This article explains how CoreDNS works in OpenShift clusters, covering DNS resolution architecture where the 10th IP from service network is reserved for CoreDNS service.

More: https://ku.bz/wY0TJ3X8N

Zain Malik, Software @ Exostellar, explains why his team moved away from their Terraform-based cluster management approach.

He details how their small team faced significant challenges, including lengthy provisioning times (1-1.5 weeks), fragmented tooling between Terraform and node pool operators, and synchronization issues between Terraform state and cloud resources. When privileged users made direct changes to cloud resources, the team would only discover these during subsequent Terraform runs, creating a firefighting situation where engineers had to fix state inconsistencies and complete their original tasks. These limitations put considerable strain on their small team and delayed cloud migrations.

Watch the full episode: https://ku.bz/5PLksqVlk

Lynq Operator automates Kubernetes resource provisioning directly from database rows using Go templates and Sprig functions.

More: https://ku.bz/hpx6WyM4K

This week's 6 Kubernetes junior jobs are:

Software Engineer with Anduril Industries
💰 $129K to $193K a year
👨‍💻 Remote from
→ https://ku.bz/wpQ0JZMs7

Software Engineer with Attentive
💰 $140K to $170K a year
🏠🏃🏻‍♂️🌎 New York, NY, USA
→ https://ku.bz/1CBhZXgh8

Software Engineer with Palantir Technologies
💰 $145K to $155K a year
🏠🏃🏻‍♂️🌎 Seattle, WA, USA
→ https://ku.bz/Hj-fF6QJJ

Software Engineer with Verkada
💰 $125K to $140K a year
🏠 From the office in San Mateo, CA, USA
→ https://ku.bz/v7s0J-R-l

Software Engineer with Trace3
💰 $110K to $140K a year
🏠 From the office in Colorado Springs, CO, USA
→ https://ku.bz/c1csBGWY9

👉 Browse 1619 jobs on Kube Careers https://kube.careers

Most teams default to CPU and memory for autoscaling — but those metrics rarely tell you what actually matters.

Salman Iqbal's approach: skip CPU and memory entirely. Instead, scale based on requests per minute or messages in the queue — metrics that reflect real business demand.

Tools like Crossplane and KEDA are making this easier. The key is asking: what does the business actually need to scale on?



Watch the full interview: https://ku.bz/_hMPY1FGj

This interview is a reaction to Jorge Turrado's episode https://ku.bz/Q9yNhd7L2

This article explains the differences between kubectl proxy and kubectl port-forward for accessing Kubernetes services from your local machine.

More: https://ku.bz/nNPZ6f_Zb

This week's 6 best Kubernetes Platform Engineer jobs are:

Platform Engineer with Anthropic
💰 $32.66M to $44.23M a year
🏠 From the office in San Francisco, CA, USA
→ https://ku.bz/Z_tzW6FLQ

Platform Engineer with PlayStation Global
💰 $19.33M to $28.99M a year
👨‍💻 Remote from
→ https://ku.bz/y3xyztC9_

Platform Engineer with Zscaler
💰 $15.4M to $22M a year
🏠 From the office in San Jose, CA, USA
→ https://ku.bz/JhK-ktnV1

Platform Engineer with ClickHouse
💰 $14.1M to $23M a year
👨‍💻 Remote from the United States of America
→ https://ku.bz/2X8VqBpCy

👉 Browse 1874 jobs on Kube Careers https://kube.careers

Tibo on why Kubernetes isn't just for enterprise scale — it can be a practical choice for solo self-hosters too.

You will learn:

- Why Ansible's declarative promise fell short with the Podman collection, forcing sequential imperative steps instead of desired-state definitions
- How community Helm charts replace the need to write and maintain every manifest yourself
- Why GitOps isn't just a deployment workflow — it's a disaster recovery strategy when your infrastructure lives in your living room

Watch (or listen to) it here: https://ku.bz/Xk5S7VqXz

🌟 This episode is brought to you by LearnKube — get started on your Kubernetes journey through comprehensive online, in-person, or remote training https://learnkube.com/training

With @Birthmarkb

David Pech, Staff Cloud Ops Engineer, explains how to evaluate Postgres operators beyond basic deployments.

He outlines a practical evaluation framework starting with basic CRD capabilities and progressing to advanced scenarios like database cloning and anonymization. The discussion covers three key areas: high availability (measuring recovery times and SLOs), security features (TLS certificates, credential management), and operational capabilities (backup, restore, failover scenarios).

Watch the full episode: https://ku.bz/rGMF2ktdb

Valkey Operator is a Kubernetes operator that automates deployment and lifecycle management of Valkey clusters and instances with features like automated installation and configuration management.

More: https://ku.bz/WbXvSkFMG

GPUs are expensive. And if you're not tracking yield, you're burning money.

Andrew Hillier from Kubex explains the problem: most teams running inference workloads have GPUs that are barely utilized — but they don't have visibility into it. Even when they do, the GPU and workload complexity makes it unclear what to actually fix.

Their new e-book breaks down practical strategies for optimizing GPU utilization in Kubernetes without sacrificing performance.





Get the ebook: https://ku.bz/KL4jRvsL4

This week on Learn Kubernetes Weekly 170:

📦 Could lockfiles just be SBOMs?
🌐 Dynamic Istio Ingress Gateway Management with Kyverno
🎮 Factorio in Kubernetes? Well, why not?
🤖 Running DeepSeek Models on Kubernetes: A Backend Engineer's Experiment
⚡ Ephemeral Infrastructure: Why Short-Lived is a Good Thing

Read it now: https://kube.today/issues/170

⭐️ This issue is brought to you by vCluster and LearnKube — join "Multi-Tenancy March" starting Feb 24: a free 3-part hands-on series on namespace isolation, virtual clusters, GPU sharing, and AI agent sandboxing on Kubernetes https://ku.bz/multitenant26

We're giving away 10 tickets to DeveloperWeek.

The world's largest independent dev conference: 8 events in one: AI, Cloud Native, DevOps, Frontend, Security. 250+ speakers, workshops, and a hackathon.

📅 18-20 Feb
📍 San Jose, CA + Virtual

Get in touch at hello@kube.events to claim yours.

→ https://www.developerweek.com/

This article explains how OpenTelemetry unifies observability by correlating metrics, logs, and traces through standardized telemetry data collection in Kubernetes environments.

More: https://ku.bz/8p11g8gy6

This case study shows how Codefresh optimized Kubernetes for CI workloads by creating a custom scheduler that packs pods tightly and uses ballast pods to eliminate build start delays.

More: https://ku.bz/mRdHKYC6p

Dokku is a Docker-powered PaaS that helps you build and manage the lifecycle of applications, essentially offering a mini-Heroku you can self-host.

More: https://ku.bz/t0Qb2SJy9

Mac Chaffee explains his viral blog post "Dear Friend, You Have Built a Kubernetes" - a satirical letter that follows developers who stubbornly reject Kubernetes while deploying a simple application, only to end up rebuilding it piece by piece.

Mac explores the psychology behind this common pattern in software engineering, where teams reject established complex tools without fully understanding the problems they solve, and how the deliberately smug tone resonates with developers who recognize this anti-pattern in their own experience.

Watch the full episode: https://ku.bz/9nFPmG85f

This tutorial teaches how to generate kubeconfig files for Kubernetes service accounts using a simple Go tool that works with both modern clusters (1.24+) and legacy versions.

More: https://ku.bz/pqyC53xhH

This week's 6 Kubernetes junior jobs are:

Software Engineer with Galileo Financial Technologies
💰 $105.6K to $198K a year
👨‍💻 Remote from
→ https://ku.bz/ZrZddDMnS

Software Engineer with Palantir Technologies
💰 $145K to $155K a year
🏠🏃🏻‍♂️🌎 Seattle, WA, USA
→ https://ku.bz/Hj-fF6QJJ

Software Engineer with T-Rex Solutions
💰 $130K to $160K a year
🏠 From the office in Ft. Meade, MD, USA
→ https://ku.bz/C_r8Lfkk9

Software Engineer with Verkada
💰 $125K to $140K a year
🏠 From the office in San Mateo, CA, USA
→ https://ku.bz/v7s0J-R-l

Software Engineer with Trace3
💰 $110K to $140K a year
🏠 From the office in Colorado Springs, CO, USA
→ https://ku.bz/c1csBGWY9

👉 Browse 1997 jobs on Kube Careers https://kube.careers

The Kubernetes community isn't waiting for AI vendors to self-regulate. It's setting the rules itself.

Mario Fahlandt explains the Kubernetes AI Conformance Program, announced at KubeCon North America: a standard that defines what platforms must enforce — like pod isolation, resource boundaries, and hardware access controls — before they can claim AI readiness.

If you're evaluating AI platforms for Kubernetes, check whether they comply. The guardrails are already built into the ecosystem.



Watch the full interview: https://ku.bz/36wSy63cC

This interview is a reaction to Mai Nishitani's episode https://ku.bz/3hWvQjXxp