<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[k0s — The Kubernetes Distribution - Medium]]></title>
        <description><![CDATA[k0s is the simple, solid &amp; certified Kubernetes distribution that works on any infrastructure: public &amp; private clouds, on-premises, edge &amp; hybrid. It’s 100% open source &amp; free. This is the official blog of k0s community. - Medium]]></description>
        <link>https://medium.com/k0sproject?source=rss----8f45b8534c0d---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>k0s — The Kubernetes Distribution - Medium</title>
            <link>https://medium.com/k0sproject?source=rss----8f45b8534c0d---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 05:03:15 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/k0sproject" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The k0sproject Blog Has a New Home!]]></title>
            <link>https://medium.com/k0sproject/the-k0sproject-blog-has-a-new-home-c9334409e893?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/c9334409e893</guid>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[k0s]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Wed, 22 Oct 2025 07:21:19 GMT</pubDate>
            <atom:updated>2025-10-24T09:30:02.108Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Nn5IELnDbcY87u7h" /><figcaption>Photo by <a href="https://unsplash.com/@retrosupply?utm_source=medium&amp;utm_medium=referral">RetroSupply</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Hey everyone 👋</p><p>We’re excited to share that the <strong>k0sproject blog</strong> has moved to a new home:<br> 👉 <a href="https://blog.k0sproject.io/"><strong>blog.k0sproject.io</strong></a></p><p>From now on, all new posts — release notes, technical deep dives, and community stories — will live there.</p><h3>Why We Moved</h3><p>As an <strong>open source project</strong>, we believe in openness, transparency, and collaboration.<br> While Medium has served us well, we wanted a space that’s <strong>fully under our control</strong>, built with open tools, and part of the <strong>k0sproject.io</strong> ecosystem.</p><p>The new blog allows us to:</p><ul><li>Publish content aligned with our open-source philosophy</li><li>Integrate better with our website and documentation</li><li>Keep everything versioned and open — just like our code</li><li><strong>Empower anyone in the community to contribute easily</strong> — every article is now just a <strong>pull request away! </strong>Just fork <a href="https://github.com/k0sproject/blog">https://github.com/k0sproject/blog</a> repo, write your kick-ass article in Markdown, and create a PR.</li></ul><h3>🚀 What’s Next</h3><p>Expect more frequent updates, technical content, and stories from contributors — all hosted at <a href="https://blog.k0sproject.io/"><strong>blog.k0sproject.io</strong></a>.</p><p>If you’ve been following us here on Medium, thank you for your support ❤️<br> We’d love for you to continue the journey with us at our new home.</p><p>See you there!<br> — The k0sproject Team</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c9334409e893" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/the-k0sproject-blog-has-a-new-home-c9334409e893">The k0sproject Blog Has a New Home!</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[k0s 1.32 Release: What’s New in Lightweight Kubernetes]]></title>
            <link>https://medium.com/k0sproject/k0s-1-32-release-whats-new-in-lightweight-kubernetes-47b1b0bb7123?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/47b1b0bb7123</guid>
            <category><![CDATA[containers]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[k0s]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Wed, 29 Jan 2025 08:11:17 GMT</pubDate>
            <atom:updated>2025-01-29T08:11:17.482Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*n49JObaXmFri1wFF" /><figcaption>Photo by <a href="https://unsplash.com/@lorenzoherrera?utm_source=medium&amp;utm_medium=referral">Lorenzo Herrera</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>We’re excited to announce the release of k0s <strong>1.32</strong>, packed with powerful enhancements, critical fixes, and new features. This release continues to solidify <a href="https://k0sproject.io/">k0s</a> as the lightweight and secure Kubernetes distribution for modern workloads.</p><h3>What’s new in Kubernetes 1.32</h3><p>As always, a new Kubernetes minor release comes with tons of fixes and some new features. Here are a few changes that we wanted to highlight in the Kubernetes 1.32 release.</p><p><strong>Security Improvements</strong>: Addressed CVE-2024–9042, a Windows node vulnerability, ensuring better security for querying node logs.</p><p><strong>API Enhancements</strong>:</p><ul><li>Increased pod limits for shared ResourceClaims from 32 to 256.</li><li>Introduced support for asynchronous pod preemption and resizing pod resources using a /resize subresource.</li><li>Added the singleProcessOOMKill flag for cgroups v2 to prevent entire containers from being OOM killed when a single process exceeds memory limits.</li><li>Enabled stricter validation and features for mutating admission policies.</li></ul><p><strong>Feature Updates</strong>:</p><ul><li>Promoted multiple features to GA, including StatefulSetAutoDeletePVC, RetryGenerateName, and the Kubernetes memory manager.</li><li>Improved scheduler performance with advanced QueueingHint and reduced memory consumption.</li><li>Added new kubelet metrics and support for systemd watchdog integration.</li></ul><p><strong>Bug Fixes</strong>: Resolved issues in kube-controller-manager, kube-scheduler, and kubelet, including StatefulSet handling, storage bugs, and race conditions during node restarts.</p><p><strong>Dependency Updates</strong>: Built with Go 1.23.4 for performance and compatibility improvements​</p><p>For more details, see the full Kubernetes 1.32 <a href="https://github.com/kubernetes/kubernetes/blob/master/CHANGELOG/CHANGELOG-1.32.md">release notes</a>.</p><h3>What’s New in k0s 1.32</h3><p>This k0s release comes with a few noteworthy user-facing improvements but as always, behind the scenes, there are plenty of fixes, enhancements, and refactorings that make k0s more stable and secure. See the full changelog on the GitHub <a href="https://github.com/k0sproject/k0s/releases">release page</a>.</p><h3>Key Features and Enhancements</h3><ul><li><strong>Runtime Configuration Enhancements</strong>:<br><strong>New Output Formats</strong>: The k0s sysinfo command now features a flag for structured output, streamlining integration into automated workflows.<br><strong>Streamlined API Configurations</strong>: API subcommands now support runtime configurations via stdin, removing the need for defaults and simplifying complex setups.</li><li><strong>Kubernetes and Core Component Updates</strong>:<br><strong>Kubernetes</strong> upgraded to <strong>1.32.1</strong>, bringing the latest features, security fixes, and deprecation handling for legacy APIs.<br>Networking improvements with <strong>Calico 3.29.1</strong> and enhanced DNS reliability with <strong>CoreDNS 1.12.0</strong>.</li><li><strong>User space proxy support in Control Plane Load Balancing (CPLB):<br></strong>For environments where IPVS is unavailable or the networking layer doesn’t behave well with IPVS. Stay tuned for a more detailed blog post for the entire CPLB feature.</li><li><strong>Build and Deployment Optimizations</strong>:<br>Updated CI/CD pipelines to use <strong>Ubuntu 24.04</strong> for increased compatibility with modern environments.<br>Added support for custom kubelet root directories with the--kubelet-root-dir flag, allowing users more flexibility in managing storage and configurations.</li><li><strong>Security and Maintenance</strong>:<br>Patched several CVEs, including <strong>CVE-2024–9042</strong>, ensuring a secure runtime environment.<br>Added linters such as <em>nolintlint</em> and <em>perfsprint</em> to improve code maintainability and performance over time.</li></ul><h3>Core Component Updates</h3><ul><li><strong>Calico</strong> upgraded to <strong>3.29.1</strong> for improved networking performance.</li><li><strong>CoreDNS</strong> updated to <strong>1.12.0</strong>, providing more reliable DNS functionality.</li><li><strong>Containerd</strong> upgraded to <strong>1.7.24</strong>, ensuring better compatibility and performance.</li></ul><h3>Major Fixes</h3><h3>Etcd Stability Improvements</h3><ul><li>Fixed an issue where nodes could inadvertently create new clusters during initial etcd sync failures.</li><li>Introduced retry logic for more robust handling of etcd leader resynchronization.</li></ul><h3>Locking and Configuration Reliability</h3><ul><li>Replaced PID-based runtime configuration locking with file-based locks (flock), improving reliability in multi-instance setups.</li><li>Resolved cache issues related to node configuration, ensuring consistent and reliable reads.</li></ul><h3>Key Statistics for 1.32</h3><ul><li><strong>Total Commits:</strong> 132</li><li><strong>Merged Pull Requests:</strong> 87</li><li><strong>Contributors:</strong> 14</li><li><strong>Updated Dependencies:</strong> 20+</li></ul><h3>Get Started Today</h3><p>Download the latest release of<a href="https://github.com/k0sproject/k0s"> k0s 1.32</a> and explore the updated documentation to experience the improvements firsthand.</p><p>Let us know what you think of the new release and how we can continue to improve!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=47b1b0bb7123" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0s-1-32-release-whats-new-in-lightweight-kubernetes-47b1b0bb7123">k0s 1.32 Release: What’s New in Lightweight Kubernetes</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[k0s 1.31 Released]]></title>
            <link>https://medium.com/k0sproject/k0s-1-31-released-7196ee8743f9?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/7196ee8743f9</guid>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[containers]]></category>
            <category><![CDATA[k0s]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Mon, 30 Sep 2024 18:47:04 GMT</pubDate>
            <atom:updated>2024-09-30T18:51:10.432Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cES-y71afSj6XYSLytqJNQ.jpeg" /></figure><h3>Major updates, features, and enhancements for k0s, including nftables. Plus: k0s applies for CNCF Sandbox project status</h3><p>The k0s project has seen significant updates over the past couple of months, introducing new features, improving existing functionalities, and ensuring the platform’s stability through several critical fixes and dependency upgrades. Here’s a breakdown of the most important changes and what they mean for you. To experience it, just <a href="https://github.com/k0sproject/k0s/releases">download the 1.31 release of k0s</a>.</p><p>The most notable new feature in this release is of course <a href="https://kubernetes.io/blog/2024/08/13/kubernetes-v1-31-release/">Kubernetes 1.31, called Elli</a>. Per kubernetes.io, “Elli is a cute and joyful dog, with a heart of gold and a nice sailor’s cap, as a playful wink to the huge and diverse family of Kubernetes contributors.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*QJu1GVaHd2FpUEHQ.png" /></figure><h3>New Features and Enhancements</h3><p>Team k0s thinks the following Kubernetes 1.31 new features deserve special mention:</p><p><strong>nftables backend</strong> offers significant performance and scalability improvements over the iptables API. Enabled by default behind the NFTablesProxyMode feature gate, nftables mode excels at processing service endpoint changes and packets more efficiently, particularly in clusters with tens of thousands of services. It leverages advanced kernel processing capabilities, enhancing overall network efficiency and throughput.</p><p><strong>OCI images as volumes</strong> is an alpha feature that allows OCI images to be used directly as volumes in Pods. This enhancement supports AI and ML use cases by enabling users to specify OCI image references as volumes and mount them within containers. To use this feature, enable the ImageVolume feature gate.</p><p><strong>cgroup v1 in maintenance mode</strong>: In Kubernetes v1.31, support for cgroup v1 is moving to maintenance mode, reflecting the industry’s shift towards cgroup v2. While critical security fixes will continue, no new features will be added, and bug fixes will be on a best-effort basis. Users are encouraged to transition to cgroup v2 for enhanced functionality and scalability.</p><h3>k0s as a CNCF sandbox project?</h3><p>About a year ago, we posted a note in the k0s open source repository asking our community whether k0s should apply to become a CNCF project at ‘sandbox’ level. The result was a resounding yes, and now we’ve submitted the application. If you’d like to see k0s as a CNCF project, and want to contribute to the process, please feel free to push your endorsement on the tracking issue: <a href="https://github.com/cncf/sandbox/issues/125">https://github.com/cncf/sandbox/issues/125</a></p><h3>k0s Changes</h3><p><strong>Dependency Updates. </strong>As we all know, we cannot run and build k0s or Kubernetes without dependencies. Keeping dependencies up to date helps mitigate bugs and ensures the project stays secure.</p><p>For k0s 1.31, the major components are at the following versions:</p><ul><li>Kubernetes: 1.31.1</li><li>Containerd: 1.7.22</li><li>Runc: 1.1.14</li><li>Etcd: 3.5.16</li><li>Kine: 0.13.0</li><li>Metrics server: 0.7.2</li><li>Calico: 3.28.1</li><li>Kube-router: 2.2.1</li></ul><p>For all the major components, k0s is now pretty much on the latest versions.</p><p>One special note on Kine: version 0.13.0 of Kine offers greatly improved performance when using PostgreSQL as the backend. According to the<a href="https://github.com/k3s-io/kine/pull/323"> enhancement</a> authors, they’ve seen almost 35x improvements when the DB holds 1M objects. I mean … wow, that’s just incredible.</p><p><strong>Dependabot.</strong> One of the tools we utilize to help us stay up to date is <a href="https://github.com/dependabot">Dependabot</a>. Since the last minor release, 1.30, Dependabot has made more than 80 dependency updates. These updates ensure that k0s remains up-to-date with the broader ecosystem, providing users with the latest features and security enhancements.</p><p><strong>Dual-stack (IPv4 &amp; IPv6) with Kube-Router. </strong>k0s now supports running dual-stack clusters with <strong>kube-router</strong>. In 1.31, we updated kube-router to the new 2.x series, which includes long-awaited support for dual-stack clusters. Starting to enable IPv6 in the cluster helps you prepare for the inevitable future of IPv6 and provides flexibility for integrating with various network environments.</p><p>As part of this effort, we’ve also done a big overhaul for <a href="https://docs.k0sproject.io/stable/dual-stack/?h=dual+stack">dual-stack documentation</a>.</p><p><strong>Ability to configure bind address. </strong>k0s 1.31 comes with new config options to configure which addresses the <strong>Kubernetes API server</strong> binds to. This allows people running on hosts with multiple NICs/addresses to configure on which interface the API is visible.</p><p><strong>Cleanup obsolete images imported via airgap bundles. </strong>One of the challenges of using k0s airgap image bundles has always been cleanup. To prevent kubelet from removing critical system images during its cleanup routines, k0s pins these images on containerd. With this enhancement, k0s now tracks the images it imports and can un-pin the ones no longer in use, making them free for cleanup.</p><p>Apart from these user-facing bigger changes, we’ve done a ton of smaller quality-of-life improvements, bug fixes, and chore work to keep the k0s code base stable and maintainable. Altogether, the 1.31 release cycle includes 500 or so commits. Read the <a href="https://github.com/k0sproject/k0s/releases/tag/v1.31.1%2Bk0s.0">k0s 1.31 release notes</a> for the full scoop.</p><h3>Shout Outs</h3><p>We want to give a special shout-out to our friends at <a href="https://www.replicated.com/">Replicated</a>. The Replicated team helped with several fixes and some new features during this release. We also have plans to deepen our relationship over the k0s project, so stay tuned.</p><p>We also want to highlight the folks submitting their first contribution to k0s:</p><ul><li><a href="https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Aplaes">https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Aplaes</a></li><li><a href="https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Aianb-mp">https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Aianb-mp</a></li><li><a href="https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Anekwar">https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Anekwar</a></li><li><a href="https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Abizhao">https://github.com/k0sproject/k0s/issues?q=is%3Apr+author%3Abizhao</a></li></ul><p>Without the community, k0s would not be where it is today. Huge shout-out to the entire community: thank you all!</p><h3>Community Hours</h3><p>We’re running monthly <strong>community hours</strong> calls for the project on the last Tuesday of the month.</p><p>Check the details at:<a href="https://dateful.com/eventlink/2735919704">https://dateful.com/eventlink/2735919704</a></p><p>We’ve also got plans to improve discussion channels for both k0s users and maintainers. Stay tuned!</p><h3>What’s Next?</h3><p>As we move forward, the k0s project team remains committed to improving the platform’s stability, performance, and user experience. Stay tuned for more updates, and as always, we welcome feedback and contributions from the community.</p><p>If you happen to be at KubeCon+CloudNativeCon in Salt Lake City in November, come say hi at the Mirantis booth (R22). We’re always super happy to meet with community members and learn more about their use cases.</p><h3>Try k0s 1.31!</h3><p>If you haven’t already, make sure to <a href="https://github.com/k0sproject/k0s/releases">upgrade to the latest version of k0s</a> to take advantage of these new features and improvements. For more detailed information, you can always check out our <a href="https://github.com/k0sproject/k0s">GitHub repository</a> and <a href="https://docs.k0sproject.io/stable/">docs</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7196ee8743f9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0s-1-31-released-7196ee8743f9">k0s 1.31 Released</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[k0smotron is growing up]]></title>
            <link>https://medium.com/k0sproject/k0smotron-is-growing-up-22cb6be97daf?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/22cb6be97daf</guid>
            <category><![CDATA[cluster-api]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[k0s]]></category>
            <category><![CDATA[containers]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Tue, 11 Jun 2024 16:02:42 GMT</pubDate>
            <atom:updated>2024-06-11T16:02:41.270Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*26Y6vFR2Pg7tE1u6iMq1SA.png" /></figure><p>k0smotron 1.0 announces general availability with new enterprise capabilities, including enabling remote deployment of whole clusters on any infrastructure, improving control plane high availability, and enabling updates-in-place. Making things more seamless and simple for Kubernetes multi-cluster deployments.</p><p>Organizations using Kubernetes face the challenge of defining, provisioning, scaling, and lifecycle-managing clusters. Many need to do this on more than one kind of infrastructure — private cloud, public cloud, bare metal, etc.</p><p>Platform teams need ways to define k8s clusters and related services, so that operators and developers can deploy them on demand. Platform teams and architects need ways to specify deployments with novel footprints — conventional or hosted control planes, local or remote workers — to suit requirements of modern hybrid, edge, and other use cases.</p><p>k0smotron — introduced last year and announced in November — offers an open source, Kubernetes-native solution to these multi-cluster/hybrid challenges. It works hand-in-hand with Kubernetes ClusterAPI to enable declarative infrastructure provisioning on your choice of public clouds, private clouds, managed or unmanaged bare metal, or any combination thereof. k0smotron helps you configure, deploy, and lifecycle manage k0s Kubernetes child clusters across all these infrastructures: as many as you’ve installed CAPI providers to support. (<a href="https://www.mirantis.com/blog/how-to-use-cluster-api">Learn more about ClusterAPI</a>.)</p><p>But there’s more. By leveraging the unique advantages of k0s Kubernetes (like control plane/worker network separation), k0smotron 1.0 makes it easier to configure exotic cluster footprints for hybrid, edge, and even wilder use cases:</p><ul><li>Control planes in the cloud with high-powered worker nodes on customer premises</li><li>Containerized control planes on a public cloud ‘mothership’ cluster, with remote workers at the far edge, behind a firewall</li><li>IoT with fully-mobile workers</li></ul><p>k0smotron can do all this — and k0s helps keep the networking fairly simple.</p><h3>Pave the World with Kubernetes</h3><p>By putting everything you need for platform management behind the Kubernetes API and declarative configuration, k0smotron provides what you need for Kubernetes-native multi-cluster GitOps, and gets you a big step closer to using Kubernetes to ‘pave the world’ of infrastructures under a Kubernetes substrate.</p><p>With k0smotron 1.0 (combined with Helm for efficiently installing applications and services on child clusters), platform engineers can package up full solution-stacks in a single set of version-controlled YAML files: composing everything a software development team needs (i.e., infra(s), cluster(s), installed services and applications) to cover a broad range of use cases, such as:</p><ul><li>Entire SaaS point-of-presence configurations</li><li>Complete edge applications with endpoint architectures</li><li>Multi-service mobile network platforms and services</li></ul><p>All now easily-deployed and self-maintaining.</p><p>One obvious benefit for platform engineers: simplicity. With k0smotron, you don’t need to install, maintain, update, manage a ton of different deployment tools (and their own collections of YAML or other configuration codebases).</p><p>k0smotron also makes sense in terms of macro trends:</p><ul><li>First, Kubernetes is beyond a doubt the future platform for everything. It’s currently the best tool we have for running complex modern applications. And it provides the core platform for all the higher-order services we need (like serverless and PaaS and functions-as-a-service, etc). So using it to pave over infrastructure simplifies our lives, and even lets us think about infrastructure in new ways — we can even think about public clouds as commodity utilities. (Remember when we all thought about ‘cloud’ that way and imagined a future in which compute/storage/network providers would compete for our workloads without locking us in and charging us a ton of money? This was the original goal.)</li><li>Second, Kubernetes is good at this. It’s a standardized machine for looking at declarative files describing a desired state, and then (either directly or through standard middleware) converging reality on that state. That’s why modern configuration-automation tools all work basically this way.</li></ul><p>Having &lt;everything&gt; behind the Kubernetes API (and making sure this is entirely open source and standardized) gives us superpowers:</p><ul><li>One way to describe and implement deep solution-stacks</li><li>One way to inject and implement observability</li><li>One point of integration for DevSecOps</li><li>Maybe most importantly, one point of integration for AIOps.</li></ul><h3>Tested by Actual Customers</h3><p>k0smotron has been in the hands of key Mirantis customers for a while now, and we’re so grateful for their help and enthusiasm. Today we’re announcing general availability of k0smotron 1.0, along with enterprise support. k0smotron 1.0 has some exciting new features that make it an even more complete solution. No longer primarily a “hosted control plane manager,” k0smotron 1.0 is now a true multi-cluster/hybrid management solution.</p><p>New features in k0smotron 1.0 include:</p><h3>Remote Machine Support</h3><p><strong><em>k0smotron RemoteMachine is now a full-fledged infrastructure provider for Kubernetes ClusterAPI (see the </em></strong><a href="https://cluster-api.sigs.k8s.io/reference/providers"><strong><em>Providers list</em></strong></a><strong><em>), enabling provisioning of remote machines via SSH.</em></strong> k0smotron can now use RemoteMachine to (for example) build and lifecycle-manage distributed or monolithic k0s clusters with hosted or conventional control planes.</p><p>Benefit: Convenient operations on unmanaged bare metal, opening up use cases in data centers and at the edge. For example, you can build a cluster with a k0s hosted control plane on a mothership cluster in a public cloud, with remote workers on bare metal Linux machines at an edge location or on a customer’s premises.</p><p>k0smotron + CAPI providers (including RemoteMachine) can also work as a high-level infrastructure provider with non-k0s control plane providers. So in principle, you can now use k0smotron RemoteMachine to deploy k3s or MicroK8s (or many other) cluster flavors, as well. k0smotron RemoteMachine can also support custom SSH connectors, so if your infrastructure requires you to use Teleport, for example, you can <a href="https://docs.k0smotron.io/stable/capi-remotemachine-teleport/">customize the provider</a> to run provisioning with Teleport-required infrastructure authorization.</p><h3>Control Plane Scaling and Updates</h3><p>Conventional k0s clusters can automatically and non-disruptively update themselves (controllers and workers) by leveraging the Autopilot operator. Previous versions of k0smotron, however, didn’t fully support this paradigm, since k0smotron had no means for signaling the in-cluster Autopilot to update hosted child cluster control planes. <strong><em>Now, for child clusters running in virtual machines, k0smotron 1.0 integrates with the child cluster’s Autopilot to update the full cluster (controllers and workers) — using a node-by-node strategy that keeps clusters available.</em></strong></p><p>Benefit: Easy scaling of child clusters on multiple infrastructure platforms, in a standardized way.</p><h3>External etcd Support for Control Plane HA</h3><p>To enhance high availability (HA) for hosted control planes (running a k0s control plane in pods), <strong><em>k0smotron 1.0 will now deploy etcd in a separate pod (and statefulset) from the hosted control plane component.</em></strong> Previously, running a highly available hosted control plane (i.e., multiple containerized controllers, deployed to different fault domains) was challenging, due to potential split-brain scenarios as etcd (effectively part of each control plane) got scaled up and down. With the new update, etcd is managed independently of other HCP components (in a separate set of pods), letting it scale independently. etcd can also be snapshotted and restored, enabling robust full cluster upgrades and restoration of state.</p><p><em>Note: this involves a breaking change with regard to versions of k0smotron prior to 0.9.5 — see </em><a href="https://github.com/k0sproject/k0smotron/releases/tag/v0.9.5"><em>this note</em></a><em> for more information and a simple workaround.</em></p><p>Benefit: Freely-scalable hosted control planes — so you can balance redundant availability with utilization.</p><h3>Clusterctl Support</h3><p>Finally, <strong><em>k0smotron 1.0 adds support for installing k0smotron through ClusterAPI’s clusterctl CLI.</em></strong> Users can install and manage k0smotron directly through clusterctl instead of separately — making it easier to, in effect, turn a ClusterAPI setup for infrastructure management into a complete multi-cluster solution. Simple:</p><pre>clusterctl init --bootstrap k0sproject-k0smotron \<br> -- control-plane k0sproject-k0smotron \<br> -- infrastructure k0sproject-k0smotron</pre><p>Benefit: Simpler setup. From zero to ‘complete multi-cluster manager’ in one command.</p><h3>ClusterClass Support</h3><p>Defining a complete Kubernetes cluster plus its infrastructure for ClusterAPI previously required pretty long and complicated YAML files — sometimes in excess of 500 lines. <strong><em>ClusterClass — now an alpha feature in ClusterAPI — lets you standardize an infrastructure definition, then reference it (with variations) in just a few lines of code.</em></strong> This is supported in k0smotron 1.0 — streamlining files, making them more readable and manageable, enabling development of reusable configurations, and simplifying automation.</p><p>Benefit: Improved reusability and easier-to-read code. Less chance of errors creeping in.</p><h3>Try k0smotron today!</h3><p>k0smotron 1.0 is now a full multi-cluster manager — hosted control planes are no longer the whole story (though they are extremely cool and useful).</p><p>Also exciting, k0smotron will now be fully supported globally by Mirantis — so enterprises wishing to explore solutions or bring use cases to life with k0smotron at scale can do so with the backing of Mirantis’ deep Customer Success engineering bench, as well as the particular expertise of Team k0s itself. Support will be available in three tiers:</p><ul><li>LabCare — 8 x 5 (M-F) business hours support. For non-production workloads</li><li>Opscare — 24/7 support, with 30-minute response time SLA on critical issues. For enterprise production environments</li><li>Opscare Plus — Fully managed, proactive support, with &gt;99.9% availability</li></ul><p>Please note that support for k0smotron includes support for <a href="https://k0sproject.io/">k0s</a> — the zero friction, zero dependencies, CNCF-validated Kubernetes.</p><p>Try k0smotron 1.0 today and experience the future of Kubernetes cluster management. For more information, visit <a href="http://k0smotron.io/">k0smotron.io</a> — our <a href="https://github.com/k0sproject/k0smotron">GitHub</a> for releases, community, or contributing — or schedule a <a href="https://mirantis.com/contact">live demo</a> with Mirantis.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=22cb6be97daf" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0smotron-is-growing-up-22cb6be97daf">k0smotron is growing up</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[K0s 1.30 Released]]></title>
            <link>https://medium.com/k0sproject/k0s-1-30-released-b1bbb9f2147a?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/b1bbb9f2147a</guid>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[etcd]]></category>
            <category><![CDATA[k0s]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Mon, 13 May 2024 10:31:12 GMT</pubDate>
            <atom:updated>2024-05-13T10:31:12.131Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*2YxVumf-tRpvtTis" /></figure><h3>New k0s delivers Kubernetes 1.30 (“UWUbernetes”) plus new k0s-only features — including improvements to control plane management, “virtual IPs” (internal LB) out-of-box, and an exciting new Support Insight feature, powered by open source Replicated Troubleshoot</h3><p>K0s 1.30 just dropped, delivering upstream Kubernetes 1.30 (“UWUbernetes”). Tiny heart fingers to the contributors, and to Team k0s for delivering the cutest Kubernetes ever in an extremely adorable (not to mention usable) form.</p><p>Kubernetes 1.30 advanced several enhancements to stable status, and introduces a flock of new features as beta or alpha. Overall, it feels (no surprise to anyone) that Kubernetes is maturing — features promoted with 1.30 include ‘fit and finish’ improvements enhancing basic container security and making things like Horizontal Pod Autoscaler more flexible and usable. For the complete scoop, read the <a href="https://kubernetes.io/blog/2024/04/17/kubernetes-v1-30-release/">official release notes</a>. We’d also like to give a shout-out to OVHCloud developer advocate Aurélie Vache, author/illustrator of <a href="https://www.amazon.com/Understanding-Kubernetes-visual-way-sketchnotes/dp/B0BB619188">Understanding Kubernetes in a Visual Way</a>, who just published an illustrated <a href="https://dev.to/aurelievache/understanding-kubernetes-part-51-kubernetes-130-changelog-59l7">guide to the Kubernetes 1.30 changelog</a> as unit 51 of her continuing series on <a href="https://kubernetes.io/blog/2024/04/17/kubernetes-v1-30-release/">DEV.to</a>.</p><h3>Some Kubernetes 1.30 Highlights</h3><p><strong>Pod Scheduling Readiness </strong>(now stable) cuts load on the scheduler when Kubernetes is confronted with scheduling a pod for which resources aren’t yet available. The feature introduces the notion of schedulingGates as part of the pod spec: these can be used (or removed) to control when and how Kubernetes considers the pod for scheduling. The feature can be leveraged, for example, to avoid making Kubernetes try (and fail, at cost of time and resources) to schedule a pod on not-yet-available nodes; or, for another example, to prevent cluster autoscaling from spinning up a new node to host a low-priority workload. In a world where clusters are empowered to self-scale (e.g., with ClusterAPI) in more and more dynamic ways, this seems very important. The feature is also being used in the community to implement quotas, and for security applications.</p><p><strong>HPA based on ContainerResource (vs. pod) metrics</strong> (now stable) lets you trigger Horizontal Pod Autoscaling on metrics returned by a specific container, vs. a pod, giving developers more flexibility in packaging containers in pods while also being able to designate specific containers as performance bottlenecks, and thus as scaling triggers.</p><p><strong>Min domains in PodTopologySpread</strong> (now stable) lets you define the minimum number of domains Cluster Autoscaler will maintain, preventing situations where pods can’t be scheduled because too few domains exist to fulfill requirements stated in PodTopologySpread constraints.</p><p><strong>User namespaces </strong>(now beta) isolates the UIDs and GIDs of containers from the host. This prevents a container running as root from having root access to the host node it’s running on.</p><p><strong>Contextual logging </strong>(now beta) lets you inject context information like service names and transaction IDs into logs, using WithName and WithValues. This is a pretty big step forward in enabling more context-aware observability and clearer insights into how Kubernetes environments and applications are working.</p><h3>New Features in k0s 1.30</h3><p>Two new k0s features — both added in response to user requests — are headlined in this latest k0s release.</p><p><strong>Declarative management of etcd state</strong> overcomes a particular challenge of using k0smotron and ClusterAPI to horizontally scale k0s Kubernetes control planes on virtual machines. For example, you might want to scale down the control plane on a self-scaling k0s cluster running on OpenStack or AWS VMs.</p><p>In such a situation, the ClusterAPI controllers run as workloads. And the controller/worker separation architecture of k0s means that these infrastructure controllers can’t access etcd — a control plane element. That’s a problem, because when ClusterAPI wants to get rid of a control plane node, it needs to tell etcd that the node has left the cluster — and can’t. Eventually your etcd dies from trying to deal with ghost members that never boot up.</p><p>Declarative management of etcd state routes around the problem by providing a mechanism that lets the ClusterAPI controller (or Terraform, or other tools) tell etcd what’s going on when control plane nodes are added or deleted. Quite cool: the facility even leaves an audit trail of ‘breadcrumbs’ in the config, like timestamps for when nodes leave the cluster.</p><p><strong>Control plane load balancer (aka. “Virtual IPs”)</strong> solves a big challenge when k0s clusters are deployed in environments that don’t provide integrated external load balancing for the Kubernetes API (e.g., vSphere, which has no default facility for load balancing). With control plane load balancer, you give k0s an IP within the local subnet, and k0s sets things up with Keepalived to load-balance traffic to this IP across available controllers. (We examined KubeVIP to provide this functionality, but it didn’t meet all our requirements, some of which are unique to k0s’ architecture.)</p><p>This is roughly what KubeVIP does — the difference being that KubeVIP expects to pick up back end addresses for APIservers from the node objects hosting controllers. And k0s, because it runs controller components as bare processes, doesn’t really have controller nodes, per se — so this won’t work. KubeVIP can also only balance one port — whereas k0s needs two ports: one for the Kubernetes API and one port reserved for joining new nodes to the cluster.</p><h3>Introducing Support Insight</h3><p>In the past, raising support issues around k0s could involve a lot of back-and-forth, with Team k0s members saying things like “what do you see when you run this command?” Annoying and time-consuming.</p><p>K0s Support Insight, introduced with k0s 1.30, solves this problem. It’s a new way of diagnosing Kubernetes issues, so we can help k0s users solve problems more efficiently.</p><p>K0s Support Insight is powered by the open source Troubleshoot project, from the team at Replicated. Check out <a href="https://troubleshoot.sh/">the Troubleshoot website</a> (especially the animated terminal session videos) and their <a href="https://github.com/replicatedhq/troubleshoot">GitHub repo</a>. We’ve started collaborating (in both directions) with the team at Replicated, and we’re excited to deliver a first taste of the resulting improvements with k0s 1.30.</p><p>What’s Troubleshoot? Replicated created the Troubleshoot system to help Kubernetes platform engineers and cloud-native application builders communicate better with their users. Team k0s is adopting their <em>support-bundle</em> component (called with kubectl support-bundle &lt;URL&gt;), which downloads a declarative description of data collection and analysis operations. These get run against your cluster with kubectl (without installing anything) and the resulting data model gets saved locally, with all sensitive information (e.g., secrets) redacted. You can review it yourself or upload it to Mirantis (who provide commercial support for k0s) or to Team k0s via GitHub issues (for community support). Experts can then analyze the support bundle locally as if it were your cluster, quickly surfacing issues.</p><p>For k0s 1.30, the team has developed a support bundle configuration that efficiently collects every scrap of data, logs, etc., that we might need to solve virtually any problem, or help independent-minded users solve it all by themselves. For details, visit the <a href="https://docs.k0sproject.io/head/support-dump/">k0s docs</a>.</p><h3>Try k0s 1.30</h3><p>If you haven’t tried k0s Kubernetes, it’s a great moment to start. Simple, easy to install and operate on almost anything, from public clouds to datacenter servers to Raspberry Pis and other IoT-scale edge devices. Please visit the <a href="https://github.com/k0sproject/k0s/releases">Releases page</a> of our GitHub repo, and <a href="https://docs.k0sproject.io/stable/">check out the docs</a> today!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b1bbb9f2147a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0s-1-30-released-b1bbb9f2147a">K0s 1.30 Released</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[k0smotron Anywhere: Now a complete ClusterAPI infrastructure provider]]></title>
            <link>https://medium.com/k0sproject/k0smotron-anywhere-now-a-complete-clusterapi-infrastructure-provider-7da7b633331f?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/7da7b633331f</guid>
            <category><![CDATA[k0s]]></category>
            <category><![CDATA[containers]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[cluster-api]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Thu, 18 Jan 2024 09:54:36 GMT</pubDate>
            <atom:updated>2024-01-18T09:54:35.974Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vcrBwi7h0qvxhRaDNPeTVg.png" /></figure><p>Managing multiple Kubernetes clusters, maybe even over multiple infrastructures, is definitely not easy. Cluster API can help you by bringing in a declarative way to manage both the infrastructure and the cluster setup. But what if you want to use on-premises, bare metal or other like such environments where existing Cluster API providers do not really have coverage? Not to worry, we’ve got your back. Now with <a href="https://k0smotron.io/">k0smotron</a> Anywhere, you can use Cluster API on ANY infrastructure, even bare metal, without the need to add any complexities to your infrastructure.</p><p>Today, we’re thrilled to announce availability of <a href="https://github.com/k0sproject/k0smotron/releases/tag/v0.8.0">k0smotron v0.8.0</a> and the completion of k0smotron Anywhere (a.k.a <a href="https://docs.k0smotron.io/stable/capi-remote/">Remote Machine Infrastructure Provider</a>). The completion means k0smotron Anywhere can now function as a <a href="https://cluster-api.sigs.k8s.io/">Cluster API</a>-compliant infrastructure provider. This opens up new possibilities for managing remote machines seamlessly in your Kubernetes clusters.</p><p>In case you’ve forgotten what k0smotron is, it is essentially a set of Kubernetes controllers that allows you to run and manage multiple Kubernetes cluster control planes as pods within a single Kubernetes cluster. This allows for more granular control and much greater resiliency due to reducing the “blast radius” of any control plane failure. As Kubernetes gets deployed into more distributed applications, this more granular approach has become critical for reliability and resiliency. Since the <a href="https://medium.com/k0sproject/introducing-k0smotron-c2ed6535ddc8?source=collection_home---6------4-----------------------">initial release</a> around 6 months ago it has evolved into much more than just control planes, it also acts as a general purpose <a href="https://k0sproject.io/">k0s</a> provider for ClusterAPI.</p><p>For those of you who’ve been using or following our k0s project, you know that we take our “Zero story” very seriously. In case of k0smotron Anywhere this means zero dependencies. The latest version is a testament how serious we are with that, practically allowing you to use ClusterAPI in provider-less mode. What do we mean by that? Read along and we’ll show you what it is and how it works.</p><h3>What is k0smotron Anywhere?</h3><p>k0smotron Anywhere extends k0smotron’s capabilities as a ClusterAPI provider to seamlessly manage remote machines via SSH connections. This means you can now provision machines in any infrastructure, not limited to environments that support ClusterAPI and are API powered. You could think of it being the providerless provider for Cluster API.</p><h3>Key Benefits</h3><ul><li><strong>Flexibility: </strong>Run your Kubernetes workloads on machines located anywhere, whether in the cloud, on-premises, or in hybrid setups.</li><li><strong>Infrastructure Neutrality</strong>: k0smotron Anywhere supports ANY infrastructure, providing a consistent approach to managing your Kubernetes clusters via Cluster API.</li><li><strong>Use Existing Machines: </strong>Leverage your existing machines by integrating them into your Kubernetes clusters with ease.</li></ul><h3>How Does it Work?</h3><p>With the k0smotron Anywhere feature, the possibilities are endless. Create RemoteMachines that define your remote infrastructure, specify their details like address, port, user, and SSH key. These remote machines can then be used as part of your Kubernetes cluster.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sHhY7KPojPJex_t3vwcRpw.png" /></figure><p>Essentially the RemoteMachine controller waits until it sees the cloud-init secret created by the bootstrap provider. Once the cloud-init is ready, it’ll connect to the RemoteMachine using SSH and executes the cloud init commands.</p><p>So say for example that you have a bare metal machine, running in your private data centre. Here’s an example how to define that as a RemoteMachine:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d119bc874c6f23b30c51a7b239f1084c/href">https://medium.com/media/d119bc874c6f23b30c51a7b239f1084c/href</a></iframe><p>As you see, you only need to provide the SSH connectivity details like address, user and SSH key, via a secret, and you’re good to go. k0smotron RemoteMachine controller then uses these details, connects to the machine and executes the Bootstrap provider created cloud-init configuration. And with k0s “Zero Dependencies”, you basically only need to have the base Linux OS installed, nothing else. You’re free to use what ever Linux distro is your favourite and there’s absolutely no need to add any provisioning or other complexities to your infrastructure.</p><p>You can then use the RemoteMachine as part of the cluster, just like with any other infrastructure provider. For example you could use it as a worker node via a Machine object:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/1c999b386497ecb399c28681f348b30b/href">https://medium.com/media/1c999b386497ecb399c28681f348b30b/href</a></iframe><h3>Using RemoteMachine with MachineDeployments and ControlPlanes</h3><p>As obvious, a RemoteMachine is a single machine ready to run as a Kubernetes node, either as a worker node or a control plane node. However, in some cases we need to be able to use those with ClusterAPI higher level constructs such as MachineDeployment and ControlPlanes. These require the usage of MachineTemplates. So how do we bridge the gap between a single RemoteMachine and a MachineTemplate?</p><p>This is where the concept of RemoteMachine pooling comes into play. Essentially, you can create several <a href="https://docs.k0smotron.io/stable/capi-remote/#using-remotemachines-in-machinetemplates-of-higher-level-objects">PooledRemoteMachine</a> objects to mark machines that are ready to be used.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5zTLqWui4KSktdbfexrS2g.png" /></figure><p>Here’s an example how to utilise PooledRemoteMachines via RemoteMachineTemplate objects:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/0ccdbf6bee0c863dbfbb36148fdf76b6/href">https://medium.com/media/0ccdbf6bee0c863dbfbb36148fdf76b6/href</a></iframe><p>When k0smotron sees a corresponding RemoteMachineTemplate, it’ll pick one free machine from the pool and create the corresponding RemoteMachine object. Naturally, k0smotron RemoteMachine controller keeps track of which PooledRemoteMachine is in use and for which RemoteMachine.</p><p>Once you have some pooled machines, you can use them with other ClusterAPI constructs via RemoteMachineTemplate objects:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/f07d4f899d9c110bbc60a58e890d15cd/href">https://medium.com/media/f07d4f899d9c110bbc60a58e890d15cd/href</a></iframe><p>Wait what? Does this mean I could utilize even things like autoscaling with my bare metal infrastructure? Well, sort of. Of course autoscaler cannot bring you new machines to the data center, but all this would allow you to use autoscaler, and other such things, in a way where you get a signal that we’d need another machine in the cluster. Then by adding a new machine to the pool, autoscaler and Cluster API would automatically pick that up and join it to the cluster.</p><h3>Benefits to the Ecosystem</h3><p>The introduction of k0smotron Anywhere aligns with the broader goal of making Kubernetes clusters more adaptable and accommodating various infrastructure needs. With the support of RemoteMachines and pooling, k0smotron essentially opens up Kubernetes ClusterAPI to support ANY infrastructure, not only those already having a CAPI provider available. Not to mention cases like bare metal where the infrastructure is very hard to automate.</p><h3>What’s next?</h3><p>Our upcoming focus includes the integration of <a href="https://cluster-api.sigs.k8s.io/reference/glossary.html#clusterclass">ClusterClass</a> support across k0smotron. This significant enhancement aligns with our commitment to providing a comprehensive Kubernetes management solution. Additionally, we’re actively working on incorporating built-in support for <a href="https://cluster-api.sigs.k8s.io/clusterctl/provider-contract.html#clusterctl-provider-contract">clusterctl</a>. This advancement aims to streamline the initialization of k0smotron within the management cluster.</p><p>Once these features are seamlessly integrated, k0smotron will be nearing feature completeness. To us, feature completeness signifies a pivotal step towards General Availability (GA). This transition involves extensive efforts in refining documentation to ensure clarity and simplicity for users. Furthermore, our dedication extends to establishing top-notch maintenance and support processes, solidifying k0smotron as a robust and reliable solution.</p><p>Oh, and for those of you who manages the clusters using <a href="https://k8slens.dev/">Lens</a>, we’ve started to work on a Lens extension to bring full Cluster API visibility to Lens.</p><h3>Get Started Today!</h3><p>Ready to harness the power of k0smotron Anywhere? Upgrade to the l<a href="https://github.com/k0sproject/k0smotron/releases/latest">atest version</a> and start exploring the flexibility and convenience of managing remote machines in your Kubernetes clusters.</p><p>Stay tuned for more updates and exciting features from the k0smotron team and ecosystem. Happy clustering! Oh, and do not forget to add a ⭐️ to our <a href="https://github.com/k0sproject/k0smotron/">GitHub repo</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7da7b633331f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0smotron-anywhere-now-a-complete-clusterapi-infrastructure-provider-7da7b633331f">k0smotron Anywhere: Now a complete ClusterAPI infrastructure provider</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[k0smotron v0.7.0 released]]></title>
            <link>https://medium.com/k0sproject/k0smotron-v0-7-0-released-807d9954f332?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/807d9954f332</guid>
            <category><![CDATA[k0s]]></category>
            <category><![CDATA[containers]]></category>
            <category><![CDATA[cluster-api]]></category>
            <category><![CDATA[kubernetes]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Tue, 17 Oct 2023 09:42:13 GMT</pubDate>
            <atom:updated>2023-10-17T09:42:13.324Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FWStnEXgfWJRxPZc6c0NiA.png" /></figure><p><strong>Team k0s proudly presents: k0smotron v0.7.0.</strong></p><p>This release highlights two new features: client connection tunneling and the new ClusterAPI provider for “remote machines.”</p><h3>Remote machine CAPI provider, a.k.a ‘k0smotron Anywhere’</h3><p>ClusterAPI is a Kubernetes sub-project working to provide Kubernetes-native, declarative means for provisioning, upgrading, and operating multiple clusters on various infrastructures. One of the tenets of CAPI is that it can be easily used with cloud providers and other infrastructure that is API-powered. However, there are a lot of use cases where infrastructure is not API-powered (think: bare metal servers, IoT devices, etc.). In many of these, it doesn’t make sense to invest resources into making the infrastructure API powered.</p><p>There are some CAPI providers that can work with bare metal infrastructure, but all of them require some other components — adding complexity and potentially cost</p><p>One of the promises of k0s, the Kubernetes distro that powers k0smotron-managed clusters, is that it runs anywhere. So we started to look into how we could enable k0smotron clusters even in the simplest of infrastructures. What we mean by that is that k0smotron should be able to create and manage clusters and infrastructure admins should not need to run ANY additional tooling to make this possible.</p><p>When thinking about this, we remembered the success of our existing <a href="https://github.com/k0sproject/k0sctl">k0sctl tool</a>. For those of you who have not used k0sctl, it’s a tool that can provision and manage a single cluster via yaml configuration. Essentially, you need to provide addresses and SSH credentials for each host you want to set up as part of the cluster. k0sctl connects to the hosts using SSH and orchestrates the whole cluster setup.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/544ed49739c5d1895ae55e18b5b92258/href">https://medium.com/media/544ed49739c5d1895ae55e18b5b92258/href</a></iframe><p><strong>Enter `RemoteMachines`</strong></p><p>A RemoteMachine is a machine that can be provisioned remotely via SSH. k0smotron now controls the process of bootstrapping this kind of minimal node. Like in any other Cluster API provider, the scope for a machine controller is very limited. They “only” need to provision the machine using a bootstrap provider generated cloud-init configuration. So in the case of `RemoteMachine` controller, it simply does the following:</p><ul><li>Wait to see the bootstrap cloud-init secret for the machine</li><li>Connect to the machine via SSH</li><li>Execute the cloud-init steps via SSH</li></ul><p>Let’s see an example:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/2a7f1c347c6a48f2c28afed04ca1c70b/href">https://medium.com/media/2a7f1c347c6a48f2c28afed04ca1c70b/href</a></iframe><p>So what you need to configure for your RemoteMachine is the address to connect to, the username, and which SSH key to use — via a Kubernetes secret, for security. As simple as that.</p><p>Of course, with RemoteMachines you cannot reap the full benefits of ClusterAPI. Things like automated updates via MachineDeployment are not possible. But what it does enable is, for example, managing your on-premise bare metal nodes via Cluster API. How cool is that!?</p><h3>Client connection tunneling</h3><p>In the previous k0smotron release, we introduced a control plane provider to also run the control plane using CAPI Machines. This enables you to use CAPI to provision more traditional clusters where both control and worker planes run on CAPX provider-managed infrastructure.</p><p>What if the infrastructure to run the Machines can’t be connected to directly by cluster admins? How can cluster admins connect to k0smotron child clusters with tools like kubectl and Lens?</p><p>For this use case, we’ve enabled tunneling configuration. What this does is set up a tunnel between the child cluster and management cluster in a way that can be used by clients. In the end, clients connect to a tunnel endpoint on the management cluster and via that, can access the child cluster’s API.</p><p>Read more in the k0smotron <a href="https://docs.k0smotron.io/stable/capi-controlplane-bootstrap/#client-connection-tunneling">docs</a>.</p><h3>User survey</h3><p>In case you’ve missed it, we launched a k0s <a href="https://medium.com/k0sproject/help-shape-the-future-of-k0s-kubernetes-participate-in-our-2023-users-survey-9549d3a72e87?sk=f432e56b6bee4525b45b6141ca26c8ec">user survey</a> a couple weeks ago. By answering this survey, you’ll help us prioritise features that mean the most to YOU, and in general, you’ll help us to shape the future of k0s and k0smotron.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=807d9954f332" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0smotron-v0-7-0-released-807d9954f332">k0smotron v0.7.0 released</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[k0s 1.28 released]]></title>
            <link>https://medium.com/k0sproject/k0s-1-28-released-f48e50677ab3?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/f48e50677ab3</guid>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[cloud-native]]></category>
            <category><![CDATA[container-orchestration]]></category>
            <category><![CDATA[k0s]]></category>
            <category><![CDATA[devops]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Wed, 11 Oct 2023 08:24:57 GMT</pubDate>
            <atom:updated>2023-10-13T10:20:26.824Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*C_icZmQzGJmo_MvQ" /><figcaption>Photo by <a href="https://unsplash.com/@goian?utm_source=medium&amp;utm_medium=referral">Ian Schneider</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><strong>Team k0s proudly presents: k0s 1.28 with Kubernetes 1.28 built in.</strong></p><p>Of course, the biggest thing in this brand-new version 1.28 of k0s is support for <a href="https://kubernetes.io/blog/2023/08/15/kubernetes-v1-28-release/">1.28 Kubernetes</a>, a.k.a Planternetes.</p><p>But as per usual, the team has not been just working on updating the embedded Kubernetes components. We’ve also packed in some noteworthy new features and fixes. In this post, we dive into a few of the more exciting new features and enhancements in version 1.28 of k0s.</p><h3>Updated components</h3><ul><li>Kubernetes: 1.28.2</li><li>ContainerD: 1.7.6</li><li>RunC: 1.1.9</li><li>Etcd: 3.5.9</li><li>Kine: 0.10.3</li><li>Konnectivity: 0.1.4</li><li>Kube-router: 1.6.0</li><li>Calico: 3.26.1</li><li>CoreDNS: 1.11.1</li></ul><h3>Enhanced k0s autopilot</h3><p>k0s autopilot, bundled initially into our earlier 1.24 release, lets operators easily keep clusters up-to-date by applying an update Plan object. The update Plan contains all the information needed to perform an automated update, orchestrated by the cluster itself.</p><p>Building on top of this, 1.28 now supports scheduled updates that follow a specific update channel. So how does this work?</p><pre>apiVersion: autopilot.k0sproject.io/v1beta2<br>kind: UpdateConfig<br>metadata:<br>  name: example<br>  namespace: default<br>spec:<br>  channel: latest<br>  updateServer: https://updates.k0sproject.io/<br>  upgradeStrategy:<br>    type: periodic<br>    periodic:<br>      # The folowing fields configures updates to happen only on Tue or Wed at 13:00-15:00<br>      days: [Tuesday,Wednesday]<br>      startTime: &quot;13:00&quot;<br>      length: 2h<br>  planSpec: # This defines the plan to be created IF there are updates available</pre><p>The general idea is that each minor version will have its own dedicated update channel. That lets you very easily keep up-to-date with frequent patch updates — updates that both upstream Kubernetes and k0s provide. Once you’re ready to hop on a newer minor release, just update the UpdateConfig to follow a different channel. For those who like living on the edge, we’re also providing channels for beta releases and a latest channel.</p><p>To keep things easy, the update server “protocol” is very simple. Each channel only has a simple index.yaml that contains info about the latest version that<em> </em>channel provides. This enables, for example, building your own update servers for airgapped environments. A reference implementation can be found at <a href="https://github.com/k0sproject/update">https://github.com/k0sproject/update</a>.</p><h3>ipset and kernel version incompatibility</h3><p>As some of you might remember, about a year ago, team k0s <a href="https://www.mirantis.com/blog/networking-problems-after-installing-kubernetes-1-25-or-after-upgrading-your-host-os-this-might-be-your-problem">uncovered</a> an iptables version incompatibility bug. Then a few weeks ago, some of our integration tests started to fail unexpectedly when running on the newer 6.2 Linux kernel. While we started to debug and triage those tests, we pretty soon realized that we’d again stumbled over something very similar to the iptables saga.</p><p>In the Linux kernel, there’s a framework called ipset which manages sets of addresses within the kernel. Those sets can be used for lightning fast matching on rules. Naturally, in the world of Kubernetes, there are a few things that interface with ipset, like kube-proxy and some CNI providers.</p><p>As with Kubernetes, and containers in general, the main philosophy is that each container comes with all the tools and libraries they need bundled within the container image. While that works pretty well when considering run-of-the-mill applications, it poses some challenges for system level things. In this particular case, kube-proxy bundles one version within the image and a CNI provider (Calico in our case) might bundle some other version. Why is this a problem? Well, in most cases it is not and should not be as the Linux kernel has very strong backwards compatibility guarantees. However, on kernels &gt;= 6.2, there’s a new bitmask parameter for ipset, which newer ipset versions support and use, but which older versions don’t know how cope with. This leads to a failure, with the error message:</p><pre>Kernel and userspace incompatible: settype hash:ip with revision 6 not supported by userspace.</pre><p>The kube-proxy image was shipping ipset v7.17, whereas the calico-node image was shipping ipset v7.15. That fails on newer kernels, if kube-proxy uses IPVS. Now rules set up with v7.17 — rules that utilize the new bitmask parameter — fail on loading by the older v7.15 version. 💥</p><p>The fix in itself is easy for k0s. We <em>just</em> (okay, harder than it looks) need to ensure that ALL system components bundle the same compatible version of ipset. But this raises the question: how can system admins ensure all low-level kernel and user space tooling IS actually compatible between all system components? The longer (and painful)answer is: by testing EVERY single update — k0s, Kubernetes upstream, Linux kernel, or other components should be rigorously tested BEFORE hitting a production environment.</p><p>The short (and painless) answer is by using a well-tested Kubernetes distro like k0s. :) We believe sysadmins should not need to worry about stuff like this.</p><p>All this brings up one other major thing the team has been working on in the scope of the 1.28 release.</p><h3>Extended OS testing matrix</h3><p>As you know, one of the promises of k0s is that it runs on ANY Linux setup. As k0s bundles in ALL the needed dependencies, there’s really nothing a cluster admin needs to set up as a prerequisite before running k0s. But how can we really be sure that is the case? With rigorous testing.</p><p>We’ve now extended our OS testing matrix to cover the following Linux distros and versions:</p><ul><li>Alpine: 3.17</li><li>Centos: 7, 8 and 9</li><li>Debian: 10, 11 and 12</li><li>Fedora CoreOS: 38</li><li>Fedora: 38</li><li>Flatcar</li><li>Oracle: 7.9, 8.7, 9.1</li><li>RHEL: 7, 8 and 9</li><li>Rocky: 8 and 9</li><li>Ubuntu: 20.04, 22.04 and 23.04</li></ul><p>Not only do we test different OS versions, but we also test different configurations of k0s. The matrix “dimensions” also cover different CNI providers and different kube-proxy modes (iptables vs. IPVS). So the total number of permutations we test is going up fast: right now, we test 84. We’re also planning to add tests for arm64, which will double the amount of permutations permutations.</p><p>For each OS and configuration combination we run the full Kubernetes conformance test suite. That gives us pretty good guarantees that everything is working on that combination as expected.</p><p>As you see, we take our promise of “runs anywhere” very seriously.</p><h3>SBOM</h3><p>Starting with the 1.28 release, we now build a full signed SBOM of k0s build artifacts. This has several benefits to the users:</p><ul><li><strong>Enhanced Security</strong>: By providing a full signed SBOM, k0s enables users to verify the authenticity of their Kubernetes distribution, reducing the risk of using tampered or malicious components.</li><li><strong>Transparency</strong>: Users gain insight into the components and dependencies included in the k0s distribution, improving transparency and aiding in vulnerability management.</li><li><strong>Compliance</strong>: SBOMs are valuable for organizations striving to meet compliance requirements, such as those outlined by regulatory bodies or internal policies.</li></ul><p>Going forward we’re also planning to sign ALL the build artifacts to further enhance the security of the build pipeline.</p><h3>User survey</h3><p>In case you’ve missed it, we launched a k0s <a href="https://medium.com/k0sproject/help-shape-the-future-of-k0s-kubernetes-participate-in-our-2023-users-survey-9549d3a72e87?source=friends_link&amp;sk=f432e56b6bee4525b45b6141ca26c8ec">user survey</a> last week. By answering this survey, you’ll help us prioritize features that mean the most to YOU, and in general, you’ll help us to shape the future of k0s.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f48e50677ab3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0s-1-28-released-f48e50677ab3">k0s 1.28 released</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Help Shape the Future of k0s Kubernetes: Participate in our 2023 Users Survey!]]></title>
            <link>https://medium.com/k0sproject/help-shape-the-future-of-k0s-kubernetes-participate-in-our-2023-users-survey-9549d3a72e87?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/9549d3a72e87</guid>
            <category><![CDATA[surveys]]></category>
            <category><![CDATA[community]]></category>
            <category><![CDATA[k0s]]></category>
            <category><![CDATA[kubernetes]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Tue, 03 Oct 2023 09:19:03 GMT</pubDate>
            <atom:updated>2023-10-03T09:22:32.145Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AD2qHgTVnW20IE0wTkSLcw.png" /></figure><p>Greetings to the k0s community!</p><p>As we continue to strive towards making k0s the best Kubernetes distribution for you, we believe in the importance of listening to your experiences, feedback, and insights. That’s why we’ve rolled out the “k0s Kubernetes Users Survey 2023,” and we’re inviting you to share your thoughts!</p><h4>Why Does Your Feedback Matter?</h4><ul><li><strong>User-Centric Development:</strong> Your insights will guide our development priorities, ensuring that k0s remain responsive to your needs.</li><li><strong>Understanding Diverse Use Cases:</strong> Every user is unique, and by understanding your specific challenges and successes, we can make k0s more versatile and robust.</li><li><strong>Building a Stronger Community:</strong> This survey is a reflection of our commitment to the k0s community. We believe that by understanding and connecting with our users, we foster a stronger, more collaborative ecosystem.</li></ul><h4>What’s in the Survey?</h4><p>The survey aims to understand the different user segments, use cases, and experiences related to k0s. We’ve kept the questions concise to respect your time, and in return, you’ll be helping to shape the direction and future of k0s.</p><h4>Ready to Share Your Thoughts?</h4><p>Participating in the survey is easy and should take only a few minutes of your time. <a href="https://forms.gle/yNoVKJJYnUUyrqCR6">Click here to start the k0s Kubernetes Users Survey 2023!</a></p><p>We’re excited to hear from you! Your voice plays a vital role in the evolution of k0s. Thank you for being a part of our journey, and for helping us ensure that k0s remains the Kubernetes distribution of choice for professionals worldwide.</p><p><strong>Warm Regards</strong>,</p><p><em>The k0s Team</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9549d3a72e87" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/help-shape-the-future-of-k0s-kubernetes-participate-in-our-2023-users-survey-9549d3a72e87">Help Shape the Future of k0s Kubernetes: Participate in our 2023 Users Survey!</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[k0smotron with full CAPI support]]></title>
            <link>https://medium.com/k0sproject/k0smotron-with-full-capi-support-65b61a571a6b?source=rss----8f45b8534c0d---4</link>
            <guid isPermaLink="false">https://medium.com/p/65b61a571a6b</guid>
            <category><![CDATA[k0s]]></category>
            <category><![CDATA[containers]]></category>
            <category><![CDATA[cluster-api]]></category>
            <category><![CDATA[kubernetes]]></category>
            <dc:creator><![CDATA[Jussi Nummelin]]></dc:creator>
            <pubDate>Wed, 30 Aug 2023 11:57:39 GMT</pubDate>
            <atom:updated>2023-08-30T11:57:39.038Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*_1uSGVDzB69fda3F" /><figcaption>Photo by <a href="https://unsplash.com/@aldebarans?utm_source=medium&amp;utm_medium=referral">Aldebaran S</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Since we’ve <a href="https://medium.com/k0sproject/introducing-k0smotron-c2ed6535ddc8">initially launched</a> k0smotron in June, team k0s has been busy with working on enhancing k0smotron. Today we’re glad to share that k0smotron now has pretty much full ClusterAPI support built-in.</p><p>In k0smotron version 0.4.0, released only couple weeks after the initial k0smotron launch, we already added the basic capability to spin up both the control and worker planes via ClusterAPI.</p><p>0.5.0, released in mid July, came with support for Template types and thus enabling you to work with MachineDeployment etc. ClusterAPI constructs. MachineDeployments allows you to easily scale your nodes as it behaves very similarly as normal workload Deployment.</p><p>Now with the latest 0.6.0 release we’re happy to complete the ClusterAPI support by introducing the “traditional” control plane provider for Machines . This allows you to run the child cluster control planes on Machines.</p><p>Like we said in the initial launch blog:</p><blockquote>we are definitely looking at Cluster API direction of having k0smotron providing the controlplanes for the clusters.</blockquote><p>So yes, we’ve been a bit more than looking at it. :)</p><h3>ClusterAPI providers</h3><p>In ClusterAPI there’s many different providers providing various “parts” of the clusters. k0smotron can now acts as the following providers:</p><h4>Controlplane provider — in-cluster</h4><p>k0smotron can acts as the Control Plane provider in two different modes. Either running the control plane <em>within</em> the management cluster, this being the main use case for k0smotron. This way you can bootstrap the control plane super fast and easy and then attach your worker plane from practically ANY infrastructure. Imagine the use cases this unlocks such as multi cloud clusters, far Edge cluster and so on. The possibilities are pretty much endless.</p><h4>Controlplane provider — out-of-cluster</h4><p>This is the “traditional” way how ClusterAPI provisions clusters. The control plane of the child cluster is configured on Machines provisioned and managed by some ClusterAPI infrastructure provider, AWS for example. In this sceneario k0smotron configures the control plane Machines to run as the control plane for the child cluster.</p><h4>Bootstrap provider</h4><p>Bootstrap provider was the very first provider we implemented on ClusterAPI. It provides the needed configuration to provision worker plane Machines via an infrastructure provider. Needless to say, in the scope of k0smotron and k0s it provides the bootstrap config to spin up k0s worker nodes connecting to the control plane provisioned via the control plane provider. It can naturally work with either way the control plane is provisioned, in-cluster or out-of-cluster.</p><h3>How can You help?</h3><p>As you can imagine, k0s team cannot practically test all the different ClusterAPI infrastructure providers out there. While k0smotron does follow the contracts outlined in ClusterAPI itself there can be some details in each infrastructure provider that does not work out of box. Hence we urge you to test k0smotron with your favourite infrastructure (cloud) provider. Check out the k0smotron <a href="https://docs.k0smotron.io/">docs</a> for <a href="https://docs.k0smotron.io/stable/cluster-api/">ClusterAPI providers </a>to get started.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=65b61a571a6b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/k0sproject/k0smotron-with-full-capi-support-65b61a571a6b">k0smotron with full CAPI support</a> was originally published in <a href="https://medium.com/k0sproject">k0s — The Kubernetes Distribution</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>