Skip to content

Latest commit

 

History

History
116 lines (81 loc) · 3.95 KB

README.md

File metadata and controls

116 lines (81 loc) · 3.95 KB
Raw

Clusterlust Documentation

Clusterlust facilitates the deployment of Kubernetes clusters using the Kubespray Ansible collection. It provides a structured approach to setting up and managing Kubernetes environments.

Key Components

  • Ansible Playbooks: Automate the deployment and configuration of Kubernetes clusters.
  • Inventory Management: Define and manage the nodes within your cluster.
  • Group Definitions: Specify the groups and responsibilities of each node (control plane, workers, bastion).

Getting Started

1. Clone the Repository

cd
git clone https://github.com/playingfield/clusterlust.git
cd clusterlust
git checkout -b yourcluster

2. Install Dependencies

Run the included setup scripts to install Ansible in a python virtualenv, and install required collections:

source ansible.sh
./prepare.sh

3. Configure Inventory

In the inventory directory there are several examples of different layouts for Kubernetes clusters. If you'd like to experiment with 'dev', one master and one worker, then copy the directory and adjust IP addresses and such:

cd inventory
cp -R dev yourcluster
...
cd ..

Edit the inventory/yourcluster/hosts file to list your cluster nodes' IP addresses. Edit the inventory/yourcluster/group_vars/all/vars.yml to select features in your cluster.

For detailed information, refer to KUBESPRAY.md or https://kubespray.io/#/.

Run the host-initialization.yml playbook to prep the machines.

4. Manage the Cluster

Run the cluster.yml playbook to deploy the cluster:

./cluster.yml -i inventory/mycluster/hosts --become -K

Run the reset playbook to delete the cluster:

./reset.yml -i inventory/mycluster/hosts --become -K

Post-Deployment

After deployment, run the kube-config.yml playbook. NOTE: ~/.kube/config is copied to the user running that playbook, and the cluster_admin, if defined, else to the root user too.

Use k9s as tool to work with the cluster.

# Start K9s in readonly mode - with all modification commands disabled
k9s --readonly

Troubleshooting

  • Node Issues: Verify network connectivity and SSH access to all nodes.
ansible all -m ping -i inventory/yourcluster/hosts
  • Deployment Failures: Check the Ansible playbook output for errors and consult the logs on the affected nodes.

Development

We deploy the Kubernetes clusters using Ansible, you'll need a Hypervisor to host your Kubernetes machine(s), and linux to run Ansible. Doing this on a Windows laptop is possible, but takes a bit more care than on Linux/macOS.

Hypervisor

Preferred local hypervisors are VMWare Desktop, or Virtualbox. When you are restricted by policy or firewall to install, but you are a local admin, you can enable Hyper-V and WSL, both Windows components. You'll need to enable using SecureBoot in Hyper-V for linux, otherwise they won't boot at all. On-prem hypervisors can be used as well, VMWare and Proxmox are preferred. You can run this in the cloud when you have two virtual machines scale-sets.

Hyper-V

Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V -All
Set-VMFirmware -VMName "k8s-node" -EnableSecureBoot On -SecureBootTemplate "MicrosoftUEFICertificateAuthority"

Vagrant

Use Vagrant to have an abstract interface to your local hypervisor. Install Vagrant from a download from HashiCorp, or with choco.

choco install vagrant -y

Please note that some security policies might block binaries run from %appdata%. E.g. choco installed software. To bypass this, download the specific binary from the website yourself.

Windows Sybsystem Linux

WSL allows you to run AlmaLinux on your Windows laptop. First install msixbundle, then:

wsl --install
winget install 'AlmaLinux 8 WSL'
wsl --set-default AlmaLinux-8