⎈ Google Cloud Shell Escape, Azure CNI Pod Subnet, How I Think About Kubernetes, Sidecar FROM 421MB to 90MB, Manage Resources as One Unit

This newsletter is brought to you by Kubex — Automated Resource Optimization for Kubernetes, GPUs and AI Workloads

---

We released a new ebook: “Right-Sizing GPUs in Kubernetes”!

Gulcan and I have been busy over the past months dissecting how to make sense of GPU monitoring and metrics.

The eBook explores in detail how:

• Kubernetes, nvidia-smi, and workload each have different views of GPU utilization.

• A 54% GPU Util is a misleading signal for GPU yield.

• Collecting the right metrics and making sense of them.

• GPU waste cascades into CPU and memory costs.

We also included several flowcharts to help you diagnose utilization and yield.

Kubex sponsored the book, and you can grab your free copy here.

— Dan

---

📚 Articles

1. 🔥 Google Cloud Shell Container Escape

2. Kubernetes Resource Management ebook for Platform Owners

3. Azure Kubernetes Service Deep Dive Into Azure CNI Pod Subnet

4. How I think about Kubernetes

5. How We Shrunk a Kubernetes Sidecar from 421MB to 90MB (With No OS Inside)

6. Kube Resource Orchestrator: Manage any group of resources as one unit

---

🌟 [FREE eBook] Right-Sizing GPUs in Kubernetes

A GPU cluster can look 100% allocated while delivering 30% yield.

This book shows you how to measure the gap and close it.

→ Grab your copy

---

📖 Tutorials

1. Data Streaming in Practice (Kafka + Flink): Building a Flight Baggage Tracking System

2. Building a Home Kubernetes Cluster with 3 Raspberry Pis

3. Scaling AI Document Processing on Kubernetes with Ray & KubeRay

4. From Chaos to 99.9% Uptime: Rebuilding a Kubernetes Platform for GPU Workloads

---

📺 This week on the KubeFM podcast

From ECS to Kubernetes: A Real Migration Story

---

💼 Kubernetes jobs

Machine Learning Engineer 💰 $120K to $145K a year · 🏢 based in the office in Charlotte, NC, USA

Software Engineer 💰 $275K to $300K a year · 🌎 remote from

Platform Engineer 💰 $200K to $235K a year · 🏢 based in the office in Toronto, CA

System Administrator 💰 $128K to $176K a year · 🌎 remote from

DevSecOps Engineer 💰 $216.68K to $269.17K a year · 🏢 based in the office in San Mateo, CA, USA

👉 Discover more opportunities on Kube Careers.

---

🛠 Tools and libraries

1. Wozz: Kubernetes Cost Tool

2. Over-Engineered GitOps Homelab

3. Purr: kubectl TUI

4. CasaVue: Ingress Catalog

5. MySQL Cluster Operator for Kubernetes

6. k8sgpt: Kubernetes analyzer

7. kseal: kubeseal CLI companion

8. Linnix: eBPF Observability & AI Incident Detection for Kubernetes

9. Terraform + Ansible for OKD/OCP on Proxmox

---

📅 Upcoming Kubernetes events

Modernize & Scale: Why and how to migrate your workloads to GKE 📅 Feb 25

Introducing KRO: Simple Custom APIs for Kubernetes 📅 Feb 26

🔥 KubeFest 2026 📅 Feb 28

🔥 Kubernetes Community Days Guadalajara 2026 📅 Feb 28

From Code to Cloud: An End-to-End Journey 📅 Mar 4

🔥 Make autoscaling boring by improving how HPA, VPA & Karpenter make decisions 📅 Mar 5

🔥 Advanced Kubernetes course 📅 Mar 12

👉 You can find more events on Kube Events.

---

📢 Call for papers closing soon

🔥 Container Days Hamburg 2026 ⏳ closes Feb 28

Cloud Native Days Italy 2026 ⏳ closes Mar 6

Cloud Native Days Amsterdam ⏳ closes Mar 20

🔥 Kubernetes Community Days Panama 2026 ⏳ closes Mar 3

Kubernetes Community Days Kochi 2026 ⏳ closes Mar 1

🔥 Kubernetes Community Days Czech & Slovak - Prague 2026 ⏳ closes Apr 1

🔥 Kubernetes Community Days Kuala Lumpur 2026 ⏳ closes Feb 28

KubeCon + CloudNativeCon Japan 2026 ⏳ closes Mar 29

Kubernetes Community Days Istanbul 2026 ⏳ closes Mar 2

👉 You can find more Call for Papers on Kube Events.

---

---

Until next time!

— Gulcan