The scale-ci-deploy repo is split into two major directories, playbooks for OpenShift 4 and playbooks for OpenShift 3.
| Environment | IPI Install | Scale-up | Upgrade | Cleanup / Destroy |
|---|---|---|---|---|
| AWS | Yes | Yes | Yes | Variable OPENSHIFT_CLEANUP |
| Azure | Yes | Yes | Yes | Variable OPENSHIFT_CLEANUP |
| GCP | Yes | Yes | No | Variable OPENSHIFT_CLEANUP |
| OpenStack | Yes | No | Yes | Separate Playbook |
See the OpenShift 4 scale document for the scale playbook documentation.
See the OpenShift 4 upgrade document for the upgrade playbook documentation.
| Environment | Install | Scale-up |
|---|---|---|
| OpenStack | Yes | Yes |
For Jenkins setup, each install has a Jenkins job implemented in scale-ci-pipeline
To implement in your own CI, use a SCM plugin/module to clone the code into a build step. Each install variable in the docs should be implemented as a parameter for each job. An example build step for OCP 4 IPI AWS install as implemented in Jenkins would look like:
set -o pipefail
set -eux
echo "[orchestration]" > inventory
echo "${ORCHESTRATION_HOST}" >> inventory
ansible-playbook -vv -i inventory OCP-4.X/install-on-aws.ymlWhat is Cerberus? Cerberus is a project that will watch an Openshift/Kubernernetes cluster for dead nodes, component failures, etc and provide a healthly/unhealthy (True/False) signal.
Cerberus is optionally used for all install/scale/upgrades by defining the CERBERUS_URL variable in the format "http://1.2.3.4:8080"
For installation and startup instructions for Cerberus please see https://github.com/openshift-scale/cerberus