OpenShift 4 Automated Cluster Installation (UPI on KVM) using Ansible

This automation helps you create a Single Node OpenShift (SNO) or 3-node cluster on a KVM host using Libvirt. It's possible to configure OpenShift Data Foundation (ODF) and OpenShift Virtualization when 3-node cluster.

⚠️ Important Notice

Please be sure to exercise caution when running this playbook, as the initial tasks modify network settings. Ensure that you fully understand the changes being made and have appropriate backups or recovery plans in place before proceeding. Running these tasks without careful review may lead to connectivity issues or unintended network configuration changes.

Tested Environments

This playbook has been successfully tested on the following distributions:

Fedora 40
RHEL 8
RHEL 9

Please make sure that your environment meets these system requirements for the best results.

Prerequisites

User

You must run this playbook with sudo passwordless ( do not use 'root' user)

Install Ansible Core

To install ansible-core, run the following command:

sudo dnf install -y ansible-core

Generate an SSH Key

If you don't already have an SSH key, generate one using:

ssh-keygen -t rsa

Edit Ansible Variables

Before running the playbook, edit the Ansible variables file and fill in the required parameters, such as sshrsa and pullsecret. You can obtain the pull secret by accessing Red Hat OpenShift Local (https://console.redhat.com/openshift/create/local) and clicking "Download pull secret".

Variable	Value (Example)	Description
clustername	funny-flanders	The name of the cluster
basedomain	chiaret.to	Base domain of the cluster (do not use .local)
ocpversion	4.16.16	OpenShift version
clusters_dir	"/labs"	Directory to create cluster files
sno	'true'	Defines the type of cluster to create. Use 'true' for Single Node OpenShift and 'false' for a 3-node cluster.
destroy_if_exists	'true' or 'false'	Removes the existing cluster before creating a new one.
kvmnetwork	default	Network configured on KVM
worker_mem	'4096'	Worker memory size in MB
worker_cpu	4	Worker CPU (int)
n_worker	0	Int
master_mem	'32000'	Memory
master_cpu	8	Int
extra_disks	0	Number of extra disks to add
extra_disk_size	100	Extra disk dize in GB
installodf	'true' or 'false'	true if the cluster is a 3-node and ODF will be configured
installocpvirt	'true' or 'false'	true if the cluster is a 3-node with ODF and OpenShift Virtualization will be configured
local_quay_registry	''	Local Quay URL to use as local mirror for OCP (if Red Hat Quay is configured)
admin_user	"admin"	Configures a user to access the cluster using HTPasswd. Empty Value "" means the user will not be configured
htpasswd_pass	'Redhat@123'	Password to admin_user
remove_kubeadmin_user	'true' or 'false'	Remove kubeadmin user after install the cluster
ssh_rsa	ssh-rsa AAAAB3Nza...	SSH Pubkey to access OCP nodes over SSH
pullsecret	'{"auths":{"cloud.op..}}'	pull secret to download OpenShift images

vim ansible-vars-kvm.yaml

Install Required Ansible Collections

Some Ansible collections are required to run this playbook. Install them by running:

ansible-galaxy collection install -r requirements.yml

Usage

Creating the cluster

Once the prerequisites are met and variables are configured, run the playbook as follows:

ansible-playbook -e @ansible-vars-kvm.yaml create-cluster-upi-kvm.yaml

Accessing the cluster

The access to the cluster can be using the kubeconfig file generated on '{{ clusters_dir }}/{{ clustername }}/auth/kubeconfig'

export KUBECONFIG='{{ clusters_dir }}/{{ clustername }}/auth/kubeconfig'

Day 2 Operations

Add new worker nodes

To add new nodes after installing the cluster you can use the playbook add-new-nodes.yaml

Edit the variables file and configure n_worker value

vim ansible-vars-kvm.yaml

Execute the playbook

ansible-playbook -e @ansible-vars-kvm.yaml add-new-nodes.yaml

Configure LVM Storage

To configure LVM storage using the Logical Volume Manager (LVM) Operator after the cluster is up, you can use the playbook d2-lvm-storage.yaml.

Prerequisites:

Ensure that your cluster was created with at least 1 extra disk and the ansible-vars-kvm.yaml file is correctly configured, including the following parameters:

• extra_disks: Number of extra disks you have configured for extra disks creation. The playbook will automatically generate the corresponding disk names (e.g., vdb, vdc).

Steps:

1. Execute the playbook:

ansible-playbook -e @ansible-vars-kvm.yaml d2-lvm-storage.yaml

This playbook performs the following tasks:

• Creates the necessary namespace (openshift-storage).
• Installs the LVM Operator from the Red Hat Operator Hub.
• Configures the LVMCluster resource based on the available extra disks.
• Creates a test PersistentVolumeClaim (PVC) and a Pod in the default namespace to verify the storage functionality.