# Kubernetes cluster setup using Ansible for Debian

## Description

The project configures a Kubernetes cluster on Debian machines with kubeadm.\
It uses systemd-networkd for interface configuraition, CRI-O for containerization, Calico as a CNI plugin, and stacked etcd for control-plane database.

## Configuration

- General node config: [*roles/node/defaults/main.yml*](roles/node/defaults/main.yml)
- Cluster config:
  - [*roles/init-cluster/defaults/main.yml*](roles/init-cluster/defaults/main.yml)
  - [*group_vars/main.yml*](group_vars/main.yml)

## Usage

### 1. Configure ssh on all nodes

1. ssh in and enable root login by editing */etc/ssh/sshd_conf*
2. `ssh-copy-key root@NODE`
3. `ssh-agent $SHELL`
4. `ssh-add`

### 2. Add nodes to inventory

Edit the *inventory* file.

### 3. Prepare nodes

`ansible-playbook -i inventory -u root node.yml`

### 4. Initalize cluster

Uses the `[first_master]` group from the *inventory* to select the first master.

`ansible-playbook -i inventory -u root init-cluster.yml`

### 5. Join other masters

Uses the `[other_masters]` group from the *inventory*.\
Do it **ONE-by-ONE**.

`ansible-playbook -i inventory -u root join-control-plane.yml`

### 6. Join workers

Uses the `[workers]` group from the *inventory*.\
Do it **ONE-by-ONE**.

`ansible-playbook -i inventory -u root join-workers.yml`