moai-domain-monitoring

End-to-end observability combines metrics, logs, and traces across services. Use Prometheus for metrics collection, Loki for structured logging, OpenTelemetry for instrumentation, and Jaeger for distributed tracing. This Skill provides a complete stack with prebuilt Grafana dashboards to visualize all three signals together in Kubernetes environments.

Prometheus metrics are time-series data (Counters, Gauges, Histograms) for system performance. Structured logs capture timestamped events with context in JSON format. Traces follow requests across microservices using Trace IDs and parent-child relationships. Together they provide complete visibility into production system behavior.

Yes. OpenTelemetry instruments your code to generate spans and traces with Trace IDs, while Jaeger exports and visualizes those traces. This Skill integrates both: OpenTelemetry collects instrumentation, Jaeger/Tempo backends store traces, and Grafana dashboards display results across your Kubernetes services.

Prometheus Alertmanager integration triggers alerts based on metric thresholds. This Skill includes Alertmanager configuration alongside Prometheus collection and Grafana visualization, enabling automated notifications when production systems exceed defined alert conditions.

Loki indexes and queries structured JSON logs efficiently within your observability stack. While structured logging provides the format, Loki enables fast log retrieval, correlation with metrics and traces in Grafana, and reduces storage overhead compared to traditional log aggregators in Kubernetes environments.

Use Counters for cumulative counts, Gauges for point-in-time values, and Histograms for request latency distribution. This Skill documents and demonstrates all three types with prebuilt Grafana dashboards that aggregate these metrics across microservices in Kubernetes for comprehensive production monitoring.