This is a script to reproduce the problems we're experiencing with OVN-kubernetes on OCP4.8.
We are aware that this does not represent normal use of a cluster.
- Be logged in as a user with
cluster-admin
permissions. - Have Ansible installed.
- Have pip3 installed. (For installing the Python modules.)
- Have a writable registry configured within the cluster.
- (For the error to occure: ovn-kubernetes SDN.)
cd
to the same directory as this file.- Install Python requirements:
pip3 install --user -r requirements.txt
- Install Ansible requirements:
ansible-galaxy collection install -r requirements.yml
- Run the playbook:
ansible-playbook deploy-script.yml
- Run the script:
./run.sh
This playbook and script applies some changes to your cluster.
Changes made by the playbook:
- Creates a ServiceAccount within the
openshift
namespace. (I had to put it somewhere...) - Creates a ClusterRoleBinding ("
health-check-script-self-provisioner
"). - It configures config.ini in the
app
directory with your default clusterdomain, the API token of the SA and your API URL.
Changes made by the script:
- The script creates a namespace
health-check
and DELETES it afterwards. - It deploys an S2I app within that namespace.