505 posts tagged with "Kubernetes"
Container orchestration, Cluster API, and self-hosted control planes
France's Railway Runs on Cluster API: What SNCF's Monthly Zero-Drift Fleet Teaches a Self-Hosted PaaS
SNCF provisions Kubernetes clusters in 30 minutes and updates its whole fleet monthly with zero configuration drift — on infrastructure it owns. Here is the stack behind those numbers and which parts transfer to a small self-hosted fleet.
Tau Skipped Kubernetes Entirely: What a Wasm-Native PaaS Gains — and Gives Up
Taubyte's Tau runs a self-hosted PaaS on WebAssembly and libp2p with zero Kubernetes. This comparison tables what deleting the control plane buys — sub-millisecond cold starts, no etcd — against what it forecloses: no Docker images, no operators, no GPU scheduling.
The $50K 'Skip Kubernetes' Claim, Audited Against Owned Hardware
A March 2026 essay claims skipping Kubernetes saves $50K a year. We audit the number line by line — managed control-plane fees and node markup versus a CAPH fleet on Hetzner — and find the real variable is tenant count.
Seven Agent Sandbox Vendors Shipped Snapshot by June 2026 — What Happens When a Differentiator Becomes Table Stakes
By June 2026, seven agent-sandbox vendors shipped snapshot and resume — but only three preserve running processes. A side-by-side comparison of what each one saves, what it costs underneath, and why memory resume is the hard part for self-hosted platforms.
Cilium 1.20 Retires Two More Boxes: TCP Routes, Gateway Auth, and the Case for a CNI-Run Fleet
Cilium 1.20 adds TCPRoute/UDPRoute and gateway ExternalAuth to its Gateway API support, closing the last gaps that kept small fleets running a separate ingress controller and auth proxy. Here is the consolidation map, the concrete migration costs, and where a service mesh still earns its keep.
After You Leave Render, Heroku, or Fly.io: Which Cost-Visibility Tools Survive the Move to Flat-Rate Hardware
Qovery's guide walks a 25-person team through eight cost tools for life after Render, Heroku, or Fly.io — but every pick assumes AWS, GCP, or Azure. This post re-runs the shortlist for flat-rate owned hardware, where the meter disappears and only OpenCost plus auto-stop still earn their keep.
Kubernetes 1.36 In-Place Pod Resize Goes Beta: What Restart-Free Vertical Scaling Replaces in a PaaS Autoscaler
Kubernetes 1.36 graduates in-place vertical scaling for pod-level resources to beta: the kubelet can now grow a running pod's CPU and memory budget without a restart. This post shows the resize operation, the before/after for a PaaS autoscaler, and the beta caveats to clear first.
Nine Years in Beta Is Over: What Kubernetes 1.37's Stable Metrics API Means for Your Autoscaler
Kubernetes 1.37 promotes the metrics.k8s.io API to stable after nearly nine years in beta. What the frozen contract changes for autoscaling and usage metering on self-hosted clusters — plus the HPA gotcha and everything the API still doesn't do.
One CLI Command to a Full Remote Stack: What Moniepoint's Crossplane-Built Dev Environments Cost to Build Versus Buy
Moniepoint's eshu turns Score YAML into a full remote stack with one CLI command via Crossplane composition functions. The architecture, the two hard lessons (render-cycle steady state, 63-character names), and the build-vs-buy math: building wins past roughly 17 developers.