DevOps&SRE Library

📹Вебинар: Выбор между Serverless и Kubernetes для AI-ворклоадов: как определить оптимальную платформу под задачу На открытом уроке рассмотрим: - В чем различаются Serverless-подходы и Kubernetes при работе с AI-ворклоадами; - Какие преимущества и ограничения есть у каждого подхода с точки зрения масштабируемости, стоимости и сложности эксплуатации; - Какие трейдоффы нужно учитывать при выборе платформы: холодный старт, управление состоянием, поддержка GPU; - Как обосновывать выбор архитектуры для разных AI-сценариев на практическом воркшопе. После занятия вы будете знать: - Как сравнивать Serverless и Kubernetes для различных AI-задач; - Как выбирать платформу оркестрации в зависимости от требований к нагрузке, бюджету и архитектуре решения; - Как учитывать ключевые технические ограничения при проектировании AI-инфраструктуры; - Как аргументированно обосновывать выбор платформы для задач масштабирования, потоковой обработки данных и построения гибридных сред. ⚠️ Открытый урок проходит в преддверии старта курса «ИИ-архитектор». 👉 Для участия зарегистрируйтесь: https://vk.cc/cZywHe Реклама. ООО «Отус онлайн-образование», ОГРН 1177746618576, www.otus.ru, erid: 2VtzqxE1hEa

Kubernetes v1.35: Extended Toleration Operators to Support Numeric Comparisons (Alpha)

In Kubernetes v1.35, we're introducing Extended Toleration Operators as an alpha feature. This enhancement adds Gt (Greater Than) and Lt (Less Than) operators to spec.tolerations, enabling threshold-based scheduling decisions that unlock new possibilities for SLA-based placement, cost optimization, and performance-aware workload distribution.

https://kubernetes.io/blog/2026/01/05/kubernetes-v1-35-numeric-toleration-operators

Aetòs: From Chaos to Engineering Excellence — A 3-Year Transformation

How we transformed engineering productivity by building an Internal Developer Platform (IDP) that now processes ~50M API calls per day, manages 14,000 VMs, and powers 80+ releases per year and what you can learn from our journey.

https://medium.com/@charudatha/aet%C3%B2s-from-chaos-to-engineering-excellence-a-3-year-transformation-51a3f778678a

Kubernetes Remote Code Execution Via Nodes/Proxy GET Permission

In this post I'll describe how to execute code on every Pod in many Kubernetes clusters when using a service account with nodes/proxy GET permissions. This issue was initially reported through the Kubernetes security disclosure process and closed as working as intended.

https://grahamhelton.com/blog/nodes-proxy-rce

kubecfg

kubecfg is a Kubernetes kubeconfig manager for people who switch contexts often and need more than a thin wrapper around kubectl config.

https://github.com/kadirbelkuyu/kubecfg

Stakpak Agent

An open source agent that lives on your machines 24/7, keeps your apps running, and only pings when it needs a human. All the upside of a PaaS, none of the lock-in.

https://github.com/stakpak/agent

external-dns-provider-mikrotik

ExternalDNS is a Kubernetes add-on for automatically managing DNS records for Kubernetes ingresses and services by using different DNS providers. This webhook provider allows you to automate DNS records from your Kubernetes clusters into your MikroTik router.

https://github.com/mirceanton/external-dns-provider-mikrotik

cert-manager-webhook-pdns

A PowerDNS webhook for cert-manager

https://github.com/zachomedia/cert-manager-webhook-pdns

New Conversion from cgroup v1 CPU Shares to v2 CPU Weight

I'm excited to announce the implementation of an improved conversion formula from cgroup v1 CPU shares to cgroup v2 CPU weight. This enhancement addresses critical issues with CPU priority allocation for Kubernetes workloads when running on systems with cgroup v2.

https://kubernetes.io/blog/2026/01/30/new-cgroup-v1-to-v2-cpu-conversion-formula

Building a Centralized Multi Account AWS Monitoring Platform

I was previously part of an organization with 500+ devs with more than 25+ AWS accounts for each client which had to be managed by the DevOps team. The number of people in DevOps team was smaller than the accounts and all of them were mostly active with resources being allocated and deallocated frequently. We utilised many AWS services like EC2, S3, EKS, SageMaker, RDS, Redshift, Cloudfront, Route53, SSM etc

https://medium.com/@varunarora1408/building-a-centralized-multi-account-aws-monitoring-platform-6b14b59ae9d4

Designing for Failure: Chaos Engineering Best Practices

If you are leading an engineering team today, your job isn't to prevent failure. It is to design systems that thrive in it. This is the discipline of Chaos Engineering. It is not just about breaking things in production; it is a scientific method for validating resilience.

https://pulse.rajatgupta.work/designing-for-failure-chaos-engineering-best-practices-68503eb1b27b

Beyond Ingress Part II: Securing East-West Traffic with GKE Internal Gateway

This article will also focus on GKE Gateway, but now we're going to explore Internal Regional Gateway for load balancing east-west traffic (i.e. service-to-service) to your application. Leveraging our learnings from my previous post, we'll include the GKE external gateway in this architecture while using the Internal Regional Load Balancer gatewayclass (gke-l7-rilb) to load balance the service-to-service communications in the Bank of Anthos application.

https://medium.com/@bgillman_83663/beyond-ingress-part-ii-securing-east-west-traffic-with-gke-internal-gateway-7aef950aafe2

A field guide to sandboxes for AI

Every AI agent eventually asks for the same thing: "Let me run a program." Sometimes it's a harmless pytest. Sometimes it's pip install sketchy-package && python run.py. Either way, the moment you let an agent execute code, you're running untrusted bytes on a machine you care about.

https://www.luiscardoso.dev/blog/sandboxes-for-ai

🤖 ИИ врёт в проде А ещё может ломать процессы, уводить данные не туда и уверенно предлагать неверные решения. ⭐Слёрм запускает БЕСПЛАТНУЮ вечернюю школу «ИИ для инженеров: польза и риски». Это серия онлайн-занятий о том, как использовать ИИ в инженерной работе осознанно, безопасно и с понятной пользой. 🧩 Будем разбирать реальные инженерные сценарии: — как ИИ помогает DevOps-, SRE- и infrastructure-командам — как использовать LLM для алёртов, инцидентов, логов, тикетов и документации — где ИИ реально экономит время, а где создаёт новые риски — как проверять результат модели и не ловить галлюцинации в проде — что делать с безопасностью, данными, compliance и юридическими ограничениями — как встроить ИИ в рабочий процесс, а не просто иногда спрашивать у него команды. 🧩 В программе шесть онлайн-занятий с практиками из ИТ: — ИИ для разбора метрик и шумных алёртов — автофикс проблем прода с ИИ — ИИ-агенты в бизнес-задачах — юридические риски использования ИИ — LLM в SOC и борьба с alert fatigue — инженерное мышление в эпоху LLM Школа подойдёт DevOps-, SRE-, infrastructure-, platform- и security-инженерам, а также всем, кто уже пробовал ИИ в работе и хочет понять, как использовать его системнее и безопаснее. 📅 Старт — 21 июля. 💸 Участие бесплатное, занятия проходят онлайн. 👉🏻 Узнать подробнее и зарегистрироваться в боте Реклама. ООО "Слерм", ИНН: 3652901451, erid: 2Vtzqxjgnsc

Build a Kubernetes Cluster at Home with Raspberry Pis

Most people interact with Kubernetes through the cloud. They probably are as close to the cluster as they are to their laptop. Using a cluster in the cloud with a web browser or terminal is fine but there is something more intimate or rewarding to deploying a kubernetes cluster on bare metal. And even more rewarding is physically unplugging a node and watch Kubernetes rebalancing the workloads.

https://dev.to/anavalo/building-a-kubernetes-cluster-at-home-with-raspberry-pis-1l8c

Automated GitOps: from ECR push to EKS deploy

In this article, we'll walk through how to build a fully automated CI/CD workflow using Argo CD, Argo CD Image Updater, GitHub, Amazon Elastic Container Registry, Amazon Elastic Kubernetes Service. This setup enables automatic deployment whenever a new container image is pushed to the registry with no manual intervention.

https://medium.com/@alex.ivenin/automated-gitops-from-ecr-push-to-eks-deploy-a210cee0600c

Connecting Multi-Cloud Applications with Cilium

This article explores how Cilium Cluster Mesh simplifies multi-cluster Kubernetes connectivity, covering both the underlying challenges of so-called Mesh solution and how Cilium addresses them. It also provides a hands-on scenario that interconnects a multi-cloud application running on the native Kubernetes services of Amazon Web Services and Google Cloud Platform.

https://aws.plainenglish.io/connecting-multi-cloud-applications-with-cilium-d49e8363f05f

From Docker Compose to Kubernetes on AWS: A Hands-On Migration Story

How we took a 5-service Java application from docker-compose up to a production-grade Kubernetes cluster — and every bug we hit along the way.

https://medium.com/@zaouchwalid/from-docker-compose-to-kubernetes-on-aws-a-hands-on-migration-story-0ce5f52bb9dc

Kubernetes Optimization Beyond Requests and Limits — Node Scaling Blockers

Stop paying for idle resources. Learn to identify and eliminate the hidden blockers preventing your cluster from scaling down.

https://medium.com/@david.b.chase/kubernetes-optimization-beyond-requests-and-limits-node-scaling-blockers-649ae187f54a