Cloud Native

At KubeCon, Mirantis launches MKE 4; teases a major open source initiative

0

At KubeCon+CloudNativeCon in Salt Lake City, Utah, Mirantis announced Mirantis Kubernetes Engine (MKE) 4 for secure enterprise Kubernetes. Based on k0s, MKE 4 delivers ‘enterprise ready’ Kubernetes that is high-performance, highly-secure (with FIPS 140-2 encryption) and easy to operate, with a convenient web UI (user interface).

In this video, Shaun O’Meara, CTO at Mirantis, joined me to discuss the company’s focus on Kubernetes and the launch of MKE 4, the latest iteration of their Kubernetes distribution. MKE 4, which supports around 300,000 nodes and 1,000 customers, is designed to be more ‘composable’ and modernized compared to MKE 3. It includes a Blueprint operator for customizable deployments and is based on the k0s distribution.

Mirantis also highlighted other projects like k0smotron and Mirantis Secure Registry (MSR), transitioning to Harbor for better container image security. The MKE 4 release candidates are available, with general availability expected in early December.

O’Meara also teased an ambitious open source initiative, Project 2A, which has been a focal point for Mirantis since bringing on Randy Bias to lead their open source efforts.

Guest: Shaun O’Meara (LinkedIn)
Company: Mirantis
Show: Let’s Talk

Questions discussed

  • How would you define Mirantis in this setup?
  • What kind of discussions are you having at the event?
  • What is MKE and why did it come out?
  • How is Mirantis helping customers affected by the Broadcom-VMware deal?
  • When will MKE 4 be ready for production use?

[read more]

Unedited Transcript (Note: the text is AI generated, it has not been edited or reviewed. It may contain errors, including incorrect names. It’s provided here under Creative Commons license (CC by 4.0) to be used by bloggers, journalists and analysts for creating their own content.)

 

Swapnil Bhartiya: Hi. This is Swapnil Bhartiya. We are here at KubeCon+CloudNativeCon in Salt Lake City, Utah and we have with us once again Shaun O’Meara, CTO at Mirantis. Shaun, it’s great to have you back on the show.

Shaun O’Meara: Great to be here as always. Good to see you.

Swapnil Bhartiya : It’s my pleasure. Of course, we are going to talk about MKE 4, but before we talk about it I would just like to hear from you. Is that, how would you once again define Mirantis in this setup?

Shaun O’Meara: Mirantis is a cloud-native Kubernetes company. For us, Kubernetes is everything. The core of everything we do as a company is based on Kubernetes. For the last five years now, we’ve even run all of our open stack products on top of Kubernetes. So Kubernetes is the heart of Mirantis, and our focus is cloud native solutions for our customers, providing the right level of support for all of their needs.

Swapnil Bhartiya: This is day one. It’s almost wrapping up, which also means that you have a very good pulse of the audience. So talk a bit about what kind of discussion you’re having, what kind of folks are walking through, both what kind of conversation you’re having.

Shaun O’Meara: It’s a good question. I mean, I spend a lot of my day talking to both analysts and journalists around the event today, as well as a number of people. Lots of talk about AI. Of course, AI is the current hot topic for everybody. Lots of interest in the practicalities of running AI on top of infrastructure and how that collides. Lots of questions of whether we should be focusing so heavily on AI at this event, but I think AI is the future, and that’s where things are going. So yeah, this is the right focus for us today.

Swapnil Bhartiya: And now let’s talk about MKE for just let’s start with what is MKE and then why this came out, and then we’ll talk about the evolution,

Shaun O’Meara: sure. So let’s start with the question of what. What is MKE. MKE is the Ó distribution of Kubernetes. It’s, we consider it a platform of Kubernetes, but not a pass. It is part of the Docker heritage that we have earned back from Docker over the years. It’s, we have it running around 300,000 plus nodes, over several 1000 customers. It’s a complete Kubernetes platform. So you can take it, deploy it, and you have a working Kubernetes out of the gate with all the basics, including the Ingress, the RBAC, the cnis, the CSIS, and a full support model around that to guarantee production level, grade, working at scale, within our customers, majority of our customers today, we’ve got a very broad range, but a lot of financial services and lot of regulated environments. So that’s where M Ke is right now. As I mentioned, MKE is an iteration of the Docker enterprise that was based on a very particular architecture that Docker put together. We’re going back six or seven years now, and we’ve had a goal to modernize the M ke architecture over the last couple of years. So we’ve now with the new M ke four release. We’ve modernized that architecture, we’ve simplified that architecture, but really importantly, we’ve moved to a composable Kubernetes architecture. M ke three was a lot more monolithic in its architecture and design. M ke four has moved to a composable architecture based on k0s. So as you know, Mirantis delivered the k0s distribution for the first time in 2020. We’ve now gotten a lot of adoption of k0s, and we believe that k0s has reached that maturity point that it’s now ready for our enterprise customers. As a result, we’ve created something called the blueprint operator. The Blueprint operator allows us to compose those community platforms for our customers needs. The Blueprint operator allows customers to pick those services that they require to create that platform, as well as deploy and manage that. MKE look and feel, which is really, really important to our customer base. MKE 4 is an in place upgrade of MKE3. So really, from a customer point of view, from a user point of view, it’s just the next generation of the update that architecture change happens magically, no, completely abstracted from the customer underneath the hood, and then brings that composability into the environments for large scale Kubernetes clusters.

Swapnil Bhartiya:, when we look at Mirantis, as you mentioned, k0s, other so if you just get it, you know what other projects, products that are there to cater to this ecosystem and community,

Shaun O’Meara: for sure. I mean, k0s is at the core. So I mentioned k0s. It’s really k0s focused on delivering a really high quality Kubernetes without external dependencies, as a single binary for a wide range of use cases. But k0s is just Kubernetes. It has a lot of plugins and capabilities, but really it’s just. Kubernetes out of the box, related to k0s, we have the cosmetron project. Cosmetron is primarily a bootstrap provider for the cluster API or CAPI, but it also provides a hosted control plane mechanism that allows you to create Kubernetes control planes as pods, and then remotely manage workers as part of that. So you can do that either with CAPI or without CAPI. Other projects that we’ve got going at the moment, also in the same ecosystem, we’ve released the next edition of Mirantis secure registry. This is based on harbor, and we are contributing to the harbor community as part of that move, where we’re moving our existing customer base from MSR to harbor over time, making sure that all the capabilities that were built into MSR will also be in harbor as we move forward. Those are the key open source projects. We’ve got some other cool projects going on at the moment. I’m wearing the shirt for one of those projects. I don’t want to go into too much detail, but it’s a fully open source project focused on multi cluster management and hybrid cloud management of Kubernetes, but not just at the cluster layer. We’ve moved up into the service management, state management and Observability layer. It’s called Project two A. The two A is 0x two a, for those of you can remember your hex, it means 42 as we’re in a public forum. I won’t go into too much of what the 42 connection is, but it just proves the geeks at the Mirantis executive team, we’re very much focused on helping and producing a world where Kubernetes is the ubiquitous abstraction layer for all infrastructure and all applications, and that’s where we’re moving forward with project two. A at the moment.

Swapnil Bhartiya: When we look at the whole CNC of landscape or Kubernetes ecosystem, we have talked about some of the old pain points, which were like complexity, scale, cost, but we sometimes assume that everybody is a green horn. No, everybody is. But there are a lot of legacy players. There are a lot of traditional players who are looking at that market. But when they look at complexity, or all those things, how do you look at that player? To move them from those old legacy systems to modern Kubernetes systems?

Shaun O’Meara: I think the reality is so. So with MKE 4 as an example. Let’s, let’s look about, you know, talk about the real elephant in the room when we talk about legacy, everybody’s concerned about VMs. And obviously the broad, common VMware world has changed a lot of what we talk about these days. So as we start to look at that, we’ve included, for example, cube vert as part of our MKE. MKE is also designed to be very easy to deploy and manage. That UI for MKE will have a kube UI so that your traditional operator who’s used to working through UI, point, click, type UIs. And I don’t think they’re unsophisticated. I think those are very sophisticated operators, in many ways, they’re just used to a certain way of work, so we want to soften the transition for them by providing great UIs and great user experiences. The world is rapidly moving forward, though, and I think more and more people want to have Kubernetes in their environments. We want to make it easy to manage Kubernetes at scale. That’s where, really, where the project I spoke about earlier comes in. How do we manage Kubernetes a very, very large scale, but without everybody having to become an expert, but if they have experts, those experts still have access to the core of the system unchanged. You know, without any sort of lock in, so that they can continue to manage and create their own custom environments as they need to.

Swapnil Bhartiya : Since you mentioned cube word, did you feel any shock waves from the whole Broadcom VMware space here and how Mirantis is helping some of those customers,

Shaun O’Meara: we do feel the shock waves. We’ve had a viable alternative to the VMware product for a long time, in the form of Mirantis, OpenStack for Kubernetes. Really, where Kubert is coming in is for those customers who have smaller or medium sized VMware deployments who are looking for alternatives. They’re not necessarily replacing VMware. I mean, I still think VMware is a very solid product, but for a lot of reasons, they’re looking for parallel alternatives, and this is where something like Kube makes a lot of sense, because Kubernetes clusters are easy to manage. We’ve solved that problem, put Kubeon on top of it. Now you’ve got a very easy to manage VM environment, and you can do it very affordably. Mosque and open stack is aimed at much larger environments where people are trying to create private cloud alternatives to the VMware private cloud environments. And I hesitate to always say alternatives. It’s it’s parallel options within environments. So that’s really what we’re seeing. So we see a lot of customers. Who are deeply exploring ways to spread their workload into more baskets, rather than the old all the eggs in one basket, type model, which was the VMware space, right?

Swapnil Bhartiya: Thank you. Now I’ll go back to MKE in a bit. When you look at MKE 4, is it ready for production when people can start using it.

Shaun O’Meara: So MKE the release candidates are released. We’re announcing it here today. Today is our announcement. It’ll be the final GA available next week Wednesday. Check in with my team to make sure I haven’t got it off. It’s next Wednesday, and it will. It has already gone out to some early, early test customers, and we will start upgrade migrations with our customer base early December.

Swapnil Bhartiya: What kind of iteration Do you see of MKE going forward?

Shaun O’Meara: So MKE 4, because of its composable nature, it’s going to iterate along the same lines as k0s. So we’ve decoupled the Kubernetes iteration from the rest of the ecosystem, which, obviously there is a link, you can’t just upgrade something if you change the Kubernetes API, but customers have a lot more choice over which API. k0s is much closer to the head of Kubernetes, so customers will always have access to the latest and then we’re going to start to add more options, or more blueprint options for customers to start to compose those environments as they move forward. MKE itself is very focused, as I said, on those large scale customers. And our new project, which will be related with two with MKE parallel tracks, will start to take advantage of the more multi cluster environments.

Swapnil Bhartiya: Shaun, once again, thank you so much for joining me today. Talk about MT and beyond that. Thanks for all those great insights. And as usual, I would love to have you back on the show.

Shaun O’Meara: Thank you very much.

[/read]

BlackFog launches AI-based enterprise platform for ransomware and data loss prevention

Previous article

What kind of industries rely on Akamai’s API security solutions | Rupesh Chokshi

Next article