loadbalancers are deleted in master machine when changing authoritativeAPI from MachineAPI to ClusterAPI, then it will cause sync error when changing authoritativeAPI back to MachineAPI

    4.21.0-0.nightly-2025-10-19-181151

    always

    1.Install an AWS techpreview cluster, we can see there are loadbalancers in the AWSCluster

liuhuali@Lius-MacBook-Pro huali-test % oc get awscluster huliu-aws1020b-rsrcn -n openshift-cluster-api -oyaml
apiVersion: infrastructure.cluster.x-k8s.io/v1beta2
kind: AWSCluster
metadata:
  annotations:
    cluster.x-k8s.io/managed-by: cluster-capi-operator-infracluster-controller
  creationTimestamp: "2025-10-20T03:41:22Z"
  generation: 1
  labels:
    cluster.x-k8s.io/cluster-name: huliu-aws1020b-rsrcn
  name: huliu-aws1020b-rsrcn
  namespace: openshift-cluster-api
  ownerReferences:
  - apiVersion: cluster.x-k8s.io/v1beta1
    blockOwnerDeletion: true
    controller: true
    kind: Cluster
    name: huliu-aws1020b-rsrcn
    uid: fbb38b3c-3cad-402b-9163-149f6092af17
  resourceVersion: "20146"
  uid: 4030074c-39d4-4e20-a2f9-615e60449ba4
spec:
  bastion:
    enabled: false
  controlPlaneEndpoint:
    host: api-int.huliu-aws1020b.qe.devcluster.openshift.com
    port: 6443
  controlPlaneLoadBalancer:
    crossZoneLoadBalancing: false
    loadBalancerType: nlb
    name: huliu-aws1020b-rsrcn-int
    scheme: internet-facing
  identityRef:
    kind: AWSClusterControllerIdentity
    name: default
  network:
    vpc:
      availabilityZoneSelection: Ordered
      availabilityZoneUsageLimit: 3
      subnetSchema: PreferPrivate
  region: us-east-2
  secondaryControlPlaneLoadBalancer:
    crossZoneLoadBalancing: false
    loadBalancerType: nlb
    name: huliu-aws1020b-rsrcn-ext
    scheme: internet-facing
status:
  ready: true

    2.Delete controlplanemachineset, then we can see master machines are sync to CAPI 

liuhuali@Lius-MacBook-Pro huali-test % oc get machine huliu-aws1020b-rsrcn-master-0 -n openshift-machine-api -oyaml
apiVersion: machine.openshift.io/v1beta1
kind: Machine
metadata:
  annotations:
    machine.openshift.io/instance-state: running
  creationTimestamp: "2025-10-20T03:38:28Z"
  finalizers:
  - sync.machine.openshift.io/finalizer
  - machine.machine.openshift.io
  generation: 3
  labels:
    machine.openshift.io/cluster-api-cluster: huliu-aws1020b-rsrcn
    machine.openshift.io/cluster-api-machine-role: master
    machine.openshift.io/cluster-api-machine-type: master
    machine.openshift.io/instance-type: m6i.xlarge
    machine.openshift.io/region: us-east-2
    machine.openshift.io/zone: us-east-2c
  name: huliu-aws1020b-rsrcn-master-0
  namespace: openshift-machine-api
  resourceVersion: "48288"
  uid: 10fc9a3d-e439-4710-9d9f-6e136969db73
spec:
  authoritativeAPI: MachineAPI
  lifecycleHooks:
    preDrain:
    - name: EtcdQuorumOperator
      owner: clusteroperator/etcd
  metadata: {}
  providerID: aws:///us-east-2c/i-0fa92d671ac915000
  providerSpec:
    value:
      ami:
        id: ami-034f9c3fdfc80a314
      apiVersion: machine.openshift.io/v1beta1
      blockDevices:
      - ebs:
          encrypted: true
          iops: 0
          kmsKey:
            arn: ""
          volumeSize: 120
          volumeType: gp3
      capacityReservationId: ""
      credentialsSecret:
        name: aws-cloud-credentials
      deviceIndex: 0
      iamInstanceProfile:
        id: huliu-aws1020b-rsrcn-master-profile
      instanceType: m6i.xlarge
      kind: AWSMachineProviderConfig
      loadBalancers:
      - name: huliu-aws1020b-rsrcn-int
        type: network
      - name: huliu-aws1020b-rsrcn-ext
        type: network
      metadata:
        creationTimestamp: null
      metadataServiceOptions: {}
      placement:
        availabilityZone: us-east-2b
        region: us-east-2
      securityGroups:
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-node
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-lb
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-controlplane
      subnet:
        filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-subnet-private-us-east-2b
      tags:
      - name: kubernetes.io/cluster/huliu-aws1020b-rsrcn
        value: owned
      userDataSecret:
        name: master-user-data
status:
  addresses:
  - address: 10.0.92.153
    type: InternalIP
  - address: ip-10-0-92-153.us-east-2.compute.internal
    type: InternalDNS
  - address: ip-10-0-92-153.us-east-2.compute.internal
    type: Hostname
  authoritativeAPI: MachineAPI
  conditions:
  - lastTransitionTime: "2025-10-20T03:42:58Z"
    message: 'Drain operation currently blocked by: [{Name:EtcdQuorumOperator Owner:clusteroperator/etcd}]'
    reason: HookPresent
    severity: Warning
    status: "False"
    type: Drainable
  - lastTransitionTime: "2025-10-20T03:42:50Z"
    status: "True"
    type: InstanceExists
  - lastTransitionTime: "2025-10-20T03:42:48Z"
    message: The AuthoritativeAPI status is set to 'MachineAPI'
    reason: AuthoritativeAPIMachineAPI
    severity: Info
    status: "False"
    type: Paused
  - lastTransitionTime: "2025-10-20T04:49:19Z"
    message: Successfully synchronized MAPI Machine to CAPI
    reason: ResourceSynchronized
    severity: ""
    status: "True"
    type: Synchronized
  - lastTransitionTime: "2025-10-20T03:42:48Z"
    status: "True"
    type: Terminable
  lastUpdated: "2025-10-20T03:42:58Z"
  nodeRef:
    kind: Node
    name: ip-10-0-92-153.us-east-2.compute.internal
    uid: e181b1a6-aeb1-4fd1-aa78-86b1e480dd33
  phase: Running
  providerStatus:
    conditions:
    - lastTransitionTime: "2025-10-20T03:42:50Z"
      message: Machine successfully created
      reason: MachineCreationSucceeded
      status: "True"
      type: MachineCreation
    instanceId: i-0fa92d671ac915000
    instanceState: running
  synchronizedGeneration: 3

    3.Change one master machine's authoritativeAPI from MachineAPI to ClusterAPI, sync successfully, but loadbalancers are deleted in the MAPI machine

liuhuali@Lius-MacBook-Pro huali-test % oc edit machine huliu-aws1020b-rsrcn-master-0 -n openshift-machine-api      
machine.machine.openshift.io/huliu-aws1020b-rsrcn-master-0 edited

liuhuali@Lius-MacBook-Pro huali-test % oc get machine huliu-aws1020b-rsrcn-master-0 -n openshift-machine-api -oyaml
apiVersion: machine.openshift.io/v1beta1
kind: Machine
metadata:
  annotations:
    machine.openshift.io/instance-state: running
  creationTimestamp: "2025-10-20T03:38:28Z"
  finalizers:
  - sync.machine.openshift.io/finalizer
  - machine.machine.openshift.io
  generation: 5
  labels:
    machine.openshift.io/cluster-api-cluster: huliu-aws1020b-rsrcn
    machine.openshift.io/cluster-api-machine-role: master
    machine.openshift.io/cluster-api-machine-type: master
    machine.openshift.io/instance-type: m6i.xlarge
    machine.openshift.io/region: us-east-2
    machine.openshift.io/zone: us-east-2c
  name: huliu-aws1020b-rsrcn-master-0
  namespace: openshift-machine-api
  resourceVersion: "132718"
  uid: 10fc9a3d-e439-4710-9d9f-6e136969db73
spec:
  authoritativeAPI: ClusterAPI
  lifecycleHooks:
    preDrain:
    - name: EtcdQuorumOperator
      owner: clusteroperator/etcd
  metadata:
    annotations:
      machine.openshift.io/instance-state: running
    labels:
      machine.openshift.io/cluster-api-cluster: huliu-aws1020b-rsrcn
      machine.openshift.io/instance-type: m6i.xlarge
      machine.openshift.io/region: us-east-2
      machine.openshift.io/zone: us-east-2c
      node-role.kubernetes.io/master: ""
  providerID: aws:///us-east-2c/i-0fa92d671ac915000
  providerSpec:
    value:
      ami:
        id: ami-034f9c3fdfc80a314
      apiVersion: machine.openshift.io/v1beta1
      blockDevices:
      - ebs:
          encrypted: true
          kmsKey:
            id: ""
          volumeSize: 120
          volumeType: gp3
      capacityReservationId: ""
      credentialsSecret:
        name: aws-cloud-credentials
      deviceIndex: 0
      iamInstanceProfile:
        id: huliu-aws1020b-rsrcn-master-profile
      instanceType: m6i.xlarge
      kind: AWSMachineProviderConfig
      metadata:
        creationTimestamp: null
      metadataServiceOptions:
        authentication: Optional
      placement:
        availabilityZone: us-east-2b
        region: us-east-2
      securityGroups:
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-node
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-lb
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-controlplane
      subnet:
        filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-subnet-private-us-east-2b
      tags:
      - name: kubernetes.io/cluster/huliu-aws1020b-rsrcn
        value: owned
      userDataSecret:
        name: master-user-data
status:
  addresses:
  - address: ip-10-0-92-153.us-east-2.compute.internal
    type: InternalDNS
  - address: 10.0.92.153
    type: InternalIP
  authoritativeAPI: ClusterAPI
  conditions:
  - lastTransitionTime: "2025-10-20T03:42:58Z"
    message: 'Drain operation currently blocked by: [{Name:EtcdQuorumOperator Owner:clusteroperator/etcd}]'
    reason: HookPresent
    severity: Warning
    status: "False"
    type: Drainable
  - lastTransitionTime: "2025-10-20T03:42:50Z"
    status: "True"
    type: InstanceExists
  - lastTransitionTime: "2025-10-20T08:05:52Z"
    message: The AuthoritativeAPI status is set to 'ClusterAPI'
    reason: AuthoritativeAPINotMachineAPI
    status: "True"
    type: Paused
  - lastTransitionTime: "2025-10-20T08:05:54Z"
    message: Successfully synchronized CAPI Machine to MAPI
    reason: ResourceSynchronized
    severity: ""
    status: "True"
    type: Synchronized
  - lastTransitionTime: "2025-10-20T03:42:48Z"
    status: "True"
    type: Terminable
  lastUpdated: "2025-10-20T03:42:58Z"
  nodeRef:
    kind: Node
    name: ip-10-0-92-153.us-east-2.compute.internal
    uid: e181b1a6-aeb1-4fd1-aa78-86b1e480dd33
  phase: Running
  providerStatus:
    conditions:
    - lastTransitionTime: "2025-10-20T03:42:50Z"
      message: Machine successfully created
      reason: MachineCreationSucceeded
      status: "True"
      type: MachineCreation
    instanceId: i-0fa92d671ac915000
    instanceState: running
  synchronizedGeneration: 1    

4.Change the machine's authoritativeAPI back to MachineAPI, it reports sync error

liuhuali@Lius-MacBook-Pro huali-test % oc edit machine huliu-aws1020b-rsrcn-master-0 -n openshift-machine-api      
machine.machine.openshift.io/huliu-aws1020b-rsrcn-master-0 edited
liuhuali@Lius-MacBook-Pro huali-test % oc get machine huliu-aws1020b-rsrcn-master-0 -n openshift-machine-api -oyaml
apiVersion: machine.openshift.io/v1beta1
kind: Machine
metadata:
  annotations:
    machine.openshift.io/instance-state: running
  creationTimestamp: "2025-10-20T03:38:28Z"
  finalizers:
  - sync.machine.openshift.io/finalizer
  - machine.machine.openshift.io
  generation: 6
  labels:
    machine.openshift.io/cluster-api-cluster: huliu-aws1020b-rsrcn
    machine.openshift.io/cluster-api-machine-role: master
    machine.openshift.io/cluster-api-machine-type: master
    machine.openshift.io/instance-type: m6i.xlarge
    machine.openshift.io/region: us-east-2
    machine.openshift.io/zone: us-east-2c
  name: huliu-aws1020b-rsrcn-master-0
  namespace: openshift-machine-api
  resourceVersion: "133340"
  uid: 10fc9a3d-e439-4710-9d9f-6e136969db73
spec:
  authoritativeAPI: MachineAPI
  lifecycleHooks:
    preDrain:
    - name: EtcdQuorumOperator
      owner: clusteroperator/etcd
  metadata:
    annotations:
      machine.openshift.io/instance-state: running
    labels:
      machine.openshift.io/cluster-api-cluster: huliu-aws1020b-rsrcn
      machine.openshift.io/instance-type: m6i.xlarge
      machine.openshift.io/region: us-east-2
      machine.openshift.io/zone: us-east-2c
      node-role.kubernetes.io/master: ""
  providerID: aws:///us-east-2c/i-0fa92d671ac915000
  providerSpec:
    value:
      ami:
        id: ami-034f9c3fdfc80a314
      apiVersion: machine.openshift.io/v1beta1
      blockDevices:
      - ebs:
          encrypted: true
          kmsKey:
            id: ""
          volumeSize: 120
          volumeType: gp3
      capacityReservationId: ""
      credentialsSecret:
        name: aws-cloud-credentials
      deviceIndex: 0
      iamInstanceProfile:
        id: huliu-aws1020b-rsrcn-master-profile
      instanceType: m6i.xlarge
      kind: AWSMachineProviderConfig
      metadata:
        creationTimestamp: null
      metadataServiceOptions:
        authentication: Optional
      placement:
        availabilityZone: us-east-2b
        region: us-east-2
      securityGroups:
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-node
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-lb
      - filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-controlplane
      subnet:
        filters:
        - name: tag:Name
          values:
          - huliu-aws1020b-rsrcn-subnet-private-us-east-2b
      tags:
      - name: kubernetes.io/cluster/huliu-aws1020b-rsrcn
        value: owned
      userDataSecret:
        name: master-user-data
status:
  addresses:
  - address: 10.0.92.153
    type: InternalIP
  - address: ip-10-0-92-153.us-east-2.compute.internal
    type: InternalDNS
  - address: ip-10-0-92-153.us-east-2.compute.internal
    type: Hostname
  authoritativeAPI: MachineAPI
  conditions:
  - lastTransitionTime: "2025-10-20T03:42:58Z"
    message: 'Drain operation currently blocked by: [{Name:EtcdQuorumOperator Owner:clusteroperator/etcd}]'
    reason: HookPresent
    severity: Warning
    status: "False"
    type: Drainable
  - lastTransitionTime: "2025-10-20T03:42:50Z"
    status: "True"
    type: InstanceExists
  - lastTransitionTime: "2025-10-20T08:06:46Z"
    message: The AuthoritativeAPI status is set to 'MachineAPI'
    reason: AuthoritativeAPIMachineAPI
    severity: Info
    status: "False"
    type: Paused
  - lastTransitionTime: "2025-10-20T08:06:46Z"
    message: 'failed to convert Machine API machine to Cluster API machine: [spec.providerSpec.value.loadBalancers:
      Invalid value: []v1beta1.LoadBalancerReference(nil): must include load balancer
      named "huliu-aws1020b-rsrcn-int", spec.providerSpec.value.loadBalancers: Invalid
      value: []v1beta1.LoadBalancerReference(nil): must include load balancer named
      "huliu-aws1020b-rsrcn-ext"]'
    reason: FailedToConvertMAPIMachineToCAPI
    severity: Error
    status: "False"
    type: Synchronized
  - lastTransitionTime: "2025-10-20T03:42:48Z"
    status: "True"
    type: Terminable
  lastUpdated: "2025-10-20T08:06:46Z"
  nodeRef:
    kind: Node
    name: ip-10-0-92-153.us-east-2.compute.internal
    uid: e181b1a6-aeb1-4fd1-aa78-86b1e480dd33
  phase: Running
  providerStatus:
    conditions:
    - lastTransitionTime: "2025-10-20T03:42:50Z"
      message: Machine successfully created
      reason: MachineCreationSucceeded
      status: "True"
      type: MachineCreation
    instanceId: i-0fa92d671ac915000
    instanceState: running
  synchronizedGeneration: 0 

    loadbalancers are deleted in master machine when changing authoritativeAPI from MachineAPI to ClusterAPI, then it will cause sync error when changing authoritativeAPI back to MachineAPI

    loadbalancers should not be deleted in master machine when changing authoritativeAPI from MachineAPI to ClusterAPI, it should sync successfully when changing authoritativeAPI back to MachineAPI

    new feature testing for https://issues.redhat.com/browse/OCPCLOUD-2709