Skip to main content

Hardware

The Anton cluster runs on three MS-01 mini PCs, providing a compact yet powerful foundation for the Kubernetes homelab.

Node Specifications

MS-01 Mini PCs

  • CPU: Intel N100 (4-core, 4-thread)
  • RAM: 16GB DDR4 per node
  • Storage: 2x 1TB NVMe SSDs per node (6TB total raw)
  • Network: Gigabit Ethernet
  • Power: ~10W per node under load

Cluster Topology

Storage Configuration

Distributed Storage

  • Total Capacity: 6x 1TB NVMe SSDs
  • Replication: 3-way replication across nodes
  • Usable Space: ~2TB (with replication overhead)
  • Storage Classes: Ceph Block, Ceph Filesystem, Ceph Bucket

Per-Node Layout

Each MS-01 Node:
├── System Disk: 1TB NVMe (Talos Linux)
└── Data Disk: 1TB NVMe (Ceph OSD)

Power and Environmental

Power Consumption

  • Idle: ~6W per node (18W total)
  • Load: ~10W per node (30W total)
  • Maximum: ~15W per node (45W total)

Physical Layout

  • Form Factor: Compact mini PCs
  • Cooling: Passive cooling with heat sinks
  • Noise: Silent operation
  • Placement: Desktop/shelf mounting

Network Architecture

Internal Networking

  • Pod Network: 10.244.0.0/16 (Cilium)
  • Service Network: 10.96.0.0/12
  • Node Network: 192.168.x.0/24 (private subnet)

Connectivity

  • Uplink: Gigabit Ethernet per node
  • Inter-node: Full mesh connectivity
  • External Access: Cloudflare Tunnel

Monitoring Commands

# Check node status and resources
kubectl get nodes -o wide

# View node resource usage
kubectl top nodes

# Check node conditions
kubectl describe nodes

# View hardware info (requires privileged access)
kubectl get nodes -o yaml | grep -A5 "system-info"

Node Management

Talos Commands

# Check node status
talosctl health

# View system resources
talosctl usage

# Check disk usage
talosctl df

# View system logs
talosctl logs

Hardware Monitoring

# Check storage usage across nodes
kubectl get pv

# View node resource allocation
kubectl describe node k8s-1

# Check cluster resource summary
kubectl cluster-info

The MS-01 hardware provides an excellent balance of performance, power efficiency, and cost for a homelab Kubernetes cluster, delivering enterprise-grade features in a compact form factor.