Prometheus

Standalone Prometheus monitoring with Gateway API HTTPRoute.

Deployment

---
apiVersion: source.toolkit.fluxcd.io/v1
kind: GitRepository
metadata:
  name: flux-apps
  namespace: flux-system
spec:
  interval: 1m0s
  ref:
    tag: v1.0.0
  url: https://github.com/stuttgart-things/flux.git
---
apiVersion: kustomize.toolkit.fluxcd.io/v1
kind: Kustomization
metadata:
  name: prometheus
  namespace: flux-system
spec:
  interval: 1h
  retryInterval: 1m
  timeout: 5m
  sourceRef:
    kind: GitRepository
    name: flux-apps
  path: ./infra/prometheus
  prune: true
  wait: true
  postBuild:
    substitute:
      PROMETHEUS_NAMESPACE: monitoring
      PROMETHEUS_VERSION: "28.13.0"
      PROMETHEUS_STORAGE_CLASS: nfs4-csi
      PROMETHEUS_STORAGE_SIZE: 10Gi
      PROMETHEUS_RETENTION: 15d
      GATEWAY_NAME: cilium-gateway
      GATEWAY_NAMESPACE: default
      HOSTNAME: prometheus
      DOMAIN: example.sthings-vsphere.labul.sva.de

Variables

Variable	Default	Description
`PROMETHEUS_NAMESPACE`	`monitoring`	Target namespace
`PROMETHEUS_VERSION`	`28.13.0`	Helm chart version
`PROMETHEUS_STORAGE_CLASS`	`nfs4-csi`	StorageClass for persistent volume
`PROMETHEUS_STORAGE_SIZE`	`10Gi`	PVC size
`PROMETHEUS_RETENTION`	`15d`	Data retention period
`GATEWAY_NAME`	(required)	Gateway resource name for HTTPRoute
`GATEWAY_NAMESPACE`	`default`	Namespace of the Gateway resource
`HOSTNAME`	(required)	HTTPRoute hostname prefix
`DOMAIN`	(required)	HTTPRoute domain suffix

Useful PromQL Queries

PVC usage (GiB):

kubelet_volume_stats_used_bytes / 1024 / 1024 / 1024

PVC usage percentage:

kubelet_volume_stats_used_bytes / kubelet_volume_stats_capacity_bytes * 100

Node CPU usage:

100 - (avg by(instance) (irate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)

Node memory usage:

(1 - node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes) * 100

Notes

Uses HelmRepository from https://prometheus-community.github.io/helm-charts
Standalone Prometheus (no Grafana/Alertmanager included)
Uses Gateway API HTTPRoute for UI access