Tu es un expert SRE et observabilité. Je veux monitorer une architecture microservices [KUBERNETES/SERVERLESS]. Stack d'observabilité complète: 1. **Metrics Collection** : Prometheus, Grafana Mimir, VictoriaMetrics, OpenTelemetry Collector 2. **Distributed Tracing** : Jaeger, Tempo, Zipkin, OpenTelemetry tracing 3. **Logging Strategy** : Loki, ELK Stack, Fluentd, structured logging 4. **APM Integration** : New Relic, DataDog, Dynatrace, open-source alternatives 5. **Service Mesh** : Istio, Linkerd, Consul Connect with telemetry 6. **Custom Dashboards** : Grafana dashboards, SLI/SLO definitions, error budgets 7. **Alerting Strategy** : Alertmanager, PagerDuty, OpsGenie, escalation policies 8. **Chaos Engineering** : Gremlin, Litmus Chaos, failure injection testing 9. **Capacity Planning** : Resource utilization trends, predictive scaling 10. **Business Metrics** : User journey tracking, conversion funnels, A/B testing Fournis les configurations Docker/Kubernetes, les dashboards Grafana, les règles d'alerting, et les charts Helm.