DevOps / Site Reliability Engineer in Ho Chi Minh City. I run my own production platform end to end and keep it honest with SLOs, canary releases and on-call.
Portfolio · Résumé (PDF) · LinkedIn · contact@docaohieu.com
- 5 Linux servers in Singapore and Vietnam, about 120 containers, 30+ apps, one private Tailscale network, Traefik and Cloudflare at the edge.
- Observability: Prometheus, Alertmanager, Grafana, Loki and Tempo on a separate node; 8 SLOs with burn-rate alerts; a dead man's switch on Cloudflare Workers pages me if monitoring itself goes quiet.
- Delivery: self-managed GitLab CI, GitHub Actions, Jenkins, Harbor; canary deploys that roll back on their own.
- Data and recovery: PostgreSQL and Redis with automatic failover (Percona, Sentinel); nightly Kopia backups to two cloud drives.
- Remote, since 2024: the whole digital side of Wellington Decorators (New Zealand): website, CI/CD, email, monitoring.
| Project | What it is |
|---|---|
| fleet-pulse | SLOs with error budgets, canary deploys with automatic rollback, incidents that draft their own postmortem. |
| job-radar | A focused crawler that finds, scores and triages DevOps jobs: job boards, career pages, ATS APIs, Postgres queue. |
| ansible-toolkit | Ansible roles and playbooks for provisioning Linux servers, lint-checked on every push. |
| reusable-workflows | Zero-config CI/CD for GitHub Actions: detect, build, test, scan, deploy. |
| pdfx | Agent skill that turns Markdown into publication-grade PDFs and processes existing PDFs. |
| floatwave | Floating mini-player for YouTube Music on Windows (Electron), 20 releases. |
Linux · Docker / Compose · Kubernetes, Helm (test cluster) · Traefik, Nginx · GitLab CI, GitHub Actions, Jenkins · Prometheus, Grafana, Loki, Tempo · Terraform, Ansible · PostgreSQL, Redis · Vault · Cloudflare, Oracle Cloud · Bash, Python, Go, TypeScript
DevOps, SRE and Platform roles, in Ho Chi Minh City or remote.


