The HyperShift operator (NodePool controller) panics when attempting to reconcile a NodePool for a Hosted Control Plane (HCP) on KubeVirt when spec.management.upgradeType: InPlace is specified. The NodePool remains stuck with no nodes created, and the operator logs show a runtime error: invalid memory address or nil pointer dereference.

HCP 4.20    
OCP 4.20
MCE 2.10
ACM 2.15

    100%

1. Deploy a Hosted Control Plane on KubeVirt.
2. Apply a NodePool manifest using upgradeType: InPlace and strategy: RollingUpdate.
3. Observe the NodePool status (it stays empty).
4. Check the logs of the hypershift-operator pod in the hypershift namespace.

    The NodePool controller crashes/panics during reconciliation. No VirtualMachines or Machine objects are created.

{"level":"error","ts":"2026-02-11T12:20:49Z","msg":"Reconciler error","controller":"nodepool","error":"panic: runtime error: invalid memory address or nil pointer dereference [recovered]","stacktrace":"sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller[...]).reconcileHandler..."}

The operator should either:

Successfully reconcile and create the nodes using the In-Place upgrade logic.

Gracefully fail with a validation error if InPlace is not yet supported for the KubeVirt platform in this version.

    nodepool YAML:

apiVersion: hypershift.openshift.io/v1beta1
kind: NodePool
metadata:
  name: hosted-kubevirt-custom
  namespace: clusters
spec:
  clusterName: hosted-kubevirt
  management:
    autoRepair: false
    upgradeType: InPlace 
    replace:
      strategy: RollingUpdate
  platform:
    type: KubeVirt
    kubevirt:
      compute:
        cores: 4
        memory: 8Gi