Plumb metav1.Duration through healthchecks for more consistency

Author: JoelSpeed

controllers/machinehealthcheck_controller.go

@@ -210,7 +210,7 @@ func (r *MachineHealthCheckReconciler) reconcile(ctx context.Context, logger log
 	sort.Strings(m.Status.Targets)
 
 	// health check all targets and reconcile mhc status
-	healthy, unhealthy, nextCheckTimes := r.healthCheckTargets(targets, logger, m.Spec.NodeStartupTimeout.Duration)
+	healthy, unhealthy, nextCheckTimes := r.healthCheckTargets(targets, logger, *m.Spec.NodeStartupTimeout)
 	m.Status.CurrentHealthy = int32(len(healthy))
 
 	var unhealthyLimitKey, unhealthyLimitValue interface{}

controllers/machinehealthcheck_targets.go

@@ -86,7 +86,7 @@ func (t *healthCheckTarget) nodeName() string {
 // If the target doesn't currently need rememdiation, provide a duration after
 // which the target should next be checked.
 // The target should be requeued after this duration.
-func (t *healthCheckTarget) needsRemediation(logger logr.Logger, timeoutForMachineToHaveNode time.Duration) (bool, time.Duration) {
+func (t *healthCheckTarget) needsRemediation(logger logr.Logger, timeoutForMachineToHaveNode metav1.Duration) (bool, time.Duration) {
 	var nextCheckTimes []time.Duration
 	now := time.Now()
 
@@ -146,14 +146,14 @@ func (t *healthCheckTarget) needsRemediation(logger logr.Logger, timeoutForMachi
 		}
 		logger.V(3).Info("Using comparison time", "time", comparisonTime)
 
-		if comparisonTime.Add(timeoutForMachineToHaveNode).Before(now) {
+		if comparisonTime.Add(timeoutForMachineToHaveNode.Duration).Before(now) {
 			conditions.MarkFalse(t.Machine, clusterv1.MachineHealthCheckSuccededCondition, clusterv1.NodeStartupTimeoutReason, clusterv1.ConditionSeverityWarning, "Node failed to report startup in %s", timeoutForMachineToHaveNode.String())
 			logger.V(3).Info("Target is unhealthy: machine has no node", "duration", timeoutForMachineToHaveNode.String())
 			return true, time.Duration(0)
 		}
 
 		durationUnhealthy := now.Sub(comparisonTime)
-		nextCheck := timeoutForMachineToHaveNode - durationUnhealthy + time.Second
+		nextCheck := timeoutForMachineToHaveNode.Duration - durationUnhealthy + time.Second
 
 		return false, nextCheck
 	}
@@ -272,7 +272,7 @@ func (r *MachineHealthCheckReconciler) getNodeFromMachine(ctx context.Context, c
 
 // healthCheckTargets health checks a slice of targets
 // and gives a data to measure the average health.
-func (r *MachineHealthCheckReconciler) healthCheckTargets(targets []healthCheckTarget, logger logr.Logger, timeoutForMachineToHaveNode time.Duration) ([]healthCheckTarget, []healthCheckTarget, []time.Duration) {
+func (r *MachineHealthCheckReconciler) healthCheckTargets(targets []healthCheckTarget, logger logr.Logger, timeoutForMachineToHaveNode metav1.Duration) ([]healthCheckTarget, []healthCheckTarget, []time.Duration) {
 	var nextCheckTimes []time.Duration
 	var unhealthy []healthCheckTarget
 	var healthy []healthCheckTarget

controllers/machinehealthcheck_targets_test.go

@@ -389,9 +389,9 @@ func TestHealthCheckTargets(t *testing.T) {
 			}
 
 			// Allow individual test cases to override the timeoutForMachineToHaveNode.
-			timeout := timeoutForMachineToHaveNode
+			timeout := metav1.Duration{Duration: timeoutForMachineToHaveNode}
 			if tc.timeoutForMachineToHaveNode != nil {
-				timeout = *tc.timeoutForMachineToHaveNode
+				timeout.Duration = *tc.timeoutForMachineToHaveNode
 			}
 
 			healthy, unhealthy, nextCheckTimes := reconciler.healthCheckTargets(tc.targets, ctrl.LoggerFrom(ctx), timeout)