, okay, well, you know, we kind of have cloud native stuff, but there's a whole other stack sitting right below cloud native infrastructure that could benefit from a lot of these things. And, you know, a a lot of folks were, were doing things with GI ops, um, or GI ops, like, um, you know, with things like spinnaker, with things like Puppet, with things like Ansible, um, you know, trying to, um, you know, uh, automate a lot of the, this, these things.

And now we're like, we're at the point where like we are, um, uh, we are adopting kind of some of these GI ops ideas further and further down the stack. You know, what I really like is that it's a framework, right? And that Open GI Ops is continually working on the framework.

And it's not saying you have to use these tools. Um, I love that we've mentioned a bunch of CDF projects as well as CNCF projects. I think that, again, shows the relationship between not only the two foundations, but you know, the power of open source and building on top of each other to make sure that you're building strong pipelines.

You know, CDFs mission is to make sure you're building good pipelines that are secure and that you can deploy fast. Um, so, you know, I think that's something that we don't wanna lose sight of, is that it's not set in stone. It's continually being worked on, and that's what the good work from your Open GI Ops is doing.

Um, so I think, you know, we keep, we keep talking about like the different projects, but do you guys have like a particular example of how you sort of built up like a, a GI ops flow? No. If I, if I may just perhaps, um, um, answering into your question, I think that, uh, the initial remarks made by Alan about, uh, the topic was extremely interesting because the name GitHubs could be misleading perhaps for our audience.

And it's not because you are using Git that you are doing GitHubs, and no one is better, in better position than Christian to say that one of, uh, the key, uh, I would say contribution of the open GitHubs is about the four principle. So it was super important to define the four principle, uh, of what is GitHubs. Because at moment in time and, and under the pressure of the market or whatever, everyone could say, hi, I'm, I'm GitHubs compatible.

I'm, I'm, I'm, I'm doing GitHubs. But if you are deep diving into it's absolutely not the case. So there are four key principle that for me are absolutely, I would say super structuring, uh, what is GitHub's or what is not.

And the name by itself could be perhaps misleading Christian, I dunno if you want to comment on and perhaps share about the principle, because I think it's super, super important to cut them. Yeah. So, um, the open GI ops principles is, um, kind of a way that we've allowed, um, we kind of democratize kind of like what, what it means to be doing GI ops and, um, the four principles if, uh, for those of you who kind of just wanna go more in depth to 'em, I'll, I'll kind of talk about 'em at a high level is on open GI ops dev, but there's really four guiding principles, right?

So, um, we, when we say GI ops, we, we say everything must be declarative, right? Um, have it in a system managed by GI Ops needs to have is state expressed declaratively, right? So kind of taking advantage of, you know, the declarative nature of Kubernetes, um, it needs to be, the second one is needs to be version and immutable, right?

Those desired states, right? Those declaration needs to be stored in a way that enforces immutability and versioning and retains an entire history, uh, of the changes. Notice that we didn't say get there, right?

We didn't, we didn't want to just kind of only say you have to be using GI to be doing GI ups. It's more of, okay, like you can be using an object storage, um, OCI, um, anything that preserves version and immutability, so that way you could do rollbacks and roll forwards and things like that. Um, another big thing that kind of gets, um, uh, uh, that kind of gets people with, uh, with the get ops principles is that it's pulled automatically, right?

Um, so it's, it's, it differentiates itself from like event based workflows, right? So you have an event based workflows where like something happens that triggers something. Um, GI ops isn't actually focused on event based, right?

It's, it's basically a reconciliation loop, right? It's a loop that happens, right? Um, that's not to say you won't have, uh, uh, triggers or events in your workflows.

You'll obviously will. I mean, there, there's, uh, it, I consider it a best practice, but it needs to work in tandem with a reconciliation, um, of, of the, of the system constantly, uh, reconciled, right? So, so it's pulled automatically, meaning that it, it just grabs the declaration, um, automatically.

So for, for most people, um, it's basically a get clone, right? It's pulled automatically, the system just pulls the declaration. Um, it doesn't wait for an event to actually be sent.

Uh, this information, it, it pulls them. Um, and it's always, which brings, kind of goes in hand to hand with the number four principle is continuously reconciled, right? So, um, it's, uh, it, it doesn't only pull the declarations, right?

But it continuously monitors the system, um, to make sure that the desired state is actually what is the current state. I always say that, um, you know, it's kind of like the difference between like a wall-mounted air conditioner and a thermostat, right? With a thermostat.

Um, you know, you just set your declaration, right? I want to 72 degrees or you know, 20 if you're in, in Europe, um, and, uh, and you, you just set it and the air conditioner will come on and come off trying to keep your desired state, right? With the current state, right?

With the current weather. Whereas a more event-based driven, it's like, oh, you know what? I'm a little warm.

I'm gonna stand up and turn on the air. You know what? I'm cold now.

I'm gonna stand up and turn off the air, right? And that's to kind of differentiate, um, uh, the, the kind of the two, um, work event-based workflows versus kind of like a, a, a reconciliation loop. And that is, um, and you'll always use them in tandem, but that, that's really, it's a nuanced, uh, difference.

And the dec and in all of these principles, we are not, we're not specifying a tool at all, um, or a specific protocol or anything, right? We're kind of just generally just focusing on the practice and principles. Uh, I just wanted to mention that, uh, part of that statement, and the, and the thing that is, is strongest in it is, is about monitoring and understanding state and the tools that we put in place to monitor and understand state, and you mentioned this very early on, Alan, like Spinnaker predates this model of, of GI ops, but it, it is always been a system that looked at the state of what's going on, uh, in the classes that you set up, in terms that you can take actions and it would monitor and then take those actions.

So that's a natural progression for something as, you know, as, as far back as spinnaker, that concept has been there. And strengthening that around the tools in the CDF that's present in Tecton, that's present even in our thought processes about orus, how things work. You, you have to monitor some source of, source of truth, understand that the system should also have that state that's been declared, and that's what makes it so extendable for tools.

Um, and that's part of what makes this marriage between GI ops and the CDF so important and, and so useful, uh, as we move into concepts like interoperability going forward and CD events, we're really trying to understand how do you communicate these states across systems that might not ordinarily, um, work together, right? We want those to be interoperable so that as you declare state, um, you can make those changes across multiple systems and also across however your architecture is spread out. Um, we're talking about Kubernetes, but there's, Kubernetes is not the only cloud system available that you're gonna make and monitor with, right?

And so all of these things have to be present in the toolings and the CDF, and we'll continue to build tools that do that. Excellent. I I, but I think it also speaks to the, the constant date of change that we see in, in tech tools.

And, and a lot of that is forested because it's open source, right? Open source. It, it's kind of in, in some ways, open source is like water or fluid, it finds its level, right?

You, you, you wouldn't necessarily see this in, in, in close source software or commercial software. It not that it's not adaptable and changeable, but generally the cycles take longer, right? And, and, you know, I think that's one of the beauties of the whole CDF and CNCF, the whole Linux Foundation open source system is, is it's ability to morph and evolve into its container.

No, no unintended with containers here, um, but to morph into its container now. So that begs the question, where do we go from here? Right?

What, what's next for GI ups, right? What are we seeing, you know, on the horizon? Well, you, you know, I have an opinion, but I, I, I'll let, I'll let, I'll let others chime in first.

Natali, how about you? You, what? You know, what are the new recipes in the cookbook?

Hey, Alan. Uh, that's a very good question, and I like you pointed out to the open source as a great motion for that. And also I would like to mention open standards, you know, uh, technology and open source, and comes together with the open standards and the industry just converge on those.

Uh, standard definition. So there's, uh, the open source name is an implementation, right? Is an implementation detail.

Then there's the open standards and uh, things like gear ops and also other technology. They have some kind of manifesto or working group manifesto, declaration where you define the standard, and then there's an open source implementation. That's the beauty, that's the strength of the community and the open source, uh, movement and gear ops is a, is a great part of it.

Look at open gear ops that, uh, Christian was mentioning before that that was a joint effort from multiple companies, multiple community defining the standard. Then there, there could be other, uh, open source implementation, Argo CD flux, spinnaker, you know, uh, anyone can implement in, in its own way, but the standard is open. The principle are open.

That's the important thing. And the second point, Alan, to your point, where are we going? Hey, you heard about all this gene AI thing, right?

Um, we're going more Once or twice. Yeah. Yeah.

We're, We're going to GI UPS for AI workflow, uh, ai ml, this means ops implemented with ops. So not only serving the app, deploying the app, syncing the app, also building, serving and think models consumed by the, the apps we're going in this, this, this direction. And the, the cool thing is that there's also an open source ai, uh, standard.

There are open source tools implementation. So it's not only chat GPT, there's, uh, lots of open source thing going on Austin Linux Foundation. And, uh, that's the beauty of, uh, ecosystem, which is able to interconnect that.

This was mentioning interoperability. This is possible, again, thanks to open standards. Yeah.

Alan, if I may, um, uh, I love the question about the what next. And of course, um, a hundred percent with Natalia general is, is certainly on, on the list and already here. Um, something that is totally, I would say incremental with giros is the reconciliation loop.

And when we are talking about reconciliation loop, most of the time we are thinking about, I would say infrastructure artifact or software artifact. But there is much more than this. We are security policy, network policy, sustainability.

And something that is, for me, super interesting, for instance, in what we are doing in telco industry on our, uh, Kubernetes cluster is starting to use, um, for instance, project like Kepler done by, I would say it had been initiated by team. Very, very good project that help us basically to have probs, uh, at namespace level that help us to understand how many power are used for a dedicated network function. So basically it's about the infrastructure, the workload, but we have to think about this drift management and this, uh, I would say continuous reconci, this continuous reconciliation about the network policy, the security policy, and as well, uh, the sustainability topic.

That for me is super, super critical. Yeah. And if I could extend that thought, the, from an enterprise level, I think one of the things that, uh, that GI ops is going to provide long term is this ability to have temporary environments and femoral environments that can speak across different organizations, um, that in very large companies can bring together just for testing, just for developing a really, really solid, uh, implementation of a complete system.

Because all of it's declarative, right? And you can say, Hey, I need the entire system set up to run this test at midnight, right? But you can do it between different organizations, and it's the coordination of the security of the networking that's in GI ops that will make that available to everyone else.

The the trick is, of course, to have the tooling around it that supports it and also understands that these things are gonna be fluid fair. Yeah. What I, what I like that both AI and Kepler were, were mentioned, because that's kind of like where my thought was going, but just kind of generically speaking about having these different, um, you know, modes of operating, right?

We have ML ops, right? We have, um, um, we have, uh, you know, environment sustainability. We have, uh, fin, uh, fin, uh, was that FinTech ops, right?

Like at China, the, and really like the get ops, I always say like, it's like the cornerstone, like of these other tooling that, that is coming around because like get ops kind of gets, gets you there, but these other things are like intent, right? Is, and I'm, I'm really looking forward to things like, um, you know, where AI can be coupled with like, the intent of some of these things, right? Like the intent of like project like Kepler, where it is like, Hey, let's use the equipment that we have efficiently.

Um, which goes hand in hand with FinTech ops, right? Where it's like, let's, let's just this be like the most cost optimal way and folks don't really have to like, worry about like, how it gets done. Because again, like I said, kind of GI ops is kind like the cornerstone of, of all of that.

And you really don't really have to like, think about it too much. And now you can just like focus more on intent and things. Um, you know, again, things like, uh, the Project Kepler is, is like probably one of the coolest things I think coupled with something like ai, um, is gonna be, uh, uh, and I, I think we're gonna see that's kind of like where I see GI Ops going in, in the future, being kind of like more in the background, but being more of an enabler to more of some of these things that have a great intent about, uh, using compute more efficiently, right?

However you define efficiently, right? Whether it's monetary or whether it's for the environment, um, or whether it's for like speed, right? Like, it, it can, it can mean a lot of things.

So I think that's it. It's really cool that I hear like all of these things mentioned kind of, kind of makes me think that I'm not too crazy in my, in my thinking that we're all kind of thinking along the same lines. Well, and you know, I just wanna top this off with the one thing that, um, y'all didn't, uh, mention, but you alluded to.

It's all of the projects, the, uh, addition of AI and how these things all work together, but it's the community behind everything, right? It's the fact that there is this human, uh, human oversight, this human intent to make things better, to make things more efficient, and to work in the open source format. So it's, like you said earlier, multiple companies, uh, people at home, people that are just passionate.

And I think if the audience, Washington can tell you all are very passionate about this, which is great, but it's that, it's the fact that everyone wants to move things forward. And, you know, there's all these different things that can get you there. And there's just communities beyond, communities beyond communities that are working in that direction.

So whether you're working on a project or, um, you know, you're a maintainer, you're a contributor or you're an end user and you're using things, um, everybody has a say in the game when it comes to open source. And having these sort of, to your point Christian cornerstones on how you can build on top of things, knowing that there again is human oversight. Making sure that those things are proved, tested, uh, ratified, you know, constantly reworked, scaled up.

Like you said, Alan flown, like water weaving Bob and weaving around things, I think is really, is really cute, huge. And, um, it makes me excited to be a part of the open source community. And, you know, I think you all are just great, uh, examples and representatives of what can happen when things like move forward.

Excellent. Let me, let me dive in on this, uh, AI and generative ai. 'cause three of four of you mentioned it already, right?

Um, look, everybody and their mother's doing a copilot, it seems, right? We, we all, there's a copilot for everything. You know, when we look at a copilot for GI ops, frankly, maybe I'm, you know, I'm, I don't hold myself out to be the be all, end all.

Is there a sort of co-pilot for GI ops, maybe on the horizon for open GI ops or, or any of the other, uh, tools out here? Uh, Christian, if, if I may, I guess that, go ahead. Uh, there are people, uh, there are people from, from, we've worked the company that, that was, um, I would say, uh, mainly back, uh, to the Ity project.

And, uh, they are, uh, they, they build a dedicated, uh, I would say, uh, model where you can, I would say, have a gene AI experience, uh, about your, uh, GitHub's uh, implementation. So perhaps if we, if you in the audience are curious, uh, it's quite easy to, to find it about, uh, yes, the Gene AI for, uh, flex City and Networks, um, from, uh, very amazingly, uh, and there are some demonstration that are live, I would say, uh, it's, it's quite easy to find it. So I'm, I'm not sure that it's a hundred percent, uh, production grade and already, but it's quite interesting in what's covering already.

Interesting. Anybody else to have any news on sort of a copilot for GI ops? Yeah, it's, it's very interesting.

We're I think, very early on to see what, um, what the capabilities is going to be. Um, I know at Acuity we have an AI assistant, which is, uh, which basically is a debugger, right? Like you can ask it like, you know, what your issues are and it'll come back and, you know, plain English will tell you what, you know, what the issue is.

Um, you know, based on like, you know, whatever, you know, questions that, that, that you give it, right? And it's more like a self-service, kind of like have, give the end users the ability to kind of do some debugging, um, in IT tool. But I think we barely scratched the surface on, you know, what is capable, um, with, with GI ops and ai.

I, I think, I think we're, we're still, we're we're, we're still in the beginning stages. Uh, today is, uh, it's not difficult, but it's a task to git sify an app. You, you write an application, then you need to write the manifest describing this application.

How about Gene ai or a code assistant will keep git sify your app. Exactly. So all the manifest out there and everything is there.

I think it's, we're not far from there. And Christian and, uh, Philip were mentioning real use case, uh, happening today, but this other use cases, it, it, I don't think it's too much far in the, in the future. So do you envision that as a separate open source project or is that something that maybe gets opened, you know, not opened, gets joined into the open GI UPS project or something like that?

So is it standalone or, or part of, part of sort of an existing, In this case that Christian, uh, and Philip mentioned it, that they start from a model. There's a chatbot that consuming this model, this modeling, there is inference text. So, um, I I, I, I haven't heard anything in that regard, but it would not too much difficult to implement.

Well, there, they, someone out there should grab on that run, run with that idea then. Right? Exactly.

Um, I, I will tell you as you know, one of the greatest things that I like using generative AI chat bots for is I'm not a coder, right? Uh, I'm not a coder and I, but I learned how code works. But, you know, by putting that Christian, much, much as you mentioned, debugging, I don't use it for debugging.

I use it for explaining, right, what does this mean? And in very plain English, it tells you that's what this is, that's what this does. This is why this is there.

You know, what a great tool for people who wanna learn how to use GI ops or how to use, you know, so many tools today. I think when we talk about generative AI and taking developer jobs and stuff like that, no, I think it's gonna make us more developers and better developers, right? Because think about how much you, you can learn doing that.

Now, is that something that CDCF can get behind to try to teach people more about CDC, uh, ci, excuse me. CICD. Thank you for that segue.

That was awesome, by the way. Um, I, I, I, it's, it's interesting that you mentioned this. 'cause one of the concepts that, um, I've been talking about or pushing for in our own community is a better understanding of the different types of people or the different types of use cases that the community members bring in.

And so, going back to what Lori was talking about earlier, this is built on community and each one of those members are different and their use cases are different, and their needs are different and their sides are different. And in order to build the model, you have to understand the problem in a way that you can highlight all of the information and then find the commonalities, right? And so, because that's the process, that is the process that we're going to be going through in the CDF, like really understanding those use cases, helping, um, teams find, for lack of a better word, uh, a golden path.

If you're just starting out and you know the size of, of your company and you know, the use case you kind of want to hit, like what, not just what is the best practice for the tooling that you put together, but what's the best practice for your size and your need? And can we begin to model that in a way that fits those use cases? And once we understand that in a way that's useful, we can of course then ha you know, work with the GI ops community to say, okay, we have all of these use cases.

Let's work on a model that kind of starts to pre prescribe these things. And then, uh, that should, should eventually arise. Yeah, absolutely.

I, I wanted to mention Philippe put in the chat the link to, uh, project spoke about, we'll, Philippe, we'll try to get that into the notes 'cause I'm not sure if the chat window shows up on the video screen, but, but thank you for putting that in there guys. We we're coming up to the top of the hour here, and we, we need to wrap up. Um, Lori, I'm gonna come to you last so you could start thinking about what you want to say right now.

But I wanted to give each of you kind of one last chance to leave something with our audience, right, about gi ups, about CDF. Maybe you want to mention if something's happening in Paris at, uh, CubeCon, uh, cloud native con right in another month or so from today, a little more than a month from today. Um, Natalia, I think we started with you in the beginning of the show, so let's start with you now, if it's okay.

Yeah, thanks. Uh, yeah, first of all, in CubeCon, uh, Europe, in Paris, it is gonna be the end of March. Uh, and there're gonna be also some co-locate event, uh, for Tuesday, for Ops Tuesday, uh, Thursday, right?

And, uh, there's also Argo Con, there are lots of those, uh, interesting gather. So a anyone has the chance to come over, please do, because it's a very good moment, moment to gather and create community also around ops. And my, my last thoughts about it, I think ops is a methodology like the, it is like DevOps.

It's a methodology that you need to embrace and you need to set up some process, some culture. Uh, it's an evolution of DevOps, of infrastructure as a code, whatever, but it's a methodology. Uh, I think Philippe was mentioning this before.

Once you adapt this methodology, then all the, the process comes along together, the technology comes along together, and you can bring this to infrastructure set up, user set up app set up and models, you know, gene AI as we were discussing. So it can be, uh, really some something universal to adopt. Uh, but you need this first step of cultural switch into GI ops and open ops I think is the really good place to start with.

Excellent. Thank you. Christian, how about you?

Well, as, um, as Natalia mentioned that, um, I will be in Paris at, at, um, at Argo Con, right? It's a co-located on, um, events, sorry, Tuesday Q Con. It's on Tuesday.

Yeah. Um, right before, right before Q con. So, um, I'll be there right as, as a co-chair of the events.

Um, I'll be there. So you can you catch me on, on there. And, um, really, really what I like to say about, about GI Ops and, you know, open GI ops and the GI ops working group is, um, if you wanna learn more, just really just get involved, right?

Get involved. There's always, um, there's always in, in open source. We, there's always a need for help, right?

And don't think that, um, you know, any contribution is a small contribution. dev, um, there, there, there's a link somewhere to, uh, get involved, uh, join a meeting, um, you know, and talk, get ups and ask questions and, and things like that. This is, uh, again, a conversation that I don't want just to end here.

Is, is one that I would like to extend to other folks, um, uh, to go to. Yeah. So, um, contribute, open, get ops, come to a meeting, you'll be in Paris.

CubeCon, stop by the Argo Con. I'll be there. Uh, probably jet lagged, but I will be there.

So I, before we move on, I actually mentioned that Techstrong will be there, tech Strong tv. We will be, um, streaming live all week. Uh, we're actually finalizing something right now with the CNCF Linux Foundation, where I think we're gonna be interviewing a lot of maintainers and project people.

So Christian, maybe we'll have you on there. That would be great. Um, working with some, some of the CNCF folks, but also on Tuesday, which is the, I don't think they're calling it zero day this year.

And as a security person, that does my heart good. I never like the term zero day. It has evil connotations, but, um, you know, pre, pre cube con day, let's call it, is Tuesday, and I believe it is the 19th.

And we're actually sponsoring a breakfast from 7 45 in the morning at a rooftop cafe right across the street from, uh, the conference center. Um, it's open to anyone who just goes on Eventbrite and registers. It's a good breakfast and it's a great way to start off a day full of, you know, satellite conferences that are going around here.

And, um, invite anyone watching this. Uh, we'll, we'll have in the notes, I'll put the link to the Eventbrite. Um, but let's move over to Philippe now.

Philippe, your last thoughts? Yeah, thanks so much. So of course, um, uh, we'll be in Paris, so it could be at home for me.

So if you want to understand better how basically Cloudnative and GitHubs could be applied in, in, in complex, uh, tele ecosystem, don't, don't hesitate to reach out. Give me my pleasure to talk and chat with you. Uh, my last thought would be very, very simple to just tell how much up is incremental, uh, for, uh, telco industry.

It's super important to recall that. Um, we, uh, and several, um, other, uh, telecom operator peers, we have been strongly imply in two incremental documents. The first one is what is called the Cloud Native Manifesto, released by the NGMN, where we identify seven core principle, how to make telco cloud native.

And, uh, the principle number four is only dedicated to GitHubs. Second, uh, super, um, important document is, um, a white paper that, uh, we released inside the C-N-C-F-C-N-F certification working group that now is hosted by Linex Foundation Networking, that is called Challenges of Cloud Native Telco transformation today and how to overcome them tomorrow. And there is the total dedicated part about gito.

So just to say that in telco industry, we truly believe that GitHubs is the operating model and framework that we need to build a resilient, sustainable, and flexible network. Excellent. Let's see if I got it.

I do see, am I close? The DC the dc DC the DC That was my best one yet though. DC I'm getting there.

Uh, it takes me 10 times the DC How about you? Well, I, I, hearing those wonderful statements as exit statements, I will just say that all of those things need the tools that do the monitoring, uh, and the operations of those concepts. And, um, if you are, if those tools are important to you, if interoperability of those tools are important to you, please take some time, come over, visit the CDF or the CD Foundation, uh, on, on the internet, and join a meeting, join the CD of insig, join the interoperability sig, join the conversation of how we tie these tools together, um, so that you can have monitoring across, uh, different levels.

How do we impact security in your pipeline? How are we, you know, looking at all these pieces and fitting them together and making them actionable for you, as well as packaging that software that goes through, uh, the GI ops process, right? All of those steps are really, really important and we are going to safeguard those for you.

So come join us. Fantastic. Thank you, Lori.

Last word. Thanks for giving me the last word. Um, so I think if, if anyone watching has gotten anything out of this today is that there is a huge community behind GI ops, behind, you know, working together collaboratively.

Uh, whether you're at the CDF or you're at the CNCF or you're working on an individual project or you're working in telco or you're working in a different organization altogether, and without the humans behind all of these things doing the work, we're never gonna move forward. So I think everyone sort of echoed the sentiment that there's an opportunity for you to get involved. And if you're not going to keep con, if perhaps you're going to open source Summit North America, which is where I'll be, we have CD Con there, and that's, um, the week of April 16th, and there's a in Seattle Conference day as well, lots of activities happening.

But I think what I would really love for you all to take out of this episode is that there's frameworks, there's methodologies, there's tools, and the only way you're really gonna get to see what's going on is to make yourself a seat at the table and to join a community. Whatever community, uh, is your passion, there's always room for you. So we'd love to have you at the CDFI know Christian would love to have you, uh, at Open GI Ops and all of the other great groups that are going on.

But once you get involved, I think the other thing that we are showing today is the power of collaboration, right? So the CNCF, working with the CDF, working with Open SSF, working with other organizations across the board to just make sure that we're moving things right, we're stewarding technology in a way that everyone can be on the same page, you know, level playing field and just build on top. foundation.

And Alan as always, thank you for giving us this opportunity to share a little bit more with the world. My pleasure. All right, that's gonna wrap this episode of CD Pipeline.

I don't know, Laura, if we have another one planned before CubeCon. Uh, but six weeks probably not. So we'll, we'll pick this up after CubeCon Cloud Native Con and that'll probably mean we'll be in late March, early April.

But until then, I think thank you all for excellent advice and excellent guidance to our audience. Keep guy, keep doing what you're doing. You know, there's many unsung heroes in the open source space and, and you guys are four of them, five of them.

Thank you so much. And that's gonna make it a wrap on today's show. Take care, everyone.