Kube Prometheus Stack¶

Namespace: monitoring | Sync Wave: 3 | Chart: kube-prometheus-stack v82.0.0

Deploys the Prometheus monitoring stack for cluster observability. Configured for metrics collection and remote write — Grafana and Alertmanager are disabled.

Upstream Chart¶

Chart: kube-prometheus-stack
Version: 82.0.0
Repository: https://prometheus-community.github.io/helm-charts

Key Configuration¶

Base values:

CRD installation disabled (provided by Prometheus Operator CRDs)
Grafana: disabled
Alertmanager: disabled
Node Exporter: enabled
Kube State Metrics: enabled
Scrape interval: 30s
Evaluation interval: 30s
Remote write endpoint configured
All control plane component monitoring enabled (kubelet, API server, etcd, scheduler, controller-manager, CoreDNS, kube-proxy)

Per-environment resources:

Setting	Dev	Staging	Prod
Prometheus replicas	1	1	2
Retention	2h	4h	6h
CPU request	100m	200m	500m
Memory request	256Mi	512Mi	1Gi
CPU limit	200m	500m	1000m
Memory limit	512Mi	1Gi	2Gi

Sync Configuration¶

The ArgoCD Application ignores diff on status fields for Prometheus and Alertmanager CRs, which are frequently updated by the operator and would otherwise cause perpetual out-of-sync states.

Dependencies¶

Prometheus Operator CRDs (wave 2) must be installed before this component (wave 3)

Files¶

components/kube-prometheus-stack/
├── Chart.yaml
└── values/
    ├── base.yaml
    ├── dev.yaml
    ├── staging.yaml
    └── prod.yaml