bgp,test: Properly wait for FRR container to be ready

PodIP will be assigned even if the containers are not ready. Don't
waiting for the actual readiness state here causes unexpected error in
the later test case. For example, when FRR container is not brought up
with image pull failure, the later "Connectivity to endpoint via LB"
test fails because no one installs the route to the kernel FIB.

Signed-off-by: Yutaro Hayakawa <yutaro.hayakawa@isovalent.com>

test/k8s/bgp.go

@@ -67,14 +67,15 @@ var _ = SkipDescribeIf(
 				frr = applyFRRTemplate(kubectl, ni)
 				kubectl.ApplyDefault(frr).ExpectSuccess("Unable to apply rendered template %s", frr)
 
-				Eventually(func() string {
-					frrPod, err := kubectl.GetPodsIPs(helpers.KubeSystemNamespace, "app=frr")
-					if _, ok := frrPod["frr"]; err != nil || !ok {
-						return ""
-					}
-					routerIP = frrPod["frr"]
-					return routerIP
-				}, 30*time.Second, 1*time.Second).Should(Not(BeEmpty()), "BGP router is not ready")
+				err := kubectl.WaitForSinglePod("kube-system", "frr", 30*time.Second)
+				ExpectWithOffset(1, err).ShouldNot(HaveOccurred(), "Failed to wait for FRR Pod")
+
+				frrPod, err := kubectl.GetPodsIPs(helpers.KubeSystemNamespace, "app=frr")
+				ExpectWithOffset(1, err).ShouldNot(HaveOccurred(), "Cannot determine FRR Pod IPs")
+
+				var ok bool
+				routerIP, ok = frrPod["frr"]
+				Expect(ok).To(BeTrue())
 
 				bgpConfigMap = applyBGPCMTemplate(kubectl, routerIP)
 				kubectl.ApplyDefault(bgpConfigMap).ExpectSuccess("Unable to apply BGP ConfigMap %s", bgpConfigMap)
@@ -92,7 +93,6 @@ var _ = SkipDescribeIf(
 				demoDS = helpers.ManifestGet(kubectl.BasePath(), "demo_ds.yaml")
 				kubectl.ApplyDefault(demoDS).ExpectSuccess("Unable to apply %s", demoDS)
 
-				var err error
 				ciliumPodK8s1, err = kubectl.GetCiliumPodOnNode(helpers.K8s1)
 				ExpectWithOffset(1, err).ShouldNot(HaveOccurred(), "Cannot determine cilium pod name")
 				ciliumPodK8s2, err = kubectl.GetCiliumPodOnNode(helpers.K8s2)