KubeCon + CloudNativeCon EU 2026 | Andrew Burden & Ľuboslav Pivarč, Red Hat

Clips
More from KubeCon + CloudNativeCon EU 2026

Andrew Burden

Community Facilitator

Red Hat

Ľuboslav Pivarč

Principal Software Engineer

Red Hat

play_circle_outline Kubert v1.8 Debuts at KubeCon Tomorrow: VMs with Containers, Nested Multi-Cluster, NVIDIA vGPU for AI

play_circle_outline Increasing vendor collaboration: Intel, AMD, IBM, NVIDIA contributing features

play_circle_outline Upstream-first development model and relation to OpenShift Virtualization downstream

play_circle_outline Kubert advances in CNCF graduation process; adopters list expands and end users urged to register

Info
Transcript

Andrew Burden & Ľuboslav Pivarč, Red Hat

Andrew Burden

Community Facilitator Red Hat

Ľuboslav Pivarč

Principal Software Engineer Red Hat

In this interview from KubeCon + CloudNativeCon EU 2026, Andrew Burden, community facilitator at Red Hat, joins Ľuboslav Pivarč, principal software engineer at Red Hat, to talk with theCUBE Research's Rob Strechay and Rebecca Knight about KubeVirt's journey toward becoming a first-class citizen of the Kubernetes ecosystem. With version 1.8 launching at the event, Burden explains how KubeVirt containerizes the virtualization layer, enabling legacy virtual machines to run natively alongside container workloads with stronger security and isolation. Pivarč highli... Read more

explore Keep Exploring

What is Kubert, and what are its main goals and use cases? add

How are industry partners and chip vendors collaborating to enable AI and confidential computing features in Kubernetes? add

How does the upstream‑downstream relationship between Kubert and OpenShift Virtualization affect your work and the release/process coordination? add

Is Kubert nearing graduation from the CNCF, and what is the current status of that graduation process? add

bolt Powered by CUBE AI

Andrew Burden & Ľuboslav Pivarč, Red Hat

search

Rebecca Knight

>> Hello everyone and welcome back to the Kube's live coverage of KubeCon CloudNativeCon 2026 here live from Amsterdam. I'm your host, Rebecca Knight and I am alongside Rob Stretchay, the terrific Rob Stretchay. Let's talk about Kubert and making it a first class citizen across the Kubernetes ecosystem.

Rob Strechay

>> I mean, to me, it's one of my favorite projects. I think again, it's alignment as we go forward where some things fit better in virtual machines than they do in containers. And containers came from a different place and getting to that where they're all the same and treated the same within that control plane is why I love this project so much.

Rebecca Knight

>> Well, who better to talk about it than Andrew Burden, community facilitator at Kubert Red Hat. Thank you so much for coming on the show.

Andrew Burden

>> Thank you very much for having us here.

Rebecca Knight

>> And you are joined by Ľuboslav Pivarč, principal software engineer, Kubert maintainer at Red Hat. Thank you both so much.

Ľuboslav Pivarč

>> Hi, thank you for having me.

Rebecca Knight

>> And I want to point out to our audience that you are both wearing, you have a lot of school spirit right now, wearing the uniform of Kubert. Excellent. Very cute little squid there.

Andrew Burden

>> Yeah, thank you very much. And it's a very special show because it's for our event that's happening this Thursday at the Cloud Native Theater. It's the first Kubert Summit live event that we've ever run. We're really stoked. We've got five great sessions. It's from 11:00 AM to 1:30 PM here at KubeCon.

Rebecca Knight

>> Okay. All right.

Rob Strechay

>> Live in Amsterdam. But for those who are not here, they're going to also have it on and tape it and put it on-

Andrew Burden

>> They're going to tape it, yeah, as I understand it. Yeah.

Rob Strechay

>> Of course.

Rebecca Knight

>> So Andrew, I'm going to start with you because Kubert is, it's turning into quite a milestone moment. You're dropping version 1.8 this here at KubeCon-

Rob Strechay

>> Tomorrow?

Rebecca Knight

>> Tomorrow.

Andrew Burden

>> Tomorrow.

Ľuboslav Pivarč

>> Yeah, tomorrow.

Rebecca Knight

>> We wait with bated breath. For those who are just getting familiar with the project, I know Rob did a little bit of a preview there, but tell our viewers a little bit about what Kubert is all about.

Andrew Burden

>> What Kubert is all about. Well, so it's about containerizationing the virtualization layer. It originally started as an idea of being able to support legacy virtual machines natively in Kubernetes alongside container workloads. And then as the project has matured and grown, we've come across a few other workflows. One is the increased security and isolation afforded by virtual machines over containers. There is VGPU enablement, which underpins things like Nvidia's GeForce Now. They can be used in various pipelines like Tekton pipelines, generating artifacts, what have you. And also then with the virtualization as a base layer for multi-clusters, nested virtualized Kubernetes clusters in a Kubernetes cluster. So those are the primary.

Rob Strechay

>> So how are you seeing the community come together as a maintainer and how are you seeing it grow as it's gone from before Salt Lake or before Atlanta to Atlanta and now in the last five months or so since we were in Atlanta? How has it grown? How has the momentum gone?

Ľuboslav Pivarč

>> The momentum is very great. We see a lot of collaboration with a lot of companies. We see ideas and features that we need to develop together. We see AI driving a lot of this, for example, from the hardware perspective where we need to enable it and Nvidia is coming to us, helping us with it. But also we have, for example, confidential computing, which we are trying to bring to Kubert and there you need all the vendors from the chips manufacturer. So Intel, AMD, for example, or from IBM. And they are all coming to us and helping us to do this in the Kubert, enabling these features, which is great to see and great to see that you are not alone here and that you get the help that you need sometimes. So yeah, I think it's coming well.

Rebecca Knight

>> So Andrew, I know that Microsoft has stepped in to lead a pretty major feature. And for a project that's been Red Hat heavy for most of its life, this seems like a pretty important point. What do you think it was that made Microsoft want to move from a quiet user to more of an active contributor here?

Andrew Burden

>> That's difficult to answer, but I like to think that kind of like underlying that is the fact that we've created and it's been in three no versions now have instigated a vet process, which is a lighter version of the Ket process for those more familiar with the Kubernetes side of things. And those kind of like the templates, the process and the timelines, making that transparent and providing a place for people to come and engage with us and work with the community, to enable an end user. We didn't know Microsoft was an end user until they came with this like really exciting, complicated idea that we, initially we were kind of like, "That's kind of wild." But then through negotiation, through conversation and just engagement, they were able to finally bring this thing where we came together, came up with something that made sense for all the parties. And so I just think, yeah, having that process in place, that transparency, I think, makes it possible for people to better understand how they can come with these ideas and just engage with the community. Yeah. And we greatly encourage that. I need to stress.

Rob Strechay

>> Oh, absolutely. I mean, I think we're all about community here and I think it's great to see them coming to the table with complicated and useful ideas across that. To that point, I mean, again, to me, it seems like there's ... I mean, Kubernetes, as Jonathan was talking about on stage this morning, as well is very ... It's ubiquitous across clouds. And when you start to look at it, people are looking how to be more efficient. AI, not all of AI fits solely well within containers. And sometimes you need a VM or some people like to put databases in VMs and they don't want to go and change. What are some of the use cases that you're seeing that are coming into Kubert and really helping you challenge how you have to design and build the underlying as one of the maintainers and how the core folks are actually operating in there?

Ľuboslav Pivarč

>> Yeah. I mean, the multi-hypervisor work we are doing with Microsoft is great challenge. It's the next steps for Kubert. And I think, and I believe it's going to be a great thing for us because we need to evolve a lot of technical stuff behind the scene that probably user will not see. But when we get this done, they get a lot of benefits out of it, for example, for Microsoft, but it's not the only company that can join us. We are open to any other company or any other project that wants to integrate with us with different hypervisors. So there is a lot of value that we are going to unlock there, I believe.

Rob Strechay

>> So just playing off that a little bit, again, obviously you're also with Red Hat and there's the downstream versus the upstream and Kubert's really upstream. How does that play out with what you're doing in OpenShift Virtualization versus what's in Kubert?

Andrew Burden

>> So I'm not heavily involved in the OpenShift side of things. I'm very much an upstream first. And that's one of the great things about working on Kubert and being paid by Red Hat to work upstream first. It's kind of like a little bit of a dream in many ways, props and respect. I think our release cadence, and again, this kind of like having these set timelines and the vet processes are part of this of making it available not just for OpenShift virtualization, but any of the downstream products that come out of Kubert to understand that to have something in their downstream, they need to meet the vet process, the vet deadline, the code freeze, and have this kind of like the same way that we have to operate the same way with Kubernetes. It's just kind of like, I think of it as a little bit of a waterfall. And I think again, yeah, just understanding our timelines and our processes and we're trying to get better at communicating that and making sure that, again, all open and transparent, we meet those, then it will be there to be piped into the downstream.

Rebecca Knight

>> Luboslav, can you describe a little bit about what that process looks like? So I'm talking about the Kubert enhancement proposals process, which is a bit of a mouthful, where community members are surfacing and helping make improvements over time. How does a healthy process look like in practice and how do you make it not become this bureaucratic bottleneck?

Ľuboslav Pivarč

>> Yeah, that's a very good question. So first of all, it needs to be open, right? It needs to be inclusive to everyone for everybody's idea and everybody features. It happens in a way that you basically, it's very natural in my opinion to everybody who is contributing to, for example, on GitHub or GitLab. They open an issue with, "Hey, I want this kind of feature getting into the Kubert." They describe what they want it to look like, but also they need to provide a motivation behind it for us to kind of see if this is the right feed or we can think about something else. And then because it's open, everybody can come to that issue and comment, "Oh yeah, this is great idea." Or maybe "No," or maybe, "I have another idea that we can implement there." And when they open the issue, they can open a PR, which is described in the design. And I think, I hope this is not an overhead because everybody should think about how you structure the architecture of the system. So I don't think that's overhead, but sometimes I do hear that this is overhead, unfortunately. And when you do the design, we can iterate on it with everybody in the community. Once we get that settled and we say, "Okay, we are fine with it," we need to also get a perspective of who has the time to actually shepherd this from the community further. And when we do this, we can merge the design and we settle that, "Okay, so next release is 1-9," we get this feature in the 1-9 into AlphaStage. We trying to get this in stages so we can work on a proper system which can get a feedback look as well from the users to iterate on it. And when it is GA, we want it to be there forever and don't change it. So we try to provide the stability that every enterprise is asking for, but as well, we're trying to get the speed there for this great fast-moving environment.

Rob Strechay

>> So I was going to play off of that because I think things are moving so fast in this community. I mean, just everybody's talking about the AI and we went from models to inference and agents and all of this and you have new contributors who are coming in, like you have Nvidia who's coming in as well. How does that help you or how does that challenge you with prioritization of ... Because I mean, there's only so much work you can get done in a certain timeframe. How does that all work as well? Either of you can answer.

Andrew Burden

>> That's more of a you question.

Rob Strechay

>> Yeah.

Ľuboslav Pivarč

>> That's a great question. I would say what's the most impactful thing and that we need to pick up, right? Because we cannot just work on something that, for example, Red Hat is asking me to work. We need to see that we get to all community needs. And when we don't have at the time, we ask somebody else to step up. We have Nvidia with us for quite some time, I would say. And they are a great contributor, for example, contributing to the six scalability, which we have trying to improve scalability of Kubert. So whenever we need help, we ask for help. And if ... I think we never just denied a vet until today.

Andrew Burden

>> I know.

Ľuboslav Pivarč

>> Yeah. So we still find the capacity, but we are aware that we don't have unlimited capacity in the world. And for example, the next release, we are seeing 70 deps. So 70 designs coming into Kubert. That's for sure, this is going to be like the breaking point where we are going to be more of prioritizing what to pick. And to be honest, it's going to be fun.

Rob Strechay

>> So when do you see your timeframe for your next release that ... So you can prioritize and having done that, understanding backlog management and all of that, that's what I did prior to sitting here talking all the time. But do you have quarterly releases? Do you have X number of sprints and things like that? How do you actually organize that? Curious.

Ľuboslav Pivarč

>> Well, we have this thing in Kube Eraser. We don't try to reinvent all the things. We are trying to follow Kubernetes because we are running on Kubernetes, right? We want that filled as a Kubernetes and we'll belong to a community to fill as Kubernetes as well. So we are following Kubernetes releases. We are a little bit behind because we of course need the Kubernetes release in order to develop our features on top of it. I think we are behind one month maybe.

Andrew Burden

>> I think it's more than that.

Ľuboslav Pivarč

>> Maybe two?

Andrew Burden

>> 12 weeks?

Ľuboslav Pivarč

>> Yeah, 12 weeks.

Andrew Burden

>> It's a few weeks.

Ľuboslav Pivarč

>> Few weeks.

Andrew Burden

>> Yeah.

Ľuboslav Pivarč

>> Yeah. That's it.

Andrew Burden

>> Yeah, I think so.

Rebecca Knight

>> So Andrew, I want to bring you back into this because at the very beginning of our conversation, you talked about the Kubert Summit, which is happening this week for the first time after five years of running it virtually. What made you want to bring together people in person for the first time? And what kinds of conversations do you expect to be having tomorrow?

Andrew Burden

>> The first part is super easy. It's like magic happens when you bring people together. And I think we learned through the pandemic that it's really nice and it is a balancing act because our virtual events enable anyone from around the world to attend. It's a lot more complicated when it's in person event, but there is something truly amazing that happens when you do meet in person, and that's what this event's all about. And so yeah, it's something I've been investigating and trying to put together for a couple of years now and the CNCF made ... This is a new thing they're running. So we're in the trial phase of that. And so they gave us the opportunity and we jumped at it. The conversations, we've got some really smart people from Nvidia talking about Kubert and GB200. We've got an awesome live migration story from Hitachi. We've got SK Telecom and NetApp talking about the storage aspect of the South Korean AI digital sovereignty thing. And we've got people from the core community talking about what's happening there: bridging, multi-clustered with networking and how SIG scale and performance is like upscaling their performance testing. So I'm really excited about it. It's going to be awesome.

Rob Strechay

>> Easy for you to say, right? No, but I think one of the things, kind of parting thoughts, graduation, is it around the corner for Kubert? How are things going? Because I know that's a process too.

Andrew Burden

>> Huge process. We put our ticket in last year, August, September, and just recently we've moved from the first queue into the second queue. We've got someone assigned and we're beginning DD and this is ... Yeah, it's a big deal. It's a big deal. And I think Kubert, I think, has felt mature for quite a while and our adopters list kind of shows that. So if there are any end users out there who aren't on our adopters list, would love to see you on there, that's on Kubert, kubertadopters.md. But also ... I totally lost my train of thought there.

Rob Strechay

>> But you hope to get there and you've moved into the next queue.

Andrew Burden

>> And right. This graduating for us is kind of like, it makes it more obvious for people that iron is deeply embedded in the Kubernetes ecosystem, the CNCF ecosystem. And so it's kind of like they can see us graduate and be like, "That's obviously a mature product. Let's investigate that."

Rob Strechay

>> Awesome.

Andrew Burden

>> Thanks for bringing me back.

Rob Strechay

>> No problem. That's what I'm here for, man.

Rebecca Knight

>> Well, Andrew, thank you both so much for coming on The Kube. Really interesting conversation.

Andrew Burden

>> Thank you so much.

Rebecca Knight

>> Thank you. Best of luck on the summit. Stay tuned for more of the Kube's live coverage of KubeCon 2026, CloudNativeCon, live in Amsterdam. I'm your host, Rebecca Knight for Rob Strechay. You're watching The Kube, the leader in enterprise tech news and analysis.