Curated collection of Grafana dashboards for monitoring the Goldentooth Raspberry Pi cluster infrastructure.

This repository contains production-ready Grafana dashboards designed specifically for monitoring a Raspberry Pi cluster running Kubernetes, HashiCorp stack (Consul, Nomad, Vault), and various observability tools.

Comprehensive overview of HashiCorp services:

• Consul: Cluster membership, service health, KV store metrics
• Nomad: Job status, resource allocation, client health
• Vault: Secret engine metrics, authentication status, token usage

High-level infrastructure monitoring:

• Node Health: CPU, memory, disk usage across all Pi nodes
• Network Performance: Latency, bandwidth, connectivity status
• Service Availability: Uptime metrics for critical cluster services
• Storage: NFS exports, ZFS pool status, disk I/O

Detailed system metrics from node_exporter:

• Hardware Monitoring: Temperature, voltage, frequency scaling
• Resource Utilization: Per-node CPU, memory, disk, network
• Process Monitoring: System load, context switches, interrupts
• Filesystem Details: Mount points, inode usage, disk space

SLURM workload manager dashboard:

• Job Queue: Pending, running, completed job metrics
• Partition Status: Node allocation, resource availability
• User Activity: Job submission patterns, resource consumption
• Cluster Efficiency: Utilization rates, queue times

These dashboards are automatically provisioned through the Goldentooth Ansible role:

goldentooth setup_grafana

They integrate with:

• Prometheus: Primary metrics collection
• Node Exporter: System-level metrics
• Blackbox Exporter: Service availability monitoring
• Custom Exporters: SLURM, HashiCorp services

Dashboards are automatically deployed via Ansible to /var/lib/grafana/dashboards/ and configured through provisioning files. Updates are applied through the cluster management pipeline.

Each dashboard includes:

• Variable Templating: Node selection, time ranges, service filters
• Alert Annotations: Integration with Prometheus AlertManager
• Panel Descriptions: Detailed explanations of metrics and thresholds
• Responsive Layout: Optimized for different screen sizes

For cluster-specific customizations, modify the dashboard JSON files and redeploy through Ansible.