DevOps & Infrastructure

CI/CD pipeline, observability (logs, metrics, traces), K8s vs alternatives, incident management

Technical Reference & Key Concepts

CI/CD Pipeline Stages

  1. Lint & Format: Enforce code style, catch obvious bugs early.
  2. Unit Tests: Fast feedback. Run on every commit.
  3. Build: Compile, containerize, version artifact.
  4. Integration Tests: Test with real dependencies (DB, cache, API).
  5. Deploy to Staging: Canary or blue-green. Run smoke tests.
  6. Deploy to Production: Gradual rollout (1% -> 10% -> 100%). Monitor for regressions.

Observability Triad

  • Logs: Structured (JSON). Centralized (ELK, Loki). Include request IDs for tracing.
  • Metrics: RED method (Rate, Errors, Duration) for services. USE method (Utilization, Saturation, Errors) for resources.
  • Traces: Distributed tracing (Jaeger, Zipkin). Follow a single request across services.

Container Orchestration: Kubernetes vs Nomad vs ECS - pick based on team size. K8s is powerful but operationally heavy. Nomad simpler for smaller teams.