OPERATING DOCUMENT / NARWHAL

Observability Stack

Prometheus, Grafana, Loki, Tempo, Alloy, and Hubble eBPF telemetry.

01 / Context02 / Procedure03 / Operations04 / Verification

Observability Stack

Narwhal provides full-stack telemetry integrating metrics, logs, distributed traces, and network flows into a single unified operational dashboard.

Telemetry Pipeline

Architecture diagram
Responsive vector rendering · source preserved
[ Kubernetes Pods & Nodes ] Metrics (Prometheus Exporters) ► Prometheus ► Grafana Logs (Alloy Agent) ► Loki ► Grafana Traces (OpenTelemetry SDK) ► Tempo ► Grafana Network Flows (eBPF BPF Maps) ► Hubble UI

4 Observability Pillars

  1. Metrics (Prometheus Operator): Node CPU/memory, XFS storage consumption, and APISIX latency alerts
  2. Logs (Grafana Loki + Alloy): Label-indexed container log aggregation and streaming
  3. Traces (Grafana Tempo): OpenTelemetry distributed trace correlation with logs via TraceIDs
  4. Network Telemetry (Cilium Hubble): Real-time service topology, DNS latency, and packet drops via eBPF
DOCUMENT TRACE

Return to the project record to see architecture, development pulse, proof and related field notes in one context.