Incorrect pod readiness calculation for resource metric

[tony-bondarenko]

There is an issue in pod readiness condition:

watermarkpodautoscaler/controllers/datadoghq/replica_calculator.go

Line 605 in c524600

ignorePod = condition.Status == corev1.ConditionFalse || pod.Status.StartTime.Add(delayOfInitialReadinessStatus).After(condition.LastTransitionTime.Time)

It checks that "pod start time" + "readiness delay" > "last ready condition transition" to include pod into calculations.
If ready state transition happened faster than readiness delay, pod will stay in ignored state forever.

More probably it should be:
condition.LastTransitionTime.Time.Add(delayOfInitialReadinessStatus).After(time.Now())

At the moment readinessDelaySeconds for resource metrics can't be used. Work around is to set it to 1.