After seed generation, MCP was stuck in degraded state due to a missing rendered MC. Similarly, the seed image generated failed in upgrade attempts, due to same error.

    main

    1.
    2.
    3.
    

    The master mcp had different rendered-mc references in .spec vs .status, as the cleanup of stale rendered MCs occurred while the pull-secret sanitization was still being processed by MCD. As a result, the seed image and post-image seed SNO restore get stuck with MCD trying to find the deleted MC referenced in the .status.

The pull-secret sanitization function is waiting for MCD to update the pull-secret on disk, but does not wait for it to finish updating the MCP. Including the MCP health check before moving on ensures the previous rendered-mc is no longer referenced when stale rendered-mc cleanup occurs