• k3s-paasmonkeys
  • Why k3s
  • Prerequisites
  • Getting started
    • Choose your Cluster node names
    • Flash SD Card
    • Customize the Raspberry
      • Set static IP
      • Enable container features
      • Add your ssh key
  • Create the cluster
    • Set the Master
    • Join Node
  • Api Gateway
• Credits:
• License

k3s-paasmonkeys

This is a personal project to set a Kubernetes RPi cluster to provide me a private cloud for learning.

Why k3s

K3S is a highly available, certified Kubernetes distribution designed for production workloads in unattended, resource-constrained, remote locations or inside IoT appliances. Both ARM64 and ARMv7 are supported with binaries and multiarch images available for both.

Prerequisites

• 6 SD card 32GB like SanDisk 32GB ULTRA microSDHC Card Class 10
• 6 RPi 3b+ (1 Masters and 5 Worker)
• 8-port Ethernet Gigabit Switch like Netgear GS308
• 8-way USB Power-supply
• 6 Ethernet cables
• 6 Micro-USB cables
• RPi 3B+,4B Cluster case like 8 Slot Cluster Cloud

Getting started

For every RPi that we going to have in our cluster. We have to follow these steps:

Choose your Cluster node names

I recommend you follow some kind of nomenclature in your nodes naming. It'll be helpful later when you are working with your cluster. Because of the high availability feature with an embedded DB is experimental. I decided only to use one Master node in my cluster.

Flash SD Card

• Download Raspbian Lite
• Flash SD card using Etcher
• Mount the SD card and create a text file named "ssh" in the boot partition.

Customize the Raspberry

Insert the SD card and turn on your RPi. It will be accessible on your network over ssh using the following command:

ssh [email protected]

Log in with the password "raspberry" and then type:

sudo raspi-config

Do the following actions:

• Set the GPU memory split to 16mb
• Set the hostname
• Change the password for the pi user

Set static IP

ifconfig | grep -i inet
sudo nano /etc/dhcpcd.conf

Add to end of this file:

interface eth0
static ip_address=192.168.2.xxx/24
static routers=192.168.2.1
static domain_name_servers=192.168.2.1

Enable container features

edit /boot/cmdline.txt

sudo nano /boot/cmdline.txt

and add the following to the end of the line:

cgroup_enable=cpuset cgroup_memory=1 cgroup_enable=memory

Now reboot the device:

sudo shutdown -r now

Add your ssh key

Copy public key to the RPi

ssh-copy-id [email protected]

Now you can rely on your public key to log into each RPi without typing a password in.

Create the cluster

Because of the high availability feature with an embedded DB is experimental. I decided only to use one Master node in my cluster.

Set the Master

SSH to you our RPi Master

ssh [email protected]

Install our Master node in our RPi

export SERVER_IP=192.168.2.120
export USER=pi
k3sup install --ip $SERVER_IP --user $USER

Join Node

SSH to our RPi Master

ssh [email protected]

Copy public key to the Node

ssh-copy-id [email protected]

Join RPi Node to the RPiCluster

export AGENT_IP=192.168.2.xxx
export SERVER_IP=192.168.2.120
export USER=pi
k3sup join --ip $AGENT_IP --server-ip $SERVER_IP --user $USER

Api Gateway

Here you can see how deploy an API Gateway in our cluster.

Credits:

• Will it cluster? k3s on your Raspberry Pi
• Inlets Operator — Exposing Services on Private Kubernetes Clusters with a Public IP

License

See the LICENSE file for license text and copyright information.