Tag: devops

Dragonfly: P2P Container Image Distribution at Scale

Dragonfly turns your container registry into a P2P swarm. Stop hammering one registry with 200 nodes pulling 4GB GPU images at once.

Renovate vs Dependabot: Self-Hosted Dependency Bots

Renovate vs Dependabot for self-hosted teams: config depth, ecosystem coverage, self-host paths, and which bot wins for Forgejo/GitLab shops.

Gitea Actions vs Woodpecker CI

Gitea Actions vs Woodpecker CI: two solid self-hosted CI options, one clear winner for your setup. Compare syntax, runners, footprint, and secrets handling.

ClickHouse for Self-Hosted Logs

ClickHouse can store TB of logs on one node, query in seconds, and outscale Loki and Elasticsearch in raw cost. Here's how to wire it up with Vector and Grafana.

AI Swarm Audited My 840-Post Blog

I pointed a parallel swarm of AI agents at 840 technical posts to fact-check and refresh them — the architecture, the token bill, and the guardrails.

OPA & Gatekeeper: Policy as Code

Open Policy Agent and Gatekeeper enforce rules across Kubernetes — block privileged pods, require labels, restrict images. Rego basics and real homelab constraints.

Claude Code in a Homelab Workflow

Claude Code puts an agentic AI assistant in your terminal for real homelab work — compose files, bash, Ansible, systemd. The honest take on cost and data.

Kustomize vs Jsonnet for K8s Manifests

Kustomize patches YAML with overlays. Jsonnet is a real templating language with imports and functions. Here's which K8s manifest tool to pick.

Argo Rollouts vs Flagger Progressive Delivery

Argo Rollouts uses a Rollout CRD with a slick UI. Flagger watches your Deployment with a Canary CRD. Here's which progressive-delivery tool to pick.

Argo Workflows vs Tekton

Two Kubernetes-native pipeline engines from CNCF — Argo Workflows and Tekton. Different philosophies, same goal: stop using Jenkins. Honest comparison.

k3sup vs kubeadm for Homelab Clusters

kubeadm is the official Kubernetes bootstrapper. k3sup is a one-shot SSH installer for k3s. Here's which one to use for your homelab cluster.

Dead Container Took Down Prod

A real postmortem: how a forgotten survey-tool container, an uncapped Docker log driver, and rootless networking combined to kill prod for a Monday morning.