Skill gap: Production SRE practices — incident response, service-level objectives (SLOs), observability (metrics, logging, tracing), and on-call tooling. This is important because my role owns services end-to-end and I need to reliably run systems in production.
90-day plan (weekly milestones):
Weeks 1–2: Foundations
- Read: “Site Reliability Engineering” (O’Reilly chapters on SLOs & incident response) + Google’s SRE book (selected chapters).
- Course: “Site Reliability Engineering: Measuring and Managing Reliability” (Coursera).
- Deliverable: Summary doc of SRE principles and proposed SLO templates for one of our services.
Weeks 3–4: Observability basics
- Course: “Monitoring, Logging, and Tracing” (Udemy/Pluralsight).
- Hands-on: Add Prometheus metrics and Grafana dashboard to a small microservice; instrument structured logging (json) and basic traces with OpenTelemetry.
- Deliverable: Demo dashboard + README showing metrics, example queries, and retention plan.
Weeks 5–6: Tracing & debugging
- Read: OpenTelemetry docs; workshop on distributed tracing.
- Hands-on: Integrate OpenTelemetry traces across a simple request flow; visualize traces in Jaeger.
- Deliverable: Postmortem-style runbook for diagnosing high-latency requests using traces.
Weeks 7–8: SLOs, SLIs, error budgets
- Workshop: Define SLIs/SLOs for a chosen service; set error budget policy.
- Hands-on: Implement alerting rules in Alertmanager; create alert runbooks.
- Deliverable: SLO dashboard and alert playbooks.
Weeks 9–10: Incident response & on-call
- Practice: Participate in a simulated incident (game day) with team; run blameless postmortem.
- Read: “The Phoenix Project” (selected) for cultural practices.
- Deliverable: After-action report with improvements and updated runbooks.
Weeks 11–12: Automation & handoff
- Automate common remediation (restarts, scaling) via scripts/Playbooks and CI.
- Teach: 30-minute brown-bag + walkthrough of dashboards and runbooks.
- Demonstrate: Lead a practicum where teammates resolve a seeded issue using runbooks and dashboards.
How I'll demonstrate competency:
- Deliverables above (SLO docs, dashboards, runbooks, automated remediation) in our repo.
- Run the brown-bag session and lead a game-day; collect feedback.
- Propose and implement at least one low-effort reliability improvement (e.g., effective alert tuning or an automated rollback) merged to main.
- Metrics: show reduced alert noise (target: 30% fewer false-positive alerts) and clear acceptance from teammates via checklist sign-off.
This plan combines theory, hands-on implementation, and team-facing demonstrations to ensure sustained competency and immediate impact.
