deployment-procedures

Follow a five-phase process: prepare by verifying tests and environment variables, back up the current state, deploy while monitoring, verify health endpoints and key flows, then confirm or roll back. Never deploy untested code or skip the backup step.

Static sites fit Vercel, Netlify, or Cloudflare Pages. Simple web apps work on managed platforms like Railway or Render, or a VPS with PM2 for more control. Microservices typically need container orchestration with Docker or Kubernetes.

Roll back immediately if the service is down or critical errors appear. Consider rollback when performance degrades more than 50 percent. For minor issues, a quick fix-forward may be acceptable, but prioritize speed over perfection when rolling back.

Blue-green switches all traffic between two environments, making rollback fast and suiting high-risk changes. Canary shifts traffic gradually to the new version, letting you validate with real users before full rollout.

Use kubectl rollout undo to return to the previous revision. On Vercel or Netlify, redeploy the previous commit; on Railway or Render, use the dashboard rollback; on a VPS with PM2, restore the backup and restart the process.

Late-week deployments leave little time to monitor and respond to issues before the weekend. Deploy early in the week so the team is available to watch metrics, verify stability, and handle rollbacks if problems appear.