- CentOS 7
- ideally on a bare metal host
- user with sudo access
Create a config file based on the example and set values appropriate for your local environment.
$ cp config_example.sh config_${USER}.sh
CONFIG="config_user.sh" ./01_install_requirements.sh
CONFIG="config_user.sh" ./02_run_all_in_one.sh
CONFIG="config_user.sh" ./03_configure_undercloud.sh
CONFIG="config_user.sh" ./04_ocp_repo_sync.sh
CONFIG="config_user.sh" ./05_build_ocp_installer.sh
and finally, run the OpenShift installer to do a deployment on your local single node OpenStack deployment:
CONFIG="config_user.sh" ./06_run_ocp.sh
Once the installer is running and the VMs have been created, the following
script will add an /etc/hosts
entry for the floating IP of the service VM
hosting the API load balancer. This is required for the installer to be able
to look up the API hostname and talk to the API.
CONFIG="config_user.sh" ./expose_ocp_api.sh
You may need to provide further customization to your deployment, such as
limiting the number of master and worker nodes created to fit in your
development environment. You can do this by generating and editing the
install-config.yaml
file before launching the deployment.
CONFIG="config_user.sh" ./06_run_ocp.sh create install-config
- `${EDITOR} ocp/install-config.yaml
CONFIG="config_user.sh" ./06_run_ocp.sh
Once you have a complete working environment, you do not need to re-run all sripts. If you're making changes to the installer, your workflow would look like:
ocp_cleanup.sh
CONFIG=”config_user.sh” build_ocp_installer.sh
CONFIG=”config_user.sh” 06_run_ocp.sh