Skip to content

Latest commit

 

History

History
99 lines (73 loc) · 3.72 KB

File metadata and controls

99 lines (73 loc) · 3.72 KB

Turing RK1 Kubernetes Cluster Documentation

This repository contains documentation and configuration for running a 4-node Kubernetes cluster on Turing Pi 2.5 with RK1 compute modules.

Installation Guides

Choose your preferred Kubernetes distribution:

Guide OS Kubernetes Best For
INSTALLATION.md Talos Linux Talos K8s Production, Security
INSTALLATION-K3S.md Armbian K3s Development, AI/ML

Quick Comparison

Feature Talos K3s on Armbian
Shell Access No (API only) Yes (SSH)
Security Hardened by default Manual hardening
NPU (6 TOPS) Not available Supported
GPU (Mali-G610) Not available Supported
Debugging talosctl Standard Linux
Updates Atomic apt + reinstall
Learning Curve Steeper Gentler

See COMPARISON.md for the complete feature matrix.

Cluster Architecture

┌─────────────────────────────────────────────────────────┐
│                  Turing Pi 2.5 Board                     │
├─────────────┬─────────────┬─────────────┬─────────────┬─┤
│   Node 1    │   Node 2    │   Node 3    │   Node 4    │B│
│ Control     │   Worker    │   Worker    │   Worker    │M│
│ Plane       │             │             │             │C│
│ 10.10.88.73 │ 10.10.88.74 │ 10.10.88.75 │ 10.10.88.76 │ │
├─────────────┼─────────────┼─────────────┼─────────────┤ │
│ 32GB eMMC   │ 32GB eMMC   │ 32GB eMMC   │ 32GB eMMC   │ │
│ 500GB NVMe  │ 500GB NVMe  │ 500GB NVMe  │ 500GB NVMe  │ │
└─────────────┴─────────────┴─────────────┴─────────────┴─┘

Common Components (Both Distributions)

Both installation methods deploy the same workloads:

  • MetalLB - Load balancer (IP pool: 10.10.88.80-99)
  • NGINX Ingress - Ingress controller at 10.10.88.80
  • Longhorn - Distributed storage on NVMe drives
  • Prometheus Stack - Monitoring (Grafana, Prometheus, Alertmanager)

Access URLs

Add to /etc/hosts:

10.10.88.80  grafana.local prometheus.local alertmanager.local longhorn.local
Service URL Credentials
Grafana http://grafana.local admin / admin
Prometheus http://prometheus.local -
Alertmanager http://alertmanager.local -

Additional Documentation

Scripts

Helper scripts in ../scripts/:

Script Purpose
setup-k3s-node.sh Prepare Armbian node for K3s
deploy-k3s-cluster.sh Deploy K3s cluster from workstation
wipe-cluster.sh Reset cluster for distribution switch

Switching Distributions

Use wipe-cluster.sh to reset nodes when switching between Talos and K3s:

# View cluster status
./scripts/wipe-cluster.sh status

# Reset Talos cluster
./scripts/wipe-cluster.sh talos

# Uninstall K3s from Armbian
./scripts/wipe-cluster.sh k3s

# Full reset (power off for reflash)
./scripts/wipe-cluster.sh full