Learn Kubernetes Weekly issue 138

From Prometheus to Thanos, Increasing Memory of NGINX, AI-Driven Autoscaler, Templating Alertmanager, Envoy Gateway: Rate Limiting with Cost

2 Jul 2025

This issue is brought to you by LearnKube — get started on your Kubernetes journey through comprehensive online, in-person or remote training.

TL;DR: We're now LearnKube! Newsletter moves to Kube Today

After 8 years of teaching Kubernetes as Learnk8s, we've moved to our new home: LearnKube.com.

Nothing changes about what we do - same instructors, same hands-on labs, same deep-dive Kubernetes training that over 3,000 engineers have trusted to level up their skills.

So why the change?

Sometimes, the universe nudges you in the right direction.

LearnKube actually fits better with our family of brands (Kube Events, Kube Careers, KubeFM and the news we publish on Kube Architect, Kubesploit and Kube Builders)

One more thing!

I'm launching a new project soon called Kube Today (you are the first to know, this is still a WIP). Kube Today has a mission similar to the newsletter: to surface news, tutorials, articles, events, jobs, etc., on a daily basis. I will move this newsletter to Kube Today as it aligns better with the goal.

Here is what you need to know:

  • The newsletter won't stop, and the format won't change.
  • You will receive the newsletter from a new sender ( weekly@kube.today)
  • The newsletter might end up in spam if you don't whitelist the address.

Exciting times!

  1. How Using Availability Zones Can Eat Up Your Budget: Our Journey from Prometheus to Thanos

    Grzegorz Skołyszewski

    This article details Prezi’s migration from Prometheus to Thanos, analyzing how network traffic across cloud availability zones increased monitoring costs and how Thanos reduced expenses with long-term, multi-zone metrics storage.

  2. IPA: Building AI-Driven Kubernetes Autoscaler

    Shafin Hasnat

    Instead of relying solely on predefined thresholds, why not use AI to analyze logs, reason through the data, and provide scaling recommendations?

    Learn how Shafin built an AI-Driven Kubernetes autoscaler in this article.

  3. Increasing Memory Usage of NGINX Ingress after Upgrading GKE to Version 1.30

    Irwandhi Septian

    Learn how (and why) NGINX Ingress v1.10.1 misread CPU limits on cgroup v2 nodes, spawning 32 workers instead of 1, causing memory bloat and OOM.

    Spoiler: the fix included upgrading to v1.10.4, which correctly parses CPU quota from cgroup v2.

  4. Templating Alertmanager Config in kube-prometheus-stack

    Dan Williams

    Loveholidays had a 1700-line static Alertmanager config used to route Prometheus and Loki alerts to Slack and PagerDuty across 15+ teams.

    Learn how they used Helm's tpl function to dynamically generate Alertmanager config from a simple values.yaml.

  5. Envoy Gateway: Overview of the New “Rate Limiting with Cost” Feature

    reoring

    Envoy Gateway v1.3 introduces cost-based rate limiting, allowing API operators to charge or restrict requests based on dynamic runtime metrics (e.g., token usage in AI inference) instead of simple request counts.

  6. Minimal Viable Platforms

    Natan Yellin,CEO

    This post shows how to solve DevOps bottlenecks like infra requests, Kubernetes onboarding, and fragmented tooling using high-ROI solutions before overengineering a complete platform.

Articles worth checking out:

Join the next Advanced Kubernetes course

Join LearnKube's 4-day Advanced Kubernetes workshop!

Get your hands dirty with Kubernetes and learn what makes Kubernetes tick in a session packed with hands-on labs!

Become an expert

Join the next Advanced Kubernetes course
  1. Kubernetes Troubleshooting: Fixing Pod Issues with Restricted UID in securityContext

    Asish M Madhu

    This deep dive walks through debugging a common Kubernetes issue: running containers with a non-root UID.

  2. Using GitOps to manage a KWOK deployment

    Charles Uneze

    This post walks through deploying a GitOps pipeline using Flux to deploy KWOK (Kubernetes without Kubelet) on KIND, simulate pod/node loads, and expose kube-scheduler metrics via Prometheus and Grafana.

  3. How to Build a Multi-Tenancy Internal Developer Platform with GitOps and vCluster

    Artem Lajko

    This tutorial explains how to build a multi-tenancy IDP with GitOps and vCluster for isolated Kubernetes automation.

    • Site Reliability Engineer with CoW DAO

    • Salary: €90K to €120K a year

    • Location: remote within UTC±6

    • Tech stack: Kubernetes, AWS, Flux, Docker, Go, Python, Rust, PostgreSQL, Elastic Search, Pulumi

    • Software Engineer with Jobgether

    • Salary: $180K to $260K a year

    • Location: remote from the United States

    • Tech stack: Kubernetes, AWS, Azure, GCP, SQL

    • Software Engineer with Verkada

    • Salary: $130K to $280K a year

    • Location: based in the office in San Mateo, CA, USA

    • Tech stack: Kubernetes, AWS, GCP, Javascript, Typescript, DynamoDB, Airflow

    • Product Owner with Codat

    • Salary: £100K to £125K a year

    • Location: based in the office (and remote from home) in London, GB

    • Tech stack: Kubernetes, Azure, Docker, Javascript, Typescript, C#

    • Software Engineer with Pinecone

    • Salary: $175K to $210K a year

    • Location: based in the office (and remote from home) in New York, NY, USA

    • Tech stack: Kubernetes, Go, Rust, C++

Discover more Kubernetes jobs on Kube Careers →

  1. Zeropod: scale to zero

    ctrox

    zeropod is a tool that automatically checkpoints containers to disk after a certain amount of time of the last TCP connection, allowing for fast and seamless scaling down to zero.

  2. Kargo: application lifecycle orchestration

    akuity

    Kargo is a continuous delivery and application lifecycle orchestration platform for Kubernetes.

    It builds upon GitOps principles and integrates with Argo CD to streamline and automate the progressive rollout of changes across an application's lifecycle.

  3. fairwindsops/goldilocks

    FairwindsOps

    Goldilocks is a utility that can help you identify a starting point for resource requests and limits in Kubernetes.

  4. emberstack/kubernetes-reflector

    Reflector is a Kubernetes addon designed to monitor changes to resources (secrets and configmaps) and reflect changes to mirror resources in the same or other namespaces.

  5. Kubernetes Network Policy Recipes

    ahmetb

    This repo is a collection of NetworkPolicy recipes to lock down Kubernetes traffic.

Other interesting projects:

Upcoming Kubernetes events

  1. Jul

    3

    Kubernetes Community Days Utrecht 2025

    Online conference organized by KCD Utrecht.

    • This is a virtual event

    • This event requires an entrance fee

  2. Jul

    5

    Kubernetes Community Days Taipei 2025

    In-person conference organized by KCD Taiwan.

    • Location: Taipei, TW

    • This is a free event.

  3. Jul

    5

    Getting Started with Kubernetes

    In-person meetup organized by Junior Developers Singapore.

    • Location: Singapore, SG

    • This is a free event.

  4. Jul

    9

    Root Cause Analysis in Kubernetes, GitOps & Argo CD

    In-person meetup organized by Cloud Native Boston.

    • Location: Boston, MA, USA

    • This is a free event.

  5. Jul

    10

    DevBcn

    In-person conference organized by DevBcn.

    • Location: Barcelona, ES

    • This event requires an entrance fee

      • Use KUBE-10 to get 10% off

  6. Sept

    15

    Advanced Kubernetes course (Singapore)

    In-person workshop organized by Learnk8s.

    • Location: Singapore, SG

    • This event requires an entrance fee

Discover more Kubernetes events on Kube Events →

Kubernetes Call for Papers

  1. 27

    days

    Kubernetes Community Days Sri Lanka 2025

    The Call For Paper is open until 4 August 2025 at UTC. More info →
    • Location: Colombo, LK

    • In-person conference organized by KCD Sri Lanka.

    • The conference starts on the 26 October 2025.

    • Apply here
  2. expired

    Kubernetes Community Days El Salvador

    The Call For Paper was open until 7 July 2025 at UTC. More info →
    • Location: San Salvador, SV

    • In-person conference organized by KCD El Salvador.

    • The conference starts on the 20 September 2025.

    • Apply here
  3. 24

    days

    CloudCon Sydney 2025

    The Call For Paper is open until 1 August 2025 at UTC. More info →
    • Location: Sydney, AU

    • In-person conference organized by Cloud Native Sydney.

    • The conference starts on the 9 September 2025.

    • Apply here
  4. 26

    days

    Texas Linux Festival 2025

    The Call For Paper is open until 3 August 2025 at UTC. More info →
    • Location: Austin, TX, USA

    • In-person conference organized by TXLF.

    • The conference starts on the 4 October 2025.

    • Apply here
  5. 27

    days

    Open Source Summit Japan 2025

    The Call For Paper is open until 4 August 2025 at UTC. More info →
    • Location: Tokyo, JP

    • In-person conference organized by Linux Foundation.

    • The conference starts on the 10 December 2025.

    • Apply here
  6. 39

    days

    Devopsdays Detroit

    The Call For Paper is open until 16 August 2025 at UTC. More info →
    • Location: Detroit, MI, USA

    • In-person conference organized by Devopsdays.

    • The conference starts on the 22 October 2025.

    • Apply here
  7. 23

    days

    PWNEDCR 0x8

    The Call For Paper is open until 31 July 2025 at UTC. More info →
    • Location: San José, CR

    • In-person conference organized by DC11506.

    • The conference starts on the 19 October 2025.

    • Apply here
  8. 70

    days

    Devopsdays Bogotá

    The Call For Paper is open until 16 September 2025 at UTC. More info →
    • Location: Bogotá, CO

    • In-person conference organized by Devopsdays.

    • The conference starts on the 14 October 2025.

    • Apply here
  9. 37

    days

    Devopsdays Ljubljana

    The Call For Paper is open until 14 August 2025 at UTC. More info →
    • Location: Ljubljana, SI

    • In-person conference organized by Devopsdays.

    • The conference starts on the 13 September 2025.

    • Apply here

Until next time!

— Dan

Subscribe and, every Wednesday, receive the latest Kubernetes news!

Or follow us on: