Skip to content

Latest commit

 

History

History

ipi

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
.env
.env
 
 
README.md
README.md
 
 
create-dns-delegation.sh
create-dns-delegation.sh
 
 
deploy.sh
deploy.sh
 
 
destroy.sh
destroy.sh
 
 
install-config.yaml.tpl
install-config.yaml.tpl
 
 

README.md

OpenShift on AWS with Installer-Provisioned Infrastructure (IPI)

Provision OpenShift clusters. Allow installer to provision infrastructure.

Deploy cluster

  1. Install AWS and openshift-install CLIs, and jq.
  2. Set AWS access key secrets as environment variables in .env.
  3. Get a pull secret for Red Hat's container registries from https://console.redhat.com/openshift/downloads#tool-pull-secret. Set this as OPENSHIFT_PULL_SECRET in .env.
  4. Set up a subdomain in AWS (see docs). Set its name as BASE_DOMAIN in .env.
  1. Set CLUSTER_NAME and AWS_REGION in .env or rely on defaults.
  2. Review cluster config in install-config.yaml.tpl.
  3. Run deploy.sh to install cluster.

Run destroy.sh to destroy the cluster and AWS resources.

Use cluster

Cluster state will be saved in this directory at ./_workdir/. Rename or move this directory to retain it so that you can later destroy the cluster or get credentials for it. It uses Terraform and relies on its tfstate file.

deploy.sh will exit if a workdir already exists. Configure it to overwrite the current workdir by setting env var OVERWRITE=1; configure it to resume a previous installation by setting env var RESUME=1.

Credentials for the cluster will be echoed to stdout at the end of the installation. You can also find these credentials in the workdir at ./_workdir/auth.

To use the credentials from the workdir, set export KUBECONFIG=$(pwd)/_workdir/auth/kubeconfig and then use kubectl or oc as usual.

The cluster will be accessible at this URL: https://console-openshift-console.apps.${CLUSTER_NAME}.${BASE_DOMAIN}, e.g. https://console-openshift-console.apps.ipi.aws.joshgav.com/.