Feature Overview (aka. Goal Summary)  

Overarching Goal
Move to using the upstream Cluster API (CAPI) in place of the current implementation of the Machine API for standalone Openshift.
Phase 1 & 2 covers implementing base functionality for CAPI.
Phase 2 also covers migrating MAPI resources to CAPI.

Phase 2 Goal:  

• Complete the design of the Cluster API (CAPI) architecture and build the core operator logic
• attach and detach of load balancers for internal and external load balancers for control plane machines on AWS, Azure, GCP and other relevant platforms
• manage the lifecycle of Cluster API components within OpenShift standalone clusters
• E2E tests

for Phase-1, incorporating the assets from different repositories to simplify asset management.

Background, and strategic fit

• Initially CAPI did not meet the requirements for cluster/machine management that OCP had the project has moved on, and CAPI is a better fit now and also has better community involvement.
• CAPI has much better community interaction than MAPI.
• Other projects are considering using CAPI and it would be cleaner to have one solution
• Long term it will allow us to add new features more easily in one place vs. doing this in multiple places.

Acceptance Criteria

There must be no negative effect to customers/users of the MAPI, this API must continue to be accessible to them though how it is implemented "under the covers" and if that implementation leverages CAPI is open