OpenShift Ansible playbooks to provision infrastructure on a cloud platform
To run deployment fill out vars.yml and files/pull-secret.txt, and then run:
./deploy.sh
The rhcos image and floating IP creation are not currently part of the code and need to be created/uploaded, and then referenced in vars.yml
Tested with the versions in pip-requirements.txt ...
pip install -r pip-requirements.txt
oc delete project ukc-ingress
oc delete MachineSets --all -n openshift-machine-api
oc scale IngressControllers --all --replicas=0 -n openshift-ingress-operator
oc delete pdb prometheus-adapter thanos-querier-pdb alertmanager-main prometheus-k8s -n openshift-monitoring
oc delete pdb prometheus-user-workload thanos-ruler-user-workload -n openshift-user-workload-monitoring
# Wait for workers/infra/net2 to vanish - last worker takes ages...
watch -n5 oc get nodes
ansible-playbook -i ./inventory.yaml ./down-compute-nodes.yaml
ansible-playbook -i ./inventory.yaml ./down-control-plane.yaml
ansible-playbook -e @vars.yml -i ./inventory.yaml ./down-bootstrap.yaml
ansible-playbook -i ./inventory.yaml ./down-loadbalancers.yaml
# The down-network playbook needs to be run twice due to ha_router port removal occasionally failing for net2/eg deployments
ansible-playbook -e @vars.yml -i ./inventory.yaml ./down-network.yaml
ansible-playbook -e @vars.yml -i ./inventory.yaml ./down-network.yaml
ansible-playbook -e @vars.yml -i ./inventory.yaml ./down-server-groups.yaml
ansible-playbook -i ./inventory.yaml ./down-security-groups.yaml
Destroying state in directory to prepare for a fresh test install (at your own risk - for testing only!):
rm *.ign *.json inventory.yaml openshift_ke* .openshift_install_state.json; rm -rf ./auth