80 lines
2.3 KiB
YAML
80 lines
2.3 KiB
YAML
# k8s/monitoring/blackbox-exporter-values.yaml
|
|
# Active black-box HTTP probing of every ingress-exposed service — gives an
|
|
# uptime/availability signal independent of real traffic. Homelab services
|
|
# like Vault/MinIO/Longhorn UI get almost no organic requests, so ingress RED
|
|
# metrics alone can't tell "idle" from "down"; this closes that gap.
|
|
|
|
config:
|
|
modules:
|
|
http_2xx:
|
|
prober: http
|
|
timeout: 5s
|
|
http:
|
|
valid_http_versions: ["HTTP/1.1", "HTTP/2.0"]
|
|
valid_status_codes: [] # any 2xx
|
|
follow_redirects: true
|
|
preferred_ip_protocol: "ip4"
|
|
tls_config:
|
|
insecure_skip_verify: true # homelab-ca is a private CA; skip verify for simplicity
|
|
|
|
resources:
|
|
requests:
|
|
cpu: 20m
|
|
memory: 32Mi
|
|
limits:
|
|
cpu: 100m
|
|
memory: 64Mi
|
|
|
|
tolerations:
|
|
- key: node-role.kubernetes.io/control-plane
|
|
operator: Exists
|
|
effect: NoSchedule
|
|
|
|
serviceMonitor:
|
|
enabled: true
|
|
selfMonitor:
|
|
enabled: true
|
|
defaults:
|
|
interval: 30s
|
|
scrapeTimeout: 10s
|
|
module: http_2xx
|
|
targets:
|
|
- name: grafana
|
|
url: https://grafana.riotpiao.com/
|
|
- name: loki
|
|
url: https://loki.riotpiao.com/ready
|
|
- name: authentik
|
|
url: https://authentik.riotpiao.com/-/health/ready/
|
|
- name: vault
|
|
url: https://vault.riotpiao.com/v1/sys/health
|
|
- name: minio-console
|
|
url: https://minio.riotpiao.com/
|
|
- name: minio-api
|
|
url: https://minio-api.riotpiao.com/minio/health/live
|
|
- name: prometheus
|
|
url: https://prometheus.riotpiao.com/-/healthy
|
|
- name: portainer
|
|
url: https://portainer.riotpiao.com/
|
|
- name: forgejo
|
|
url: https://forgejo.riotpiao.com/api/healthz
|
|
- name: argocd
|
|
url: https://argocd.riotpiao.com/healthz
|
|
- name: longhorn
|
|
url: https://longhorn.riotpiao.com/
|
|
- name: api-gateway
|
|
url: https://api.riotpiao.com/healthz
|
|
- name: api-gateway-models
|
|
url: https://api.riotpiao.com/v1/models
|
|
|
|
prometheusRule:
|
|
enabled: true
|
|
rules:
|
|
- alert: ServiceProbeDown
|
|
expr: probe_success == 0
|
|
for: 5m
|
|
labels:
|
|
severity: critical
|
|
annotations:
|
|
summary: "Probe failing for {{ $labels.instance }}"
|
|
description: "Blackbox probe for {{ $labels.instance }} has failed for more than 5 minutes."
|