Monitoring Metrics Data Flow

Monitoring Metrics Data Flow A data-flow diagram generated by Archify. 01 / Cluster sources 02 / Transport routing 03 / KubeOsprey query layer 04 / API endpoints 05 / Frontend Prometheus · Auto-discovery · 01 / Cluster sources · PromQL Prometheus Auto-discovery PromQL metrics-server · Pod metrics fallback · 01 / Cluster sources metrics-server Pod metrics fallback RoundTripper · Direct / proxy / tunnel · 02 / Transport routing · Per cluster mode RoundTripper Direct / proxy / tunnel Per cluster mode pkg/prometheus · Query · QueryRange · 03 / KubeOsprey query layer pkg/prometheus Query · QueryRange pkg/metrics · /api/v1/prometheus/* · 04 / API endpoints pkg/metrics /api/v1/prometheus/* Resource charts · Pod / workload usage · 05 / Frontend Resource charts Pod / workload usage Prometheus page · PromQL queries · 05 / Frontend Prometheus page PromQL queries Metric series PromQL response Pass-through HTTP Query results JSON series Pod metrics fallback metrics API Usage history Read-only Custom queries RBAC-filtered Legend primary data async batch data flow

Transport routing

  • • In-cluster URLs go through the K8s API proxy
  • • Agent mode dials the tunnel; others connect directly

Query capabilities

  • • Query / QueryRange for instant and range queries
  • • Resource usage history aggregation

Fallback

  • • Live Pod metrics can fall back to metrics-server