Blog
Insights, analysis, and updates from the AI agent economy. Browse by tag · Browse the archive.
Kubernetes 1.37 Made Pod Certificates GA: The 4-Test Signer Harness to Run Before You Kill Static Credentials
Kubernetes 1.37 promotes Pod Certificates and Cluster Trust Bundles to stable, with 24-hour X.509 certs issued via PodCertificateRequest. A four-test harness for signer authorization, trust-bundle selection, lifetime refresh, and reload behavior that a self-hosted PaaS should pass before retiring static service-account tokens.
Kubernetes 1.37's Stable Metrics API: What Actually Changes for Metrics Server, HPA, and kubectl top
Kubernetes 1.37 promotes metrics.k8s.io to stable v1 after nine years in beta — a contract graduation, not a feature. What still needs metrics-server, how kubectl top and HPA behave through the transition, and an ordered checklist for self-hosted Cluster API fleets.
One Bad GPU Shouldn't Sink the Node: Kubernetes 1.37's DeviceTaintRule Goes Stable
Kubernetes 1.37 promotes DRA device taints and tolerations to stable: a DeviceTaintRule fences one bad GPU with NoSchedule or NoExecute while the rest of the node keeps serving tenants. The YAML, the dry-run workflow, and what your health-check pipeline must build.
TCPRoute Went Stable, but Somebody Still Has to Own the IP: kube-vip vs Talos BGP vs Hetzner LBaaS
Gateway API 1.6 made TCPRoute portable, but something still has to announce the IP. A worked comparison: kube-vip ARP is free but LAN-bound, Talos BGP is free but needs a peer, and Hetzner LBaaS costs ~€7.49 per TCP service — €37 at five services, €150 at twenty.
Karmada for Cluster API Fleets: When One Control Plane Stops Being Enough
Karmada's hub-and-member model promises cross-cluster scheduling, federated quotas, and taint-driven failover for Cluster API fleets — but it doubles your control planes. A verdict-first guide to what it buys, what it costs, and the quarterly ceiling test that says when.
Instant Kubernetes Is 5-10 Minutes, Not Sub-Second: The Real Cost Math of Cluster-per-PR Preview Environments
A Show HN promised instant Kubernetes with KVM and gRPC; its own numbers say 5-10 minutes per cluster. A scored comparison of five preview strategies and the fleet-scale math — node-hours, cold caches, control-plane sprawl — that decides cluster-per-PR vs namespace-per-PR.
Ingress-NGINX Is Dead: Your Per-Route Migration Playbook Before the November 2026 Cliff
Ingress-nginx was archived in March 2026 and Azure ends managed NGINX patches in November 2026. A concrete migration plan: which annotations Traefik's drop-in provider covers, what must be hand-rewritten as HTTPRoute, and how to cut over one host at a time.
Heroku Is in Sustaining Mode: When to Migrate and What Render, Railway, and Fly.io Really Cost
Heroku's February 2026 sustaining-mode announcement closed the door on new Enterprise contracts and froze new features. A timing framework for when to leave, a five-way cost table for the same workload on Heroku, Render, Railway, Fly.io, and Hetzner hardware, and where each alternative actually fits.
Heroku's Fir Can't Deploy Your Dockerfile: a Migration Playbook With Real Before-and-After Bills
Heroku's Fir generation can't deploy Docker images and Cedar entered sustaining mode in February 2026. A migration playbook for container teams: what breaks on Fir, how Render, Railway, and Fly.io handle Dockerfiles, and one typical stack priced five ways — $87 on Heroku, ~$35–60 hosted, ~$7 self-hosted.
Subscribe
New posts land in your reader as soon as they publish. Pick a format — all three carry the same posts.
Current feeds keep roughly two days of posts so daily polling does not miss a burst. Older entries stay reachable from the feed's next-page link in readers that follow it, or from the blog archive.
Following one topic instead? Browse tags