Monitoring Stack

The monitoring stack provides real-time visibility into cluster health and performance.

Namespace: monitoring
Chart: kube-prometheus-stack

---

1. Overview

Cluster health and performance metrics are aggregated via the Prometheus stack on the single ARM64 node.

---

2. Components

• Prometheus: Time-series database for metric collection.
• Grafana: Dashboard visualization for infra metrics.
• Alertmanager: Notification routing based on defined alert rules.

---

3. Configuration and Deployment

Implementation procedures:

1. Installation: Managed via ArgoCD using kube-prometheus-stack 69.3.0.
2. Persistence: Prometheus uses a 5 GiB local-path PVC; Grafana uses 1 GiB local-path; Alertmanager uses 500 MiB local-path.
3. Retention: Prometheus retention is configured for one day.
4. Dashboards: Pre-configured dashboards for Node Exporter and Kubernetes resources are imported.

---

4. Operational Maintenance

1. Rule Management: Alerting rules are updated via GitOps.
2. Metric Retention: One-day retention is enforced for the node-bound Prometheus volume.
3. Alerting: Critical infrastructure alerts are routed to Discord.