Setup - Hetzner Bare Metal

Setting up a Kubernetes Cluster with Hetzner Dedicated Servers

This guide for Hetzner Bare Metal is WIP and not currently recommended. Proceed with caution until an update is released and this warning removed.

Checkout the repository source to manage a cluster of dedicated servers on Hetzner.

The scripts in this repository will set up and maintain one or more kubernetes clusters consisting of dedicated Hetzner servers. Each cluster will also be provisioned to operate as a node in the MAYACHain network.

Executing the scripts in combination with some manual procedures will get you highly available, secure clusters with the following features on bare metal.

• Kubespray (based)

• Internal NVMe storage (Ceph/Rook)

• Virtual LAN (also over multiple locations) (Calico)

• Load Balancing (MetalLB)

Preparations

Servers

Acquire a couple of servers as the basis for a cluster (AX41-NVME's are working well, for instance). Visit the admin panel and name the servers appropriately.

ma-k8s-node1
ma-k8s-node2
ma-k8s-node3
...

ma-k8s-master1
ma-k8s-master2
ma-k8s-worker1
ma-k8s-worker2
ma-k8s-worker3
...

Refer to the reset procedure to initialize them properly.

vSwitch

Create a vSwitch and order an appropriate subnet (it may take a while to show up after the order). Give the vSwitch a name (i.e. ma-k8s-net) and assign this vSwitch to the servers.

Check out the docs for help.

Usage

Clone this repository cd into it and download kubespray.

git submodule init && git submodule update

Create a Python virtual environment or similar.

# Optional
virtualenv -p python3 venv

Install dependencies required by Python and Ansible Glaxy.

pip install -r requirements.python.txt
ansible-galaxy install -r requirements.ansible.yml

Note: Mitogen does not work with ansible collections and the strategy must be changed (i.e. strategy: linear).

Provisioning

Create a deployment environment inventory file for each cluster you want to manage.

cp hosts.example inventory/production.yml
cp hosts.example inventory/test.yml
cp hosts.example inventory/environment.yml
...

cp hosts.example inventory/production-01.yml
cp hosts.example inventory/production-02.yml
...

cp hosts.example inventory/production-helsinki.yml
cp hosts.example inventory/whatever.yml

Edit the inventory file with your server ip's and network information and customize everything to your needs.

# Manage a cluster
ansible-playbook cluster.init.yml -i inventory/environment.yml
ansible-playbook --become --become-user=root kubespray/cluster.yml -i inventory/environment.yml
ansible-playbook cluster.finish.yml -i inventory/environment.yml

# Run custom playbooks
ansible-playbook private-cluster.yml -i inventory/environment.yml
ansible-playbook private-test-cluster.yml -i inventory/environment.yml
ansible-playbook private-whatever-cluster.yml -i inventory/environment.yml

Check this out for more playbooks on cluster management.

MAYAChain

For the cluster to operate as a node in the MAYACHain network, deploy as instructed here. You can also refer to the node-launcher repository, if necessary, or the MAYAChain documentation as a whole.

Resetting the bare metal servers

This will install and use Ubuntu 20.04 on only one of the two internal NVMe drives. The unused ones will be used for persistent storage with ceph/rook. You can check the internal drive setup with lsblk. Change it accordingly in the command shown above when necessary.

Manually

Visit the console and put each server of the cluster into rescue mode. Then execute the following script.

installimage -a -r no -i images/Ubuntu-2004-focal-64-minimal.tar.gz -p /:ext4:all -d nvme0n1 -f yes -t yes -n hostname

Automatically

Create a pristine state by running the playbooks in sequence.

ansible-playbook server.rescue.yml -i inventory/environment.yml
ansible-playbook server.bootstrap.yml -i inventory/environment.yml

Instantiation

Instantiate the servers.

ansible-playbook server.instantiate.yml -i inventory/environment.yml