kubernetes-sigs / cluster-api-provider-kubevirt / 24126594272

38%

main: 38%

Pull #382

github

fix: clear FailureReason/FailureMessage when VM recovers from terminal state

When a KubeVirt VM is stopped (runStrategy changed to Halted), the VMI
reaches a finalized state and IsTerminal() correctly sets FailureReason
on the KubevirtMachine. However, when the VM is started back
(runStrategy changed to Always), IsTerminal() returns false but the
FailureReason/FailureMessage fields are never cleared.

This causes the CAPI Machine phase to stay stuck at "Failed" even though
all Machine conditions (Ready, InfrastructureReady, NodeHealthy) are
True. Downstream consumers that rely on the Machine phase (such as
HyperShift's endpointslice reconciler) then permanently mark endpoints
as not ready, breaking ingress into hosted clusters.

The fix adds an else clause to nil out FailureReason and FailureMessage
when IsTerminal() returns false, allowing VMs to properly recover.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Enrique Llorente <ellorent@redhat.com>

3 of 3 new or added lines in 1 file covered. (100.0%)

1049 of 2765 relevant lines covered (37.94%)

2.14 hits per line