🐛 Fix Available/Ready checks on E2E test (#12549)

* Fix runtime-sdk E2E test

* More fix

* Fix log message

Author: fabriziopandini

test/e2e/cluster_upgrade_runtimesdk.go

@@ -331,6 +331,20 @@ func ClusterUpgradeWithRuntimeSDKSpec(ctx context.Context, inputGetter func() Cl
 			input.PostUpgrade(input.BootstrapClusterProxy, namespace.Name, clusterResources.Cluster.Name)
 		}
 
+		Byf("Verify Cluster Available condition is true")
+		framework.VerifyClusterAvailable(ctx, framework.VerifyClusterAvailableInput{
+			Getter:    input.BootstrapClusterProxy.GetClient(),
+			Name:      clusterResources.Cluster.Name,
+			Namespace: clusterResources.Cluster.Namespace,
+		})
+
+		Byf("Verify Machines Ready condition is true")
+		framework.VerifyMachinesReady(ctx, framework.VerifyMachinesReadyInput{
+			Lister:    input.BootstrapClusterProxy.GetClient(),
+			Name:      clusterResources.Cluster.Name,
+			Namespace: clusterResources.Cluster.Namespace,
+		})
+
 		By("Dumping resources and deleting the workload cluster; deletion waits for BeforeClusterDeleteHook to gate the operation")
 		dumpAndDeleteCluster(ctx, input.BootstrapClusterProxy, input.ClusterctlConfigPath, namespace.Name, clusterName, input.ArtifactFolder)
 
@@ -347,20 +361,6 @@ func ClusterUpgradeWithRuntimeSDKSpec(ctx context.Context, inputGetter func() Cl
 			"AfterClusterUpgrade":          "Success",
 		})).To(Succeed(), "Lifecycle hook calls were not as expected")
 
-		Byf("Verify Cluster Available condition is true")
-		framework.VerifyClusterAvailable(ctx, framework.VerifyClusterAvailableInput{
-			Getter:    input.BootstrapClusterProxy.GetClient(),
-			Name:      clusterResources.Cluster.Name,
-			Namespace: clusterResources.Cluster.Namespace,
-		})
-
-		Byf("Verify Machines Ready condition is true")
-		framework.VerifyMachinesReady(ctx, framework.VerifyMachinesReadyInput{
-			Lister:    input.BootstrapClusterProxy.GetClient(),
-			Name:      clusterResources.Cluster.Name,
-			Namespace: clusterResources.Cluster.Namespace,
-		})
-
 		By("PASSED!")
 	})
 

test/e2e/clusterctl_upgrade.go

@@ -723,25 +723,27 @@ func ClusterctlUpgradeSpec(ctx context.Context, inputGetter func() ClusterctlUpg
 				upgrade.PostUpgrade(managementClusterProxy, workloadCluster.Namespace, workloadCluster.Name)
 			}
 
-			Byf("[%d] Verify Cluster Available condition is true", i)
-			framework.VerifyClusterAvailable(ctx, framework.VerifyClusterAvailableInput{
-				Getter:    managementClusterProxy.GetClient(),
-				Name:      workloadCluster.Name,
-				Namespace: workloadCluster.Namespace,
-			})
-
-			Byf("[%d] Verify Machines Ready condition is true", i)
-			framework.VerifyMachinesReady(ctx, framework.VerifyMachinesReadyInput{
-				Lister:    managementClusterProxy.GetClient(),
-				Name:      workloadCluster.Name,
-				Namespace: workloadCluster.Namespace,
-			})
-
 			// If this is the last step of the upgrade sequence check hat the resourceVersions are stable, i.e. it verifies there are no
 			// continuous reconciles when everything should be stable.
 			if i == len(input.Upgrades)-1 {
-				By("Checking that resourceVersions are stable")
+				Byf("[%d] Checking that resourceVersions are stable", i)
 				framework.ValidateResourceVersionStable(ctx, managementClusterProxy, workloadCluster.Namespace, clusterctlcluster.FilterClusterObjectsWithNameFilter(workloadCluster.Name))
+
+				// NOTE: Checks on conditions works on v1beta2 only, so running this checks only in the last step which is
+				// always current version.
+				Byf("[%d] Verify Cluster Available condition is true", i)
+				framework.VerifyClusterAvailable(ctx, framework.VerifyClusterAvailableInput{
+					Getter:    managementClusterProxy.GetClient(),
+					Name:      workloadCluster.Name,
+					Namespace: workloadCluster.Namespace,
+				})
+
+				Byf("[%d] Verify Machines Ready condition is true", i)
+				framework.VerifyMachinesReady(ctx, framework.VerifyMachinesReadyInput{
+					Lister:    managementClusterProxy.GetClient(),
+					Name:      workloadCluster.Name,
+					Namespace: workloadCluster.Namespace,
+				})
 			}
 
 			// Note: It is a known issue on Kubernetes < v1.29 that SSA sometimes fail:

test/e2e/self_hosted.go

@@ -334,6 +334,20 @@ func SelfHostedSpec(ctx context.Context, inputGetter func() SelfHostedSpecInput)
 		}, "3m", "30s").ShouldNot(HaveOccurred(), "MachineList should be available after move to self-hosted cluster")
 		Expect(validateMachineRollout(preMoveMachineList, postMoveMachineList)).To(BeTrue(), "Machines should not roll out after move to self-hosted cluster")
 
+		Byf("Verify Cluster Available condition is true")
+		framework.VerifyClusterAvailable(ctx, framework.VerifyClusterAvailableInput{
+			Getter:    selfHostedClusterProxy.GetClient(),
+			Name:      clusterResources.Cluster.Name,
+			Namespace: clusterResources.Cluster.Namespace,
+		})
+
+		Byf("Verify Machines Ready condition is true")
+		framework.VerifyMachinesReady(ctx, framework.VerifyMachinesReadyInput{
+			Lister:    selfHostedClusterProxy.GetClient(),
+			Name:      clusterResources.Cluster.Name,
+			Namespace: clusterResources.Cluster.Namespace,
+		})
+
 		if input.SkipUpgrade {
 			// Only do upgrade step if defined by test input.
 			return
@@ -431,14 +445,14 @@ func SelfHostedSpec(ctx context.Context, inputGetter func() SelfHostedSpecInput)
 
 		Byf("Verify Cluster Available condition is true")
 		framework.VerifyClusterAvailable(ctx, framework.VerifyClusterAvailableInput{
-			Getter:    input.BootstrapClusterProxy.GetClient(),
+			Getter:    selfHostedClusterProxy.GetClient(),
 			Name:      clusterResources.Cluster.Name,
 			Namespace: clusterResources.Cluster.Namespace,
 		})
 
 		Byf("Verify Machines Ready condition is true")
 		framework.VerifyMachinesReady(ctx, framework.VerifyMachinesReadyInput{
-			Lister:    input.BootstrapClusterProxy.GetClient(),
+			Lister:    selfHostedClusterProxy.GetClient(),
 			Name:      clusterResources.Cluster.Name,
 			Namespace: clusterResources.Cluster.Namespace,
 		})

test/framework/cluster_helpers.go

@@ -452,7 +452,7 @@ func VerifyClusterAvailable(ctx context.Context, input VerifyClusterAvailableInp
 		g.Expect(input.Getter.Get(ctx, key, cluster)).To(Succeed())
 		for _, condition := range cluster.Status.Conditions {
 			if condition.Type == clusterv1.AvailableCondition {
-				g.Expect(condition.Status).To(Equal(metav1.ConditionTrue), "The Available condition on the Cluster should be set to true")
+				g.Expect(condition.Status).To(Equal(metav1.ConditionTrue), "The Available condition on the Cluster should be set to true; message: %s", condition.Message)
 				g.Expect(condition.Message).To(BeEmpty(), "The Available condition on the Cluster should have an empty message")
 				return
 			}
@@ -484,7 +484,7 @@ func VerifyMachinesReady(ctx context.Context, input VerifyMachinesReadyInput) {
 			for _, condition := range machine.Status.Conditions {
 				if condition.Type == clusterv1.ReadyCondition {
 					readyConditionFound = true
-					g.Expect(condition.Status).To(Equal(metav1.ConditionTrue), "The Ready condition on Machine %q should be set to true", machine.Name)
+					g.Expect(condition.Status).To(Equal(metav1.ConditionTrue), "The Ready condition on Machine %q should be set to true; message: %s", machine.Name, condition.Message)
 					g.Expect(condition.Message).To(BeEmpty(), "The Ready condition on Machine %q should have an empty message", machine.Name)
 					break
 				}