sno_without_ai.md

On-prem Single Node OpenShift without Assisted Installer

Special thanks to Ben Schmaus and his amazing blog post.

We also discussed the process here during the Ask an OpenShift Admin live stream on Oct 20th.

1. Pre-reqs

   Download the tools we'll need.

   # the latest OpenShift 4.9 CLI binaries
   wget https://mirror.openshift.com/pub/openshift-v4/clients/ocp/stable-4.9/openshift-client-linux.tar.gz
   wget https://mirror.openshift.com/pub/openshift-v4/clients/ocp/stable-4.9/openshift-install-linux.tar.gz
   
   # unpackage them, optionally move to the path, clean up
   tar xzf openshift-client-linux.tar.gz
   tar xzf openshift-install-linux.tar.gz
   
   chmod 755 kubectl oc openshift-install
   sudo mv kubectl oc openshift-install /usr/local/bin/
   
   rm *.gz README.md
   
   # get the coreos-installer binary
   wget https://mirror.openshift.com/pub/openshift-v4/clients/coreos-installer/latest/coreos-installer
   chmod 755 coreos-installer
   sudo mv coreos-installer /usr/local/bin
   
   # get the RHCOS live ISO
   wget https://mirror.openshift.com/pub/openshift-v4/dependencies/rhcos/4.9/latest/rhcos-live.x86_64.iso

   A DHCP reservation needs to be created for the node or a static IP needs to be configured.

   Additionally, we will need four DNS records all pointing to the IP used for the node:

   {node-name}.domain.name
   api.clustername.domain.name
   api-int.clustername.domain.name
   *.apps.clustername.domain.name
   

2. Create an install-config.yaml

   Review each field, particularly networking.machineNetwork.cidr and BootstrapInPlace.InstallationDisk for your environment.

   cat << EOF > install-config.yaml
   apiVersion: v1
   baseDomain: work.lan
   metadata:
     name: sno
   networking:
     networkType: OVNKubernetes
     machineNetwork:
     - cidr: 192.168.14.0/24
   compute:
   - name: worker
     replicas: 0
   controlPlane:
     name: master
     replicas: 1
   platform:
     none: {}
   bootstrapInPlace:
     installationDisk: /dev/sda
   pullSecret: '{"auths":{"blah":{"auth":"blah"}}}'
   sshKey: |
     ssh-ed25519 keygoeshere

3. Create the Ignition config

   # make a copy of the install-config.yaml to use for this
   mkdir cluster && cp install-config.yaml cluster/
   
   # generate the ignition
   openshift-install create single-node-ignition-config --dir=cluster

4. Create an ISO with the Ignition embedded

   coreos-installer iso ignition embed -f -i cluster/bootstrap-in-place-for-live-iso.ign rhcos-live.x86_64.iso

   This step is technically optional, you can host the bootstrap on a webserver if desired and provide the location to the installer.

5. Boot the node to the ISO, install RHCOS

   If you're using a static IP, you can append the ip= parameters at the appropriate time. If you have very complex networking, you can boot to the live ISO environment and configure it using nmcli/nmtui, then install RHCOS with the --copy-network option. Otherwise, let it boot and install.

6. Wait for install to complete

   Once the node is booted and RHCOS installed, you can monitor it in the usual ways:

   • SSH using the core user, then monitor the bootkube service.
   • Use openshift-install wait-for bootstrap-complete.

   After a few minutes, bootstrapping will finish and the node will reboot. Monitor for progress in the normal way:

   openshift-install wait-for install-complete

   After bootstrap completes, you can also connect using the oc/kubectl command with the kubeconfig created by openshift-install.

This repo automate this process for you (complete flow is using libvirt VM)
https://github.com/eranco74/bootstrap-in-place-poc