LearnKube news

This tutorial walks through the llm-d quickstart on Kubernetes, pausing after each step to explain what you just deployed.

More: https://ku.bz/-29f-Q-2W

Sai Vennam has been in the Kubernetes world for the better part of 10 years. His prediction: operators will spend less and less time on the heavy lifting of cluster management as AI agents and LLMs streamline operations — shifting teams toward an absentee-operator model where the focus returns to business logic.



Watch the full interview: https://ku.bz/VVfQ4G891

Rune is a fast native macOS app for people who debug clusters every day, with unified pod logs, YAML editing with server dry-run, port-forwarding and a command palette.

More: https://ku.bz/zRfTn2ZwJ

More AI. That's Itiel Shwartz's prediction for Kubernetes in the next 12 months.

Cloud providers, legacy players like Datadog and Dynatrace, and newcomers like Komodor are all moving in the same direction: AI-powered operations. AWS already offers AI SRE capabilities, and Azure and Google are following. The trend is clear, even if not everyone delivering on the promise is the real deal.

The industry is shifting. The question is who actually delivers.





Watch the full interview: https://ku.bz/b9bDXQ_xq

We just published Kubernetes Architecture in Financial Services, a free technical book about platform design and operations.

Drawing on public engineering talks and case studies from seven banks, it examines:

- Tenant boundaries beyond namespaces.
- Shared delivery controls and application ownership.
- Policy maintenance, exceptions, adoption, and reporting.
- Cross-cluster identity, reliability, external dependencies, and cluster replacement.

Each chapter starts with a real platform problem, compares different boundaries, and examines the operational cost of each option.

Thank you to Buoyant, Sysdig, and Nirmata for supporting the book and its research.

Download it for free:
https://learnkube.com/kubernetes-architecture-financial-services

Kor is a CLI tool that scans your cluster and lists the unused resources nobody cleaned up, from forgotten ConfigMaps and Secrets to stale Services and PVCs.

More: https://ku.bz/YzcpnL3gj

This article explains how the Mixed Version Proxy reached beta in Kubernetes 1.36, so during an upgrade an older API server forwards requests it does not know to a newer peer instead of answering 404.

More: https://ku.bz/SfbjnMQZY

Owumi Festus breaks down what Kubernetes actually does at its fundamental level. He explains the reconciliation loop - the heart of Kubernetes - as a continuous process that moves systems from their current state to a desired state.

Using a thermostat analogy, Owumi illustrates how Kubernetes works. Just as a thermostat monitors room temperature and adjusts heating/cooling to reach your target temperature, Kubernetes continuously monitors your cluster and triggers events to achieve your specified configuration. When you declare "I want this number of pods, this number of deployments," Kubernetes responds by orchestrating the necessary changes to make that declaration a reality.

Watch the full episode: https://ku.bz/pf5kK9lQF

mcp-server-kubernetes exposes a complete Kubernetes management layer via Model Context Protocol (MCP), letting tools like Claude Desktop and mcp-chat run kubectl and Helm commands securely.

More: https://ku.bz/J-BnR0D6V

Kubernetes problems often hide in controller timing, implicit defaults, and capacity assumptions.

This week in Learn Kubernetes Weekly:

🔎 A production race condition left orphaned pods blocking new deployments.
📦 Source Hydrated Infrastructure Models commit rendered manifests for explicit GitOps audits.
🖥 Headlamp replaces the archived Kubernetes Dashboard and maps familiar workflows to a maintained UI.
⚙️ Kubernetes 1.36 adds pod-level CPU and memory management.
📊 The k8s-overcommit operator reclaims idle capacity according to priority classes.

Read issue 201: https://kube.today/issues/201

This issue is brought to you by LearnKube — understand how Kubernetes works, and what to do when it breaks. Live training with 60% hands-on labs: https://ku.bz/hypSbyc-V

This article explains how Kubernetes 1.36 lets the kubelet assign CPU and memory at the pod level, so the main container keeps exclusive NUMA-aligned cores while sidecars share the rest.

More: https://ku.bz/zXsfljslJ

A pizza order became a 15-container, 200-trace system.

In Kube Signals episode two, Brian Teller and Mauricio (Salaboy) Salatino examine observability, state management, governance, and trust for non-deterministic agents.

They discuss:

- Why multi-agent runtimes can recreate monolith scaling
- How OpenTelemetry exposes prompts, completions, and decisions
- Why execution history turns agent trust into measurable evidence

Watch: https://ku.bz/TlVjXdnb6

Kubernetes moves too fast to track everything. Learn Kubernetes Weekly filters out the noise to deliver one curated email with useful articles, tutorials, tools, jobs, events, and CFPs. Subscribe to Learn Kubernetes Weekly.

Kubernetes turns 10, and the next decade looks very different from the last.

Greg Eppel predicts that agentic AI will drive a fundamental simplification of Kubernetes operations — more autonomous actions taken on behalf of the agent, with humans brought into the loop only when absolutely necessary.



Watch the full interview: https://ku.bz/KZRgrbmwt

kubectl tree is a kubectl plugin that follows ownerReferences and prints the whole object tree under a Deployment or custom resource, so you can see what created what.

More: https://ku.bz/D2qCz0wDD

This article explains what changes when you move from the archived Kubernetes Dashboard to Headlamp, mapping each familiar workflow onto the new UI.

More: https://ku.bz/v8jGN5MqH

"You always know you're going to have the capacity needed to meet the demand of your cluster."

With provisioned control planes on Amazon Web Services (AWS) EKS, you set a baseline for the control plane infrastructure so it's ready before demand hits. Alex Kestner explains this is essential for teams running large numbers of nodes or busy clusters with high churn.





Watch the full interview: https://ku.bz/blMzG_XM7

Point-in-time recovery sounds simple until you're running Patroni clusters.

Ziv Yatzik explains the challenge: every WAL file needs to go to separate storage, and after a switchover, you need to keep collecting from the new primary without breaking the timeline.

Most tools they tested added complexity or risked database availability. Their solution: pgBackRest integrated directly with Patroni's archive command.



Watch the full episode: https://ku.bz/Rg_sQYSmw

kpf is a terminal UI around kubectl port-forward that lets you pick a service from a list and reconnects automatically when a pod restarts or your laptop wakes up.

More: https://ku.bz/SnMTxDWh8

This case study walks through a production debugging session where orphaned Kubernetes pods kept blocking new deployments due to a race condition between the job controller, garbage collector, and an undocumented default propagation policy behavior.

More: https://ku.bz/NmswYT7Zd

200 issues. More than 77,000 Kubernetes engineers. Thank you to all of you reading it 🎉

In Learn Kubernetes Weekly 200:

🤖 Benchmarking LLM Inference with Production Agent Traces
🔌 Two New Headlamp Plugins
🌐 Migrating from ingress-NGINX to Envoy Gateway
🐘 Building High-Availability PostgreSQL on Kubernetes

Read it: https://kube.today/issues/200