The Kubestead Blog
SRE patterns, deployment war stories, and Kubernetes deep dives.
Canary Deployment Patterns for High-Traffic Microservices
Seven production-tested patterns for canary deployment across high-traffic microservice architectures, from single-service canaries to coordinated multi-service rollouts.
What to Do When Your Error Budget Is Exhausted
A practical guide to error budget exhaustion: detection, triage, and the rollback policies that keep your 30-day SLO window intact.
Kubernetes Rollout Strategies: Rolling, Blue-Green, and Canary Compared
A head-to-head technical comparison of the three main Kubernetes rollout strategies, with real-world failure modes and when to use each.
How to Automate 80% of Your On-Call Runbook
The deployments that wake you up at 3 AM are almost always the same five failure modes. Here is how to automate the response to all of them.
Argo Rollouts vs Flagger vs Kubestead: A Technical Comparison
Three Kubernetes canary deployment tools, three different architectural bets. Which one fits your platform engineering stack?
The Five PromQL Queries Every Canary Analysis Should Include
After analyzing 200+ canary rollbacks, these are the five Prometheus queries that catch the most bad deploys before users notice them.
Zero-Downtime Database Migrations During Canary Deployments
The hardest part of canary deployments is not the traffic split — it is keeping the database schema compatible across two concurrent application versions.
Deploy Ordering for Microservice Dependency Graphs
When you have 100 microservices and 60 dependency edges, deploy ordering is not a solved problem. Here is the approach that works.
Traffic Shaping with Istio and Linkerd for Canary Rollouts
A practical guide to configuring Istio VirtualService weights and Linkerd TrafficSplit resources for precise canary traffic control.
Writing Error Budget Policies That Actually Get Enforced
Most teams write error budget policies that live in a wiki and die in a quarterly review. This is how to write ones that the system enforces automatically.
Observing a Rollback: Telemetry Signals You Need
When the controller triggers a rollback, what does the telemetry look like? The spans, metrics, and events that give you a complete picture of a rollback event.
Instrumenting Canary Deployments with OpenTelemetry
How to use OpenTelemetry trace context propagation to distinguish canary traffic from stable traffic in your distributed traces.
Kubestead v0.8.0: Multi-Cluster Rollouts and the New REST API
Our biggest release yet: coordinated rollouts across up to 10 clusters, new REST API endpoints, and the Dashboard uptime history view.