Site Reliability Engineering Roadmap

Understand how reliable systems are run: SLOs, monitoring, incident response and automation.

10 milestones · ~130 hours total

1. 1

   Operating systems and networking depth

   ~22h

   Linux internals, performance tools like top, strace and tcpdump, and how a request travels through the network.

   Practice this →
2. 2

   Scripting and automation

   ~18h

   Python and Bash to automate repetitive operations; write small tools with proper error handling.

   Practice this →
3. 3

   Reliability concepts: SLIs, SLOs and error budgets

   ~10h

   Define what reliable means for a service, pick indicators and use error budgets to balance features and stability.

   Practice this →
4. 4

   Monitoring with Prometheus and Grafana

   ~16h

   Metrics types, PromQL basics, dashboards and the difference between symptom and cause alerts.

   Practice this →
5. 5

   Logging and tracing

   ~12h

   Structured logs, centralised search, distributed tracing with OpenTelemetry and correlating signals.

   Practice this →
6. 6

   Alerting and on-call practice

   ~8h

   Actionable alerts, avoiding noise, runbooks and escalation. Practise with a simulated on-call exercise.

   Practice this →
7. 7

   Incident response and postmortems

   ~8h

   Roles during an incident, communication, mitigation vs root cause and blameless postmortem writing.

   Practice this →
8. 8

   Capacity planning and load testing

   ~12h

   Use k6 or Locust to find limits, understand queueing and plan headroom and autoscaling.

   Practice this →
9. 9

   Resilience patterns

   ~14h

   Timeouts, retries with backoff, circuit breakers, graceful degradation, redundancy and safe deployments such as canary releases.

   Practice this →
10. 10

    Chaos experiments and game days

    ~10h

    Run small failure experiments in a test environment, record learnings and fix the weaknesses found.

    Practice this →