Live Sanitized Telemetry

Infrastructure Hardware & Cluster Telemetry

Hardware allocation, node role matrix, and cluster-wide resource utilization. Data is automatically exported from Prometheus via the sanitized Allowlist DTO.

Strict Sanitization & Air-Gapped IsolationThis public showcase is statically compiled. All internal cluster IP subnets (10.42.x.x, 10.43.x.x), node external IPs, SSH ports, auth tokens, and private hostnames are completely stripped prior to static HTML export. Zero live connections exist between public visitors and our production cluster.

Average CPU Load

34.7%

Memory Allocation

70.3%

Storage Pools

0.91 TB

Fast NVMe / SSD local storage across 4 nodes

Multi-Node Hardware Matrix

Node 1 (Control & CloudPanel)

Ready

Control Plane • etcd • CloudPanel NGINX Host

OSUbuntu 22.04 LTS

CPU / RAM6 vCPU • 12 GB

Storage200 GB SSD

Core Workloads: CloudPanel ~30 VHosts, Geocrop API, TiTiler, Redis

Node 2 (Ingress & Perimeter)

Ready

Control Plane • etcd • Ingress-NGINX Gateway

OSRocky Linux 9.7

CPU / RAM6 vCPU • 12 GB

Storage400 GB SSD

Core Workloads: Ingress-NGINX, cert-manager, Ntfy, Cockpit UI

Node 3 (Control & Workloads)

Ready

Control Plane • etcd • Core Services

OSUbuntu 22.04 LTS

CPU / RAM6 vCPU • 12 GB

Storage200 GB SSD

Core Workloads: Prometheus Server, Grafana, Postgres StatefulSet

Node 4 (Dedicated Worker)

Ready

Dedicated Worker • ML & Compute Workloads

OSRocky Linux 10.1

CPU / RAM4 vCPU • 8 GB

Storage150 GB SSD

Core Workloads: Authentik Server & Worker, GeoCrop ML Workers, TiTiler cache