From bfef0279a72ed104062dfb6b4f3986706457e201 Mon Sep 17 00:00:00 2001
From: Niels de Vos <ndevos@redhat.com>
Date: Thu, 28 Apr 2022 10:35:34 +0200
Subject: [PATCH] e2e: retry getting Pods before executing commands

On occasion the Pods have not been (re)started before they get listed.
This can result in an empty list. It can occur during RBD testing where
Pods are restarted before `uname` is executed. In case the Pods are not
available yet, the test will fail with the "podlist is empty" error.

By adding a retry when the list of Pods is empty, the tests should
become a little more stable.

Signed-off-by: Niels de Vos <ndevos@redhat.com>
---
 e2e/pod.go | 30 ++++++++++++++++++++++++------
 1 file changed, 24 insertions(+), 6 deletions(-)

diff --git a/e2e/pod.go b/e2e/pod.go
index 61fb03db8..4083589a1 100644
--- a/e2e/pod.go
+++ b/e2e/pod.go
@@ -91,13 +91,31 @@ func waitForDaemonSets(name, ns string, c kubernetes.Interface, t int) error {
 }
 
 func findPodAndContainerName(f *framework.Framework, ns, cn string, opt *metav1.ListOptions) (string, string, error) {
-	podList, err := f.PodClientNS(ns).List(context.TODO(), *opt)
-	if err != nil {
-		return "", "", err
-	}
+	timeout := time.Duration(deployTimeout) * time.Minute
 
-	if len(podList.Items) == 0 {
-		return "", "", errors.New("podlist is empty")
+	var (
+		podList *v1.PodList
+		listErr error
+	)
+	err := wait.PollImmediate(poll, timeout, func() (bool, error) {
+		podList, listErr = f.PodClientNS(ns).List(context.TODO(), *opt)
+		if listErr != nil {
+			if isRetryableAPIError(listErr) {
+				return false, nil
+			}
+
+			return false, fmt.Errorf("failed to list Pods: %w", listErr)
+		}
+
+		if len(podList.Items) == 0 {
+			// retry in case the pods have not been (re)started yet
+			return false, nil
+		}
+
+		return true, nil
+	})
+	if err != nil {
+		return "", "", fmt.Errorf("failed to find pod for %v: %w", opt, err)
 	}
 
 	if cn != "" {