Merge pull request #25192 from ingvagabund/e2e-framework-util-start-pods-dont-wait-for-zero-pods

Automatic merge from submit-queue

e2e/framework/util.StartPods: don't wait for pods that are not created

When running ``[k8s.io] SchedulerPredicates [Serial] validates resource limits of pods that are allowed to run [Conformance]`` pods can be created in a way in which additional pods have to be create to fully saturate node's capacity CPU in a cluster. Additional pods are created by calling ``framework.StartPods``. The function creates pods with a given label and waits for them (if ``waitForRunning`` is ``true``). This is fine as long as the number of pods to created is non-zero. If there are zero pods to be created and ``waitForRunning`` is ``true``, the function waits forever as there is not going to be any pods with requested label. Thus, resulting in ``Error waiting for 0 pods to be running - probably a timeout``. Causing the e2e test to fail even if it should not.

Adding condition to return from the function immediately if there is not pod to create.

test/e2e/framework/util.go

@@ -2095,6 +2095,11 @@ func (config *RCConfig) start() error {
 // Simplified version of RunRC, that does not create RC, but creates plain Pods.
 // optionally waits for pods to start running (if waitForRunning == true)
 func StartPods(c *client.Client, replicas int, namespace string, podNamePrefix string, pod api.Pod, waitForRunning bool) {
+	// no pod to start
+	if replicas < 1 {
+		Logf("No pod to start, skipping...")
+		return
+	}
 	startPodsID := string(util.NewUUID()) // So that we can label and find them
 	for i := 0; i < replicas; i++ {
 		podName := fmt.Sprintf("%v-%v", podNamePrefix, i)