Feature Description

Enable comprehensive support for restoring hosted control planes to a different management cluster (Scenario 3 from CNTRLPLANE-316). While this capability was introduced in OpenShift 4.19, it currently has limited support (bare metal only) with provider-dependent limitations, missing E2E test coverage, and incomplete node readoption support.

Current State

Restoring to a different management cluster has the following limitations:

• Kubevirt: Not compatible
• BareMetal: Works, has node readoption (lacks E2E testing)
• AWS/ROSA: Works, does NOT have node readoption (lacks E2E testing)
• Azure/ARO: Under development (compatibility layer mostly working, still collaborating with Microsoft for full support), does NOT have node readoption (lacks E2E testing)
• PowerVS: Should work fine, does NOT have node readoption (lacks E2E testing)
• OpenStack: Should work fine, does NOT have node readoption (lacks E2E testing) - Note: stale provider with limited maintenance dependant on OpenStack team.

Node Readoption

Node readoption enables CAPI to properly adopt nodes from the original nodepools without requiring reprovisioning. Without node readoption, nodes exist and are joined to the new HostedCluster but cannot be adopted by CAPI because they don't belong to the new CAPI deployment, requiring manual node reprovisioning.

Requirements

1. Provider Compatibility

• Complete Azure/ARO support development and Microsoft collaboration
• Resolve Kubevirt incompatibility or document migration path
• Verify and document compatibility status for all providers

2. Node Readoption Support

• Implement node readoption for AWS/ROSA, Azure/ARO, PowerVS, and OpenStack (where technically feasible)
• Document node readoption procedures where available
• Document node reprovisioning procedures where node readoption is not available

3. E2E Testing

• Create comprehensive E2E test suites for cross-management cluster restore across all supported providers
• Integrate E2E tests into CI/CD pipelines
• Test success scenarios, failure recovery, data integrity validation, and edge cases

4. Documentation

• Step-by-step restore procedures per provider
• Prerequisites and requirements
• Known limitations and workarounds
• Troubleshooting guides
• Node readoption/reprovisioning procedures

Acceptance Criteria

• All supported HCP providers have verified compatibility with documented support status
• Kubevirt compatibility resolved or migration path documented
• Azure/ARO support completed and validated with Microsoft
• Node readoption implemented for all platforms where technically feasible
• Comprehensive E2E test coverage across all supported providers
• E2E tests integrated into CI/CD pipelines
• Complete documentation for all supported providers
• Technical enablement materials for field teams

References

• CNTRLPLANE-316: Auto backup/restore for HostedControlPlanes
• Manually importing a hosted cluster
• High availability for hosted control planes