Learn Kubernetes Weekly issue 198 · 26 Aug 2026
Data Lakehouse Infrastructure, Kubernetes Tool Trends, Kafka Performance Lessons, CODECO Architecture, AI Deleted Wrong Deployment
This newsletter is brought to you by LearnKube — master Kubernetes with hands-on training designed for engineers who want to learn the smart way.
We published a free, 342-page technical guide to Kubernetes rightsizing.
The guide covers:
- How Kubernetes requests and limits work.
- Which container, runtime, and application metrics you can trust.
- How to apply changes safely with policy, GitOps, rollouts, and rollback conditions.
- Where AI can help without receiving unchecked production authority.
This is by far the longest book I've written with Gulcan! There's so much to say about requests/limits 😅
— Dan
Articles
Data Lakehouse: Infrastructure
medium.com
This case study walks through how Fresha built a scalable AWS-backed Data Lakehouse using EKS, Flink and StarRocks, bringing real-time streams and batch analytics together in one platform.
What the Popularity of Emerging Tools Tells Us About Kubernetes' Future
kube.today
Drawing on more than 100 KubeFM conversations, this article shows how tools like Karpenter, KRO, Argo CD, OpenTelemetry, Kueue, and Kgateway are shaping autonomous infrastructure, platform engineering, and AI workloads.
sponsored
Kafka on Kubernetes: Performance Lessons for Any Disk-Heavy Data Service
dev.to
This case study traces why Kafka started reading from disk after moving from EC2 to EKS, and lands on cgroup v2 reclaim pressure.
medium.com
This article explains the centralise-or-not trade-off behind CODECO's federated Kubernetes architecture across cloud, edge and IoT, and how neighbourhood-scoped control, decentralised AI and federated scheduling answer it.
Your AI just deleted the wrong deployment. Now what?
medium.com
This article argues that MCP servers for Kubernetes ship the accelerator without the brake, and walks through a two-tier design in which an agent can only propose a change, and a human approves the exact plan before it is applied.
My Experiments with MCP: Moving Beyond the “Agent Wrapper”
godfreym.medium.com
This article presents a proof of concept in which an Elasticsearch MCP server runs on Kubernetes so an SRE agent can search logs in plain language, and it ends with the authentication and RBAC work still needed for production.
Articles worth checking out:
The Technical Guide to Kubernetes Rightsizing
Guessing at CPU and memory limits again? We've all been there.
This guide walks through the whole rightsizing journey, from messy metrics to safe production changes, and shows where AI actually helps (and where it doesn't).
Tutorials
Building a Real k6 Test Suite Against a Live Kubernetes App
dev.to
This tutorial builds a real k6 test suite against Google's Online Boutique running on a home lab Kubernetes cluster, with a shared client library, smoke, load, stress, and browser tests, and the two bugs the first run exposed.
From Ingress to Gateway API: How We Modernized Networking on Our GKE Cluster
medium.com
This tutorial explains why Gateway API replaces Ingress on GKE, walking through GatewayClass, Gateway, and HTTPRoute, then sets up traffic splitting and TLS without vendor-specific annotations.
Deploying WSO2 API Platform Gateway on OpenShift: A Practical Guide
medium.com
This tutorial walks through deploying the WSO2 API Platform Gateway on OpenShift, from building the images and installing cert-manager to applying the gateway operator, a sample REST API, and the route that exposes it.
Autoscalable GitLab runners on AWS EC2
palark.com
This tutorial shows how to replace always-on EC2 runners with autoscaling GitLab runners using the AWS fleeting plugin and an Auto Scaling Group, so instances start per job and shut down when the queue is empty.
Kube Signals starts where the keynote ends: with the trends that platform teams will have to operationalize next.
In this special episode, Brian Teller speaks with Saiyam Pathak about his KubeCon India keynote and the shift from developer platforms to AI factories. They examine what GPU scarcity, shared accelerators, and AI workloads mean after the conference slides meet real infrastructure.
In this interview:
- Why GPU infrastructure is becoming a platform-engineering concern
- How DRA, HAMI, MIG, and MPS change GPU allocation and utilization
- Where isolation, scheduling, and observability become harder for AI platforms
- Which cloud-native AI trends and projects platform engineers need to watch
Kubernetes jobs
Platform Engineer with Roblox
Salary: $293.8K to $343.34K a year
Location: based in the office in San Mateo, CA, USA
Tech stack: Kubernetes, Kubernetes, Nomad, observability, Aurora, RocksDB, AlloyDB, Spanner, PostgreSQL, MySQL
Data Engineer with Roblox
Salary: $76.5K to $275K a year
Location: based in the office in San Mateo, CA, USA
Tech stack: Kubernetes, Kubernetes, AWS, GCP, Trino, Flink, data catalog, Kafka, Airflow, Druid
Platform Engineer with Roblox
Salary: $345.04K to $399.42K a year
Location: based in the office in San Mateo, CA, USA
Tech stack: Kubernetes, On-Prem, Kubernetes, Public Cloud, observability, identity providers, security, Cloud Infrastructure, etcd, custom controllers
Software Engineer with Roblox
Salary: $243.29K to $295.25K a year
Location: based in the office in San Mateo, CA, USA
Tech stack: Kubernetes, Kubernetes, Nomad, Kibana, alerting, monitoring, Alertmanager, Grafana, Prometheus, Consul
Platform Engineer with Roblox
Salary: $295.25K to $345.04K a year
Location: based in the office in San Mateo, CA, USA
Tech stack: Kubernetes, Kubernetes, AWS, GCP, Hive, Trino, Flink, data catalog, Kafka, Airflow
Discover more Kubernetes jobs on Kube Careers →
Code & tools
KubeElasti: scale services to zero without dropping requests
github.com/kubeelasti
KubeElasti scales idle Kubernetes services down to zero and wakes them on the first request, holding that traffic in a proxy queue so nothing is lost while the pods start.
Kubernetes IPsec Connection Operator
github.com/dialohq
IPMan is a Kubernetes operator that automates IPSec VPN setup (via StrongSwan) so your workloads can securely connect across networks.
github.com/tonedefdev
OpenDepot is a self-hosted, Kubernetes-native registry for your OpenTofu/Terraform modules and providers, so you control distribution and versions rather than relying on the public registry.
github.com/selvarajmurugesan90
Klarity is a read-only Kubernetes dashboard for GitOps teams: it shows everything running in the cluster but never lets anyone change it outside Git.
kstack: Claude Code skill pack for Kubernetes operations
github.com/kubetail-org
kstack is a skill pack for Claude Code that runs monitoring, troubleshooting and audit tasks on your clusters, handing the heavy shell work to tools like Kubetail, Helm and Trivy to keep answers fast.
Other interesting projects:
Upcoming Kubernetes events
Aug
26
In-person conference organized by Devopsdays.
Location: Lima, PE
This event requires an entrance fee
Aug
27
In-person conference organized by Devopsdays.
Location: Jakarta, ID
This event requires an entrance fee
Aug
27
AI for Kubernetes: Runtime, Data, and Scale
In-person meetup organized by Kubernetes and Cloud Native Security India.
Location: Bengaluru, IN
This is a free event.
Sept
1
Kubernetes Community Days Korea 2026
In-person conference organized by KCD South Korea.
Location: Seoul, KR
This is a free event.
Sept
1
Kubernetes Community Days San Francisco Bay Area 2026
In-person conference organized by KCD SF Bay Area.
Location: San Francisco, US
This event requires an entrance fee
Sept
10
Online workshop organized by LearnKube.
This is a virtual event
This event requires an entrance fee
Kubernetes call for papers
11
days
Kubernetes Community Days Florida 2026 Miami
The Call For Paper is open until 7 September 2026 at UTC. More info →
Location: Miami, FL, USA
In-person conference organized by KCD Florida.
The conference starts on the 11 December 2026.
- Apply here
46
days
KubeCon + CloudNativeCon Europe 2027
The Call For Paper is open until 12 October 2026 at UTC. More info →
Location: Barcelona, ES
In-person conference organized by CNCF.
The conference starts on the 18 March 2027.
- Apply here
4
days
The Call For Paper is open until 31 August 2026 at UTC. More info →
Location: Bucharest, RO
In-person conference organized by OmniOpenCon.
The conference starts on the 18 October 2026.
- Apply here
1
days
The Call For Paper is open until 28 August 2026 at UTC. More info →
This is a virtual event
Online conference organized by ScyllaDB.
The conference starts on the 11 March 2027.
- Apply here
34
days
The Call For Paper is open until 30 September 2026 at UTC. More info →
Location: Bengaluru, IN
In-person conference organized by Canonical.
The conference starts on the 15 November 2026.
- Apply here
88
days
The Call For Paper is open until 23 November 2026 at UTC. More info →
Location: Vilnius, LT
In-person conference organized by DATA MINER.
The conference starts on the 27 November 2026.
- Apply here
25
days
The Call For Paper is open until 21 September 2026 at UTC. More info →
Location: Montreal, CA
In-person conference organized by ConFoo.
The conference starts on the 26 February 2027.
- Apply here
22
days
The Call For Paper is open until 18 September 2026 at UTC. More info →
Location: Santa Clara, CA, USA
In-person conference organized by DevNetwork.
The conference starts on the 11 February 2027.
- Apply here
49
days
Global Summit on Data Science and Cloud Computing
The Call For Paper is open until 15 October 2026 at UTC. More info →
Location: Rome, IT
In-person conference organized by Noveltics Group.
The conference starts on the 20 October 2026.
- Apply here
We're two weeks away from issue 200.
See you next wednesday.
— Gulcan