<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>litek8s blog</title><description>Guides and notes about litek8s, a lightweight Kubernetes in one binary.</description><link>https://litek8s.dev/</link><language>en</language><item><title>The pod that waits: how Pending works, and the deadline that never passed</title><link>https://litek8s.dev/blog/the-pod-that-waits/</link><guid isPermaLink="true">https://litek8s.dev/blog/the-pod-that-waits/</guid><description>What a Pending pod means, how the scheduler decides whether it fits, and the bug that let a Job with a deadline wait for ever.</description><pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate><category>scheduling</category><category>jobs</category><category>controllers</category><category>pending</category></item><item><title>kubectl get 2.4× faster: the copy hidden in json!</title><link>https://litek8s.dev/blog/the-copy-hidden-in-json/</link><guid isPermaLink="true">https://litek8s.dev/blog/the-copy-hidden-in-json/</guid><description>A profile of kubectl get&apos;s Table showed half the time going to copying values. The cause was one serde_json macro, and the fix was moving instead.</description><pubDate>Sat, 31 Oct 2026 00:00:00 GMT</pubDate><category>performance</category><category>profiling</category><category>api-server</category><category>rust</category></item><item><title>Giving an LLM agent a Kubernetes cluster without fear</title><link>https://litek8s.dev/blog/giving-an-llm-agent-a-kubernetes-cluster/</link><guid isPermaLink="true">https://litek8s.dev/blog/giving-an-llm-agent-a-kubernetes-cluster/</guid><description>litek8s mcp gives an agent diagnose_pod, logs, PromQL and GitOps tools, and refuses clusters that are not litek8s. The guardrails, and what they don&apos;t cover.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>llm</category><category>agents</category><category>mcp</category><category>kubernetes</category><category>safety</category></item><item><title>A home lab across NATs: one cluster from a desktop, a cloud VM and a friend&apos;s machine</title><link>https://litek8s.dev/blog/home-lab-across-nats/</link><guid isPermaLink="true">https://litek8s.dev/blog/home-lab-across-nats/</guid><description>Three machines behind different NATs, one litek8s cluster. How the WireGuard overlay finds direct paths, when it falls back to a relay, and what that costs.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>home lab</category><category>networking</category><category>wireguard</category><category>nat</category><category>multi-node</category></item><item><title>We ran the official Kubernetes conformance suite on litek8s: what passes, what fails, and why</title><link>https://litek8s.dev/blog/kubernetes-conformance-what-fails-and-why/</link><guid isPermaLink="true">https://litek8s.dev/blog/kubernetes-conformance-what-fails-and-why/</guid><description>413 of 419 conformance specs, against 416 for k3s. The six failures, the first run that scored zero, and what each one says about Kubernetes.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>conformance</category><category>kubernetes</category><category>testing</category><category>compatibility</category></item><item><title>Kubernetes in CI in ~2.5 s instead of ~45 s: litek8s against kind</title><link>https://litek8s.dev/blog/kubernetes-in-ci-in-seconds/</link><guid isPermaLink="true">https://litek8s.dev/blog/kubernetes-in-ci-in-seconds/</guid><description>A one-line change in a GitHub Actions workflow: a cluster Ready in ~2.5 s instead of ~45 s, Helm jobs 3.7x shorter. Measured, with where it does not pay off.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>ci</category><category>github actions</category><category>gitlab</category><category>kind</category><category>helm</category><category>performance</category></item><item><title>Metrics, logs and alerts without kube-prometheus-stack and Loki: 1,371 MiB to 1.8 MiB</title><link>https://litek8s.dev/blog/observability-without-kube-prometheus-stack-and-loki/</link><guid isPermaLink="true">https://litek8s.dev/blog/observability-without-kube-prometheus-stack-and-loki/</guid><description>On a small k3s node the usual stack added 1,371 MiB. litek8s&apos;s built-in metrics, logs and alerts added 1.8 MiB. What that buys, and what it leaves out.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>observability</category><category>prometheus</category><category>loki</category><category>alerts</category><category>memory</category></item><item><title>October 2026 in litek8s: 21 releases, RBAC, NetworkPolicies and a UI for any cluster</title><link>https://litek8s.dev/blog/october-2026-in-litek8s/</link><guid isPermaLink="true">https://litek8s.dev/blog/october-2026-in-litek8s/</guid><description>What shipped in litek8s in October 2026: RBAC and NetworkPolicies enforced, a GitHub Action for CI, a terminal UI for any cluster, and litecd for GitOps.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>changelog</category><category>release notes</category><category>build in public</category></item><item><title>One binary instead of Docker Desktop, and it is also Kubernetes</title><link>https://litek8s.dev/blog/one-binary-instead-of-docker-desktop/</link><guid isPermaLink="true">https://litek8s.dev/blog/one-binary-instead-of-docker-desktop/</guid><description>litek8s dev up serves the Docker Engine API: docker, compose and Testcontainers work unchanged, and kubectl and Helm use the same cluster. What is missing, too.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>docker</category><category>compose</category><category>testcontainers</category><category>local development</category><category>macos</category><category>linux</category></item><item><title>One LLM across three 4 GiB machines: you get memory, not speed</title><link>https://litek8s.dev/blog/one-llm-across-three-machines/</link><guid isPermaLink="true">https://litek8s.dev/blog/one-llm-across-three-machines/</guid><description>Qwen2.5 7B needs 4.4 GiB of weights. No node had that. Split over three with llama.cpp&apos;s RPC backend on litek8s, it ran at 7.5 tokens/s.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>llm</category><category>llama.cpp</category><category>home lab</category><category>multi-node</category></item><item><title>Replacing Argo CD&apos;s controller with litecd: 666 MiB to 100, and what it cost</title><link>https://litek8s.dev/blog/replacing-argo-cd-with-litecd/</link><guid isPermaLink="true">https://litek8s.dev/blog/replacing-argo-cd-with-litecd/</guid><description>How we swapped Argo CD&apos;s application controller for litecd on a production k3s cluster: the plan, the traps, the numbers, and what we gave up.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>gitops</category><category>argo-cd</category><category>litecd</category><category>memory</category><category>migration</category></item><item><title>Small is not enough: measuring litek8s&apos;s speed, and a 13× faster list</title><link>https://litek8s.dev/blog/small-is-not-enough/</link><guid isPermaLink="true">https://litek8s.dev/blog/small-is-not-enough/</guid><description>We had numbers for memory and disk, none for speed. A load test for the API server, a list made 13× faster, and how CI now catches regressions.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>performance</category><category>benchmarks</category><category>api-server</category><category>ci</category><category>rust</category></item><item><title>The node that was too full: why Kubernetes evicted our database</title><link>https://litek8s.dev/blog/the-node-that-was-too-full/</link><guid isPermaLink="true">https://litek8s.dev/blog/the-node-that-was-too-full/</guid><description>One node at 97% memory, a database evicted twice, and a scheduler that saw room. Requests against usage, and how we moved pods without downtime.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>scheduling</category><category>memory</category><category>eviction</category><category>k3s</category><category>operations</category></item><item><title>Cloning a production cluster onto litek8s: what Kubernetes taught us</title><link>https://litek8s.dev/blog/cloning-a-production-cluster/</link><guid isPermaLink="true">https://litek8s.dev/blog/cloning-a-production-cluster/</guid><description>We copied a running production cluster onto litek8s without touching it. Four bugs, two missing pieces of Kubernetes, and what each says about how it works.</description><pubDate>Thu, 29 Oct 2026 00:00:00 GMT</pubDate><category>migration</category><category>rbac</category><category>networkpolicy</category><category>gitops</category></item><item><title>You don&apos;t need a space shuttle to ship your app</title><link>https://litek8s.dev/blog/you-dont-need-a-space-shuttle/</link><guid isPermaLink="true">https://litek8s.dev/blog/you-dont-need-a-space-shuttle/</guid><description>Why litek8s is so fast, what that speed costs, and why starting small on the real Kubernetes API lets you migrate later without rewriting anything.</description><pubDate>Thu, 29 Oct 2026 00:00:00 GMT</pubDate><category>opinion</category><category>performance</category><category>small teams</category></item><item><title>Kubernetes on a Raspberry Pi and an old laptop with litek8s</title><link>https://litek8s.dev/blog/raspberry-pi-kubernetes-cluster/</link><guid isPermaLink="true">https://litek8s.dev/blog/raspberry-pi-kubernetes-cluster/</guid><description>Turn a Raspberry Pi, an old laptop and a cheap cloud VM into one Kubernetes cluster with litek8s: one binary per machine, installed over SSH from your laptop.</description><pubDate>Wed, 28 Oct 2026 00:00:00 GMT</pubDate><category>home lab</category><category>raspberry pi</category><category>multi-node</category></item></channel></rss>