DevOps&SRE Library

Bifrost’s journey from Nginx to Envoy gateway for intelligent rate limiting https://medium.com/the-heimdall-platform/bifrosts-journey-from-nginx-to-envoy-gateway-for-intelligent-rate-limiting-3215d19bc315

NVIDIA HGX Containers as a Service https://topofmind.dev/blog/2025/10/21/gpu-based-containers-as-a-service

phoenix

Phoenix is an open-source AI observability platform designed for experimentation, evaluation, and troubleshooting.

https://github.com/Arize-ai/phoenix

kubernetes-autoscaling-mixin

A set of Grafana dashboards and Prometheus alerts for Kubernetes Autoscaling using the metrics from Kube-state-metrics, Karpenter, and Cluster-autoscaler.

https://github.com/adinhodovic/kubernetes-autoscaling-mixin

juicefs

JuiceFS is a high-performance POSIX file system released under Apache License 2.0, particularly designed for the cloud-native environment. The data, stored via JuiceFS, will be persisted in Object Storage (e.g. Amazon S3), and the corresponding metadata can be persisted in various compatible database engines such as Redis, MySQL, and TiKV based on the scenarios and requirements. With JuiceFS, massive cloud storage can be directly connected to big data, machine learning, artificial intelligence, and various application platforms in production environments. Without modifying code, the massive cloud storage can be used as efficiently as local storage.

https://github.com/juicedata/juicefs

zot

zot: a production-ready vendor-neutral OCI image registry - images stored in OCI image format, distribution specification on-the-wire, that's it!

https://github.com/project-zot/zot

kide

OpenObserve Kide is a lightweight and fast Kubernetes IDE.

https://github.com/openobserve/kide

kubernetes-nmstate

Declarative node network configuration driven through Kubernetes API.

https://github.com/nmstate/kubernetes-nmstate

🚀 Закрывается набор на курс «DevOps-инженер: практики и инструменты» в OTUS Если вы хотите наконец-то выйти на уверенный middle/senior уровень в DevOps и перестать бояться продакшена — сейчас самый последний момент. Группа стартует 29 января, набор закрывается в ближайшие дни. Успейте пройти вступительное тестирование, подтвердить уровень и забронировать место по спеццене 🎯 Что даёт курс: - Полный production-ready стек 2025–2026: Docker → Kubernetes + Helm → Terraform + Ansible → GitLab CI/CD → Prometheus + Grafana + Loki + Jaeger + eBPF - Реальные практики, которые используют в Сбере, IBM, Касперском: GitOps, progressive delivery, IaC, полный цикл наблюдаемости, отказоустойчивость - Большой финальный проект с защитой: с нуля разворачиваете отказоустойчивое приложение в Yandex Cloud (бесплатно для студентов) и получаете детальный разбор от экспертов - Наставники — действующие DevOps из топовых компаний, которые каждый день решают те же задачи, что и вы - Удобный формат: онлайн-занятия по вт/чт в 20:00, записи навсегда, активный чат группы, код-ревью на каждое ДЗ 👉Пройдите тест прямо сейчас и забронируйте место: https://vk.cc/cTCufW Реклама. ООО «Отус онлайн-образование», ОГРН 1177746618576, erid: 2Vtzqx8PDud

kanidm

Kanidm is a simple and secure identity management platform, allowing other applications and services to offload the challenge of authenticating and storing identities to Kanidm. The goal of this project is to be a complete identity provider, covering the broadest possible set of requirements and integrations. You should not need any other components (like Keycloak) when you use Kanidm - we already have everything you need! To achieve this we rely heavily on strict defaults, simple configuration, and self-healing components. This allows Kanidm to support small home labs, families, small businesses, and all the way to the largest enterprise needs. If you want to host your own authentication service, then Kanidm is for you!

https://github.com/kanidm/kanidm

От кода до продакшена за 2 месяца — начни свой путь в DevOps на курсе от YADRO 📨 Практические курсы от YADRO – это обучение на реальных задачах: от теории к собственному проекту вместе с инженерами компании. ➡ На курсе по DevOps ты получишь практический опыт работы с полным стеком доставки приложения и соберёшь рабочую систему с нуля за два месяца.

Что тебя ждёт: ⏺ 2,5 месяца работы над проектом; ⏺ 80% времени — практика, задания и code review; ⏺ обучение онлайн или в офисе YADRO в Санкт-Петербурге; ⏺ менторство от инженеров YADRO; ⏺ возможность попасть на стажировку по итогам обучения.

*⃣ Подходит тебе, если: ты студент 2 курса и старше, уумеешь работать с Linux, скриптами и настройкой серверов. Подробности о курсе и других направлениях подготовки смотри на сайте!

HOWTO: Use SimKube for Cost Forecasting

Recently, I’ve had a number of folks ask for some more details about how SimKube can be used to predict or forecast your Kubernetes expenditures, and I realized that I’ve said you can do this several times, but I’ve never actually gone through the details! So this post will show you how.

https://blog.appliedcomputing.io/p/howto-use-simkube-for-cost-forecasting

Transforming Kubernetes Secret Management Best Practices https://medium.com/@nitinyadav745/transforming-kubernetes-secret-management-best-practices-804e993a22d9

Investigating and fixing "StopPodSandbox from runtime service failed" Kubelet errors https://marcusnoble.co.uk/2025-09-28-investigating-and-fixing-stoppodsandbox-from-runtime-service-failed-kubelet-errors

Scaling Dagster on Kubernetes: Best Practices for 50+ Code Locations https://u11d.com/blog/scaling-dagster-kubernetes-multi-code-locations

How We Moved a 2M RPM WebSocket Service to EKS and Fixed a Critical Bottleneck

Lessons in systems because AWS deprecated OpsWorks

https://medium.com/freshworks-engineering-blog/two-million-websockets-90f63e760cfd

Eliminate sensitive values from Terraform state using write-only attributes https://skundunotes.com/2025/12/22/eliminate-sensitive-values-from-terraform-state-using-write-only-attributes

You Can’t Debug a System by Blaming a Person

“I understand why we need to be blameless, but I have this person in my team who is often reckless. How can I not blame them when their actions continuously make things worse?” Someone asked me this at the SRE meetup, right after my talk on incidents. Since then I’ve been thinking about it, because it surfaces a concern many people might have.

https://humansinsystems.com/blog/you-cant-debug-a-systems-by-blaming-a-person

Datadog, Thank You for Blocking Us

Datadog cut off our observability overnight. We migrated to an open Grafana stack in 48 hours. Here’s why vendor lock-in is fading in an AI-native world.

https://www.deductive.ai/blogs/datadog-thank-you-for-blocking-us

Resilience vs. Fault tolerance

In this post, I will discuss if there is a difference between resilience and fault tolerance when talking about IT systems.

https://www.ufried.com/blog/resilience_vs_fault_tolerance