LearnKube news

This tutorial shows how to connect two Amazon EKS clusters with an Istio multi-primary, multi-network mesh and view both through centralized Kiali. More: https://ku.bz/CSr7L2PvB

Cortney Nickerson, Head of Community @ Nirmata, shares her playbook for building credibility in the Kubernetes ecosystem without being a maintainer. She outlines a strategy centered on vulnerability and consistent effort - telling experts what you don't know while demonstrating you've done background research first. Cortney draws parallels between her softball background and tech learning, emphasizing how athletic discipline translates to sustainable learning practices. She advocates for setting smaller, attainable goals and maintaining daily consistency rather than trying to absorb everything at once. The discussion covers practical tactics like being transparent about your learning journey, making effort visible to the community, and developing the mindset that you're capable of doing hard things.

Changing proxies without revisiting telemetry can multiply observability costs. Henrik Rexed explains how an NGINX-to-Envoy migration produced metrics at every request and created extreme volume and cardinality. His recommendation is to validate telemetry during development with tools such as OpenTelemetry Weaver, before production traffic magnifies the cost. Treat observability configuration like software and validate it before release.

K9s is a terminal UI for Kubernetes that keeps watching your cluster and lets you navigate, inspect and manage resources without typing long kubectl commands. More: https://ku.bz/SwRRcNYC2

Thibault Jamet, DevOps and Technical Product Owner @ Adevinta, explains how his team debugged a critical Vertical Pod Autoscaler incident by diving directly into the VPA source code. When their VPA started continuously evicting pods across their 30-cluster platform, traditional monitoring wasn't revealing the root cause. The breakthrough came when they pulled down the VPA recommender source code and ran the updater component in debug mode to understand why it was making eviction decisions. They discovered that pods were missing crucial annotations that the mutating webhook should have injected. Watch the full episode: https://ku.bz/rf1pbWXdN

Hannah Foxwell, Co-founder at BIMP, is one of the featured lightning-talk speakers at LearnKube Day Salt Lake City 2026. Join the community for YAML Games and a reception with live DJ Amin Astaneh. Learn more: https://learnkube.com/learnkube-day-salt-lake-city-2026

LearnKube Day brings Kubernetes practitioners together for five-minute lightning talks and practical lessons from the community. Andy Suderman, CTO at Fairwinds, is one of the featured lightning-talk speakers at LearnKube Day Salt Lake City 2026. Join the community for YAML Games and a reception with live DJ Amin Astaneh. Learn more: https://learnkube.com/learnkube-day-salt-lake-city-2026

SKA OCI Daemon primes OCI registries and compute nodes with container artifacts to reduce image pull times through efficient caching and multiple authentication methods. More: https://ku.bz/tfRG5hnNr

LearnKube Day brings Kubernetes practitioners together for five-minute lightning talks and practical lessons from the community. Nabarun Pal, Principal Software Engineer at Broadcom, is one of the featured lightning-talk speakers at LearnKube Day Salt Lake City 2026. Join the community for YAML Games and a reception with live DJ Amin Astaneh. Learn more: https://learnkube.com/learnkube-day-salt-lake-city-2026

This week on Learn Kubernetes Weekly 208: ⚖️ Go, Node.js, C#, Java, and Python on Kubernetes 🗃️ Why Kubernetes depends on etcd and Raft 🧠 JVM memory: high usage does not always mean a leak ⏳ Dagster bottlenecks that more nodes cannot fix 🔀 Apache Stateful Functions on Flink 2.x and Java 21 Read it: https://kube.today/issues/208 ⭐️ This newsletter 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 argues that vLLM in a pod is only the engine, and maps the layers a real LLM platform needs around it: - tenant keys and budgets, - inference-aware routing, - and GPU scheduling. More: https://ku.bz/mlftRMBZy

chaossnake is a Go terminal game that deletes real Kubernetes pods as you play Snake. Each food is a pod, and movement triggers kubectl delete. More: https://ku.bz/ZFQkGyGh5

VolSync replicates Kubernetes persistent volumes between clusters with rsync or rclone, and can also back them up to object storage using restic. More: https://ku.bz/j53CrZXYy

More management APIs will move into Kubernetes. More GenAI and agentic tools will manage clusters autonomously. Sebastien Allamand at AWS on what the next decade holds: agents that work in Kubernetes, managing applications on the fly — not replacing engineers, but handling the operational toil. Watch the full interview: https://ku.bz/hGZChYmJ7

This article explains why a JVM pod can sit at 90% memory and still be perfectly healthy, and how to compare pod memory with real heap usage before hunting for a leak. More: https://ku.bz/MQcshv34f

LearnKube Day brings Kubernetes practitioners together for five-minute lightning talks and practical lessons from the community. Miriah Peterson, CEO and Co-founder at Haikei Labs, is one of the featured lightning-talk speakers at LearnKube Day Salt Lake City 2026. Join the community for YAML Games and a reception with live DJ Amin Astaneh. Learn more: https://learnkube.com/learnkube-day-salt-lake-city-2026

Ray Chen spent years trying to solve resource optimization at Trumid before adopting a dedicated tool. Cost dashboards showed where the waste was but didn't fix it. Grafana dashboards helped the platform team tune workloads, but they lacked the context of what each service needed. Homegrown scripts mimicked the recommendations but still required manual validation. The missing piece was never visibility — it was sustainable automation that fit how teams already worked. StormForge partnered with Trumid to build that: continuous monitoring, JVM-aware recommendations, PR-based approvals, and team-level tagging so the right people review the right changes. Watch the full interview: https://ku.bz/w50VfwtYd

burn is a CLI that shows where your Kubernetes cluster is spending money on compute, storage, load balancers and GPUs, and it can check those estimates against your real AWS or Azure bill. More: https://ku.bz/K1bNxz62Y

LearnKube Day brings Kubernetes practitioners together for five-minute lightning talks and practical lessons from the community. Artem Lajko, Head of Platform Engineering at iits-consulting, is one of the featured lightning-talk speakers at LearnKube Day Salt Lake City 2026. Join the community for YAML Games and a reception with live DJ Amin Astaneh. Learn more: https://lnkd.in/gzwrx3mx

This article explains why Kubernetes keeps all its truth in etcd, and how the Raft algorithm gets several servers to agree on one version of reality before your apply returns. More: https://ku.bz/k5GlnB2l3