Simple deployment pipelines
Push to main, run tests, deploy. Nothing clever.
DevOps
Ship it. Keep it running.
Practical infrastructure for small teams. One server or a handful. No cloud theatre.
What's in scope
Monitoring that pages you
Uptime, disk, queues, error rates. One alert channel, not ten.
Backups you can restore from
Automated, tested, offsite. I restore them once a month to prove they work.
Servers you can hand to the next person
Configuration in code. Documented. Reproducible on a new box.
Approach
What I actually build
- CI/CD pipelines on GitHub Actions or GitLab
- Deployments to a single VPS or a small cluster
- Monitoring with Prometheus, Grafana, or plain cron checks
- Infrastructure as code with Ansible or Terraform