test(e2e): optimize test execution time by reducing unnecessary waits (#6369)

* test(e2e): optimize test execution time by reducing unnecessary waits

- Reduce global poll interval from 2s to 1s for faster resource readiness detection
- Fix WaitUntil ignoring its interval parameter (was hardcoded to 2s)
- Reduce disaster test sleep from 60s to 5s (process STOP/CONT simulation)
- Reduce nc timeout from 5s to 2s in NAT policy external access checks
- Add PeriodSeconds:1 and FailureThreshold:1 to VPC pod probe tests
- Skip redundant StatefulSet IPPool iterations (replicas 1,2) and test only replicas=3
- Parallelize NAT policy ipset/iptables checks across nodes

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* test(e2e): unify poll interval to 1s and reduce pod termination grace period

Replace all 150 hardcoded 2*time.Second poll intervals with the framework
poll constant (1s) or time.Second for files outside the framework package.
Also reduce TerminationGracePeriodSeconds from 3 to 1 for pods and add it
to StatefulSets to speed up teardown.

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>

* test(e2e): fix deployment termination grace period and remaining poll intervals

- Add TerminationGracePeriodSeconds=1 to MakeDeployment, consistent with
  MakePod and MakeStatefulSet. Without this, deployment pods use the
  default 30s grace period, slowing down all deployment-based tests.
- Reduce checkIPSetOnNode poll interval from 3s to 1s to match the
  unified poll standard.
- Reduce waitForInterfaceState poll interval from 5s to 1s to match the
  unified poll standard.

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>

* test(e2e): parallelize AfterEach cleanup to reduce test execution time

Restructure AfterEach blocks in 8 test files to delete resources in
parallel within the same dependency level, then wait for all to
disappear before proceeding to the next level. This reduces cleanup
time from N × single-resource-time to max(single-resource-time) per
level.

Add PodClient.DeleteGracefully() method that initiates pod deletion
with a 1-second grace period without blocking, enabling parallel
pod cleanup alongside other resource types.

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>

* test(e2e): fix node update conflict in nodeSelector underlay test

Use retry.RetryOnConflict to wrap node label update operations to handle
409 Conflict errors caused by stale resourceVersion when concurrent
controllers modify the node between Get and Update.

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>

* test(e2e): fix deployment ippool test race condition with terminating pods

GetPodsForDeployment() does not filter out pods with DeletionTimestamp,
so when pods are deleted and recreated within the same ReplicaSet, both
old (terminating) and new (running) pods are returned, causing
"expect 3 but got 6" failures.

Fix by using DeleteGracefully (1s grace period) and explicitly waiting
for old pods to be fully removed before checking new pod count.

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>

* test(e2e): fix Restart() 409 conflict by retrying with fresh resourceVersion

Wrap Deployment and DaemonSet Restart() with retry.RetryOnConflict to
handle 409 Conflict errors caused by concurrent controller status updates
between Get and Update. Each retry fetches the latest object to ensure
a current resourceVersion.

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Signed-off-by: Mengxin Liu <liumengxinfly@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Author: oilbeater

test/e2e/anp-domain/e2e_test.go

@@ -111,7 +111,7 @@ var _ = framework.SerialDescribe("[group:admin-network-policy]", func() {
 	}
 
 	testNetworkConnectivity := func(target string, shouldSucceed bool, description string) {
-		testNetworkConnectivityWithRetry(target, shouldSucceed, description, 20, 2*time.Second)
+		testNetworkConnectivityWithRetry(target, shouldSucceed, description, 20, time.Second)
 	}
 
 	framework.ConformanceIt("should create ANP with domainName deny rule and verify connectivity behavior", func() {

test/e2e/cnp-domain/e2e_test.go

@@ -128,7 +128,7 @@ var _ = framework.SerialDescribe("[group:cluster-network-policy]", func() {
 		dnsClient := f.DNSNameResolverClient()
 		labelSelector := fmt.Sprintf("anp=%s", name)
 
-		framework.WaitUntil(2*time.Second, 30*time.Second, func(_ context.Context) (bool, error) {
+		framework.WaitUntil(time.Second, 30*time.Second, func(_ context.Context) (bool, error) {
 			resolverList := dnsClient.ListByLabel(labelSelector)
 			if len(resolverList.Items) < expectedCount {
 				framework.Logf("Found %d/%d DNSNameResolver(s) for CNP %s", len(resolverList.Items), expectedCount, name)

test/e2e/connectivity/e2e_test.go

@@ -255,7 +255,7 @@ var _ = framework.OrderedDescribe("[group:disaster]", func() {
 
 		ginkgo.By("Waiting for newly created ovs-ovn pod running on node " + suiteCtx.Node)
 		pod = nil
-		framework.WaitUntil(2*time.Second, 2*time.Minute, func(_ context.Context) (bool, error) {
+		framework.WaitUntil(time.Second, 2*time.Minute, func(_ context.Context) (bool, error) {
 			ds = dsClient.Get("ovs-ovn")
 			pods, err = dsClient.GetPods(ds)
 			if err != nil {
@@ -329,8 +329,8 @@ var _ = framework.OrderedDescribe("[group:disaster]", func() {
 			framework.ExpectNoError(err, "failed to send STOP signal to ovn sb process: %v, %s", err, string(stderr))
 		}
 
-		ginkgo.By("Waiting 60s")
-		time.Sleep(60 * time.Second)
+		ginkgo.By("Waiting 5s for process outage simulation")
+		time.Sleep(5 * time.Second)
 
 		for _, pod := range pods.Items {
 			ginkgo.By("Getting ovn sb pid of pod " + pod.Name + " running on " + pod.Spec.NodeName)
@@ -380,8 +380,8 @@ var _ = framework.OrderedDescribe("[group:disaster]", func() {
 		_, stderr, err = framework.KubectlExec(pod.Namespace, pod.Name, cmd...)
 		framework.ExpectNoError(err, "failed to send STOP signal to ovn-controller process: %v, %s", err, string(stderr))
 
-		ginkgo.By("Waiting 60s")
-		time.Sleep(60 * time.Second)
+		ginkgo.By("Waiting 5s for process outage simulation")
+		time.Sleep(5 * time.Second)
 
 		ginkgo.By("Continuing the stopped ovn-controller process by sending a CONT signal")
 		cmd = []string{"sh", "-c", fmt.Sprintf(`"kill -CONT %s"`, pid)}
@@ -422,8 +422,8 @@ var _ = framework.OrderedDescribe("[group:disaster]", func() {
 		_, stderr, err = framework.KubectlExec(pod.Namespace, pod.Name, cmd...)
 		framework.ExpectNoError(err, "failed to send STOP signal to ovs-vswitchd process: %v, %s", err, string(stderr))
 
-		ginkgo.By("Waiting 60s")
-		time.Sleep(60 * time.Second)
+		ginkgo.By("Waiting 5s for process outage simulation")
+		time.Sleep(5 * time.Second)
 
 		ginkgo.By("Continuing the stopped ovs-vswitchd process by sending a CONT signal")
 		cmd = []string{"sh", "-c", fmt.Sprintf(`"kill -CONT %s"`, pid)}

test/e2e/framework/daemonset.go

@@ -6,7 +6,6 @@ import (
 	"errors"
 	"fmt"
 	"strings"
-	"time"
 
 	appsv1 "k8s.io/api/apps/v1"
 	corev1 "k8s.io/api/core/v1"
@@ -17,6 +16,7 @@ import (
 	"k8s.io/apimachinery/pkg/util/wait"
 	clientset "k8s.io/client-go/kubernetes"
 	v1apps "k8s.io/client-go/kubernetes/typed/apps/v1"
+	"k8s.io/client-go/util/retry"
 	"k8s.io/kubectl/pkg/polymorphichelpers"
 
 	"github.com/onsi/ginkgo/v2"
@@ -100,7 +100,7 @@ func (c *DaemonSetClient) Patch(daemonset *appsv1.DaemonSet) *appsv1.DaemonSet {
 	}
 	ExpectNoError(err)
 	var patchedDaemonSet *appsv1.DaemonSet
-	err = wait.PollUntilContextTimeout(context.Background(), 2*time.Second, timeout, true, func(ctx context.Context) (bool, error) {
+	err = wait.PollUntilContextTimeout(context.Background(), poll, timeout, true, func(ctx context.Context) (bool, error) {
 		daemonSet, err := c.DaemonSetInterface.Patch(ctx, daemonset.Name, types.MergePatchType, modifiedBytes, metav1.PatchOptions{}, "")
 		if err != nil {
 			return handleWaitingAPIError(err, false, "patch daemonset %s/%s", daemonset.Namespace, daemonset.Name)
@@ -130,7 +130,7 @@ func (c *DaemonSetClient) RolloutStatus(name string) *appsv1.DaemonSet {
 	ginkgo.GinkgoHelper()
 
 	var daemonSet *appsv1.DaemonSet
-	WaitUntil(2*time.Second, timeout, func(_ context.Context) (bool, error) {
+	WaitUntil(poll, timeout, func(_ context.Context) (bool, error) {
 		var err error
 		daemonSet = c.Get(name)
 		unstructured := &unstructured.Unstructured{}
@@ -158,21 +158,31 @@ func (c *DaemonSetClient) RolloutStatus(name string) *appsv1.DaemonSet {
 func (c *DaemonSetClient) Restart(ds *appsv1.DaemonSet) *appsv1.DaemonSet {
 	ginkgo.GinkgoHelper()
 
-	buf, err := polymorphichelpers.ObjectRestarterFn(ds)
-	ExpectNoError(err)
+	var result *appsv1.DaemonSet
+	err := retry.RetryOnConflict(retry.DefaultBackoff, func() error {
+		latest := c.Get(ds.Name)
 
-	m := make(map[string]any)
-	err = json.Unmarshal(buf, &m)
-	ExpectNoError(err)
+		buf, err := polymorphichelpers.ObjectRestarterFn(latest)
+		if err != nil {
+			return err
+		}
 
-	ds = new(appsv1.DaemonSet)
-	err = runtime.DefaultUnstructuredConverter.FromUnstructured(m, ds)
-	ExpectNoError(err)
+		m := make(map[string]any)
+		if err = json.Unmarshal(buf, &m); err != nil {
+			return err
+		}
+
+		d := new(appsv1.DaemonSet)
+		if err = runtime.DefaultUnstructuredConverter.FromUnstructured(m, d); err != nil {
+			return err
+		}
 
-	ds, err = c.Update(context.TODO(), ds, metav1.UpdateOptions{})
+		result, err = c.Update(context.TODO(), d, metav1.UpdateOptions{})
+		return err
+	})
 	ExpectNoError(err)
 
-	return ds.DeepCopy()
+	return result.DeepCopy()
 }
 
 // RestartSync restarts the DaemonSet and wait it to be ready

test/e2e/framework/deployment.go

@@ -18,6 +18,7 @@ import (
 	"k8s.io/apimachinery/pkg/util/wait"
 	clientset "k8s.io/client-go/kubernetes"
 	v1apps "k8s.io/client-go/kubernetes/typed/apps/v1"
+	"k8s.io/client-go/util/retry"
 	"k8s.io/kubectl/pkg/polymorphichelpers"
 	"k8s.io/kubernetes/test/e2e/framework"
 	"k8s.io/kubernetes/test/e2e/framework/deployment"
@@ -98,7 +99,7 @@ func (c *DeploymentClient) RolloutStatus(name string) *appsv1.Deployment {
 	ginkgo.GinkgoHelper()
 
 	var deploy *appsv1.Deployment
-	WaitUntil(2*time.Second, timeout, func(_ context.Context) (bool, error) {
+	WaitUntil(poll, timeout, func(_ context.Context) (bool, error) {
 		var err error
 		deploy = c.Get(name)
 		unstructured := &unstructured.Unstructured{}
@@ -129,7 +130,7 @@ func (c *DeploymentClient) Patch(original, modified *appsv1.Deployment) *appsv1.
 	ExpectNoError(err)
 
 	var patchedDeploy *appsv1.Deployment
-	err = wait.PollUntilContextTimeout(context.Background(), 2*time.Second, timeout, true, func(ctx context.Context) (bool, error) {
+	err = wait.PollUntilContextTimeout(context.Background(), poll, timeout, true, func(ctx context.Context) (bool, error) {
 		deploy, err := c.DeploymentInterface.Patch(ctx, original.Name, types.StrategicMergePatchType, patch, metav1.PatchOptions{}, "")
 		if err != nil {
 			return handleWaitingAPIError(err, false, "patch deployment %s/%s", original.Namespace, original.Name)
@@ -159,21 +160,31 @@ func (c *DeploymentClient) PatchSync(original, modified *appsv1.Deployment) *app
 func (c *DeploymentClient) Restart(deploy *appsv1.Deployment) *appsv1.Deployment {
 	ginkgo.GinkgoHelper()
 
-	buf, err := polymorphichelpers.ObjectRestarterFn(deploy)
-	ExpectNoError(err)
+	var result *appsv1.Deployment
+	err := retry.RetryOnConflict(retry.DefaultBackoff, func() error {
+		latest := c.Get(deploy.Name)
 
-	m := make(map[string]any)
-	err = json.Unmarshal(buf, &m)
-	ExpectNoError(err)
+		buf, err := polymorphichelpers.ObjectRestarterFn(latest)
+		if err != nil {
+			return err
+		}
 
-	deploy = new(appsv1.Deployment)
-	err = runtime.DefaultUnstructuredConverter.FromUnstructured(m, deploy)
-	ExpectNoError(err)
+		m := make(map[string]any)
+		if err = json.Unmarshal(buf, &m); err != nil {
+			return err
+		}
+
+		d := new(appsv1.Deployment)
+		if err = runtime.DefaultUnstructuredConverter.FromUnstructured(m, d); err != nil {
+			return err
+		}
 
-	deploy, err = c.Update(context.TODO(), deploy, metav1.UpdateOptions{})
+		result, err = c.Update(context.TODO(), d, metav1.UpdateOptions{})
+		return err
+	})
 	ExpectNoError(err)
 
-	return deploy.DeepCopy()
+	return result.DeepCopy()
 }
 
 // RestartSync restarts the deployment and wait it to be ready
@@ -212,11 +223,11 @@ func (c *DeploymentClient) Delete(name string) {
 func (c *DeploymentClient) DeleteSync(name string) {
 	ginkgo.GinkgoHelper()
 	c.Delete(name)
-	gomega.Expect(c.WaitToDisappear(name, 2*time.Second, timeout)).To(gomega.Succeed(), "wait for deployment %q to disappear", name)
+	gomega.Expect(c.WaitToDisappear(name, poll, timeout)).To(gomega.Succeed(), "wait for deployment %q to disappear", name)
 }
 
 func (c *DeploymentClient) WaitToComplete(deploy *appsv1.Deployment) error {
-	return testutils.WaitForDeploymentComplete(c.clientSet, deploy, Logf, 2*time.Second, 2*time.Minute)
+	return testutils.WaitForDeploymentComplete(c.clientSet, deploy, Logf, poll, 2*time.Minute)
 }
 
 // WaitToDisappear waits the given timeout duration for the specified deployment to disappear.
@@ -238,5 +249,6 @@ func MakeDeployment(name string, replicas int32, podLabels, podAnnotations map[s
 	deploy := deployment.NewDeployment(name, replicas, podLabels, containerName, image, strategyType)
 	deploy.Spec.Template.Annotations = podAnnotations
 	deploy.Spec.Template.Spec.Containers[0].ImagePullPolicy = corev1.PullIfNotPresent
+	deploy.Spec.Template.Spec.TerminationGracePeriodSeconds = new(int64(1))
 	return deploy
 }

test/e2e/framework/dns-name-resolver.go

@@ -63,7 +63,7 @@ func (c *DNSNameResolverClient) WaitToBeReady(name string, timeout time.Duration
 	result := c.WaitUntil(name, func(dnsNameResolver *apiv1.DNSNameResolver) (bool, error) {
 		// DNSNameResolver is considered ready if it has at least one resolved name
 		return len(dnsNameResolver.Status.ResolvedNames) > 0, nil
-	}, "Ready", 2*time.Second, timeout)
+	}, "Ready", poll, timeout)
 	return result != nil
 }
 

test/e2e/framework/endpoints.go

@@ -58,7 +58,7 @@ func (c *EndpointsClient) Create(endpoints *corev1.Endpoints) *corev1.Endpoints
 func (c *EndpointsClient) CreateSync(endpoints *corev1.Endpoints, cond func(s *corev1.Endpoints) (bool, error), condDesc string) *corev1.Endpoints {
 	ginkgo.GinkgoHelper()
 	_ = c.Create(endpoints)
-	return c.WaitUntil(endpoints.Name, cond, condDesc, 2*time.Second, timeout)
+	return c.WaitUntil(endpoints.Name, cond, condDesc, poll, timeout)
 }
 
 // Patch patches the endpoints
@@ -69,7 +69,7 @@ func (c *EndpointsClient) Patch(original, modified *corev1.Endpoints) *corev1.En
 	ExpectNoError(err)
 
 	var patchedEndpoints *corev1.Endpoints
-	err = wait.PollUntilContextTimeout(context.Background(), 2*time.Second, timeout, true, func(_ context.Context) (bool, error) {
+	err = wait.PollUntilContextTimeout(context.Background(), poll, timeout, true, func(_ context.Context) (bool, error) {
 		s, err := c.EndpointsInterface.Patch(context.TODO(), original.Name, types.MergePatchType, patch, metav1.PatchOptions{}, "")
 		if err != nil {
 			return handleWaitingAPIError(err, false, "patch endpoints %q", original.Name)
@@ -93,7 +93,7 @@ func (c *EndpointsClient) Patch(original, modified *corev1.Endpoints) *corev1.En
 func (c *EndpointsClient) PatchSync(original, modified *corev1.Endpoints, cond func(s *corev1.Endpoints) (bool, error), condDesc string) *corev1.Endpoints {
 	ginkgo.GinkgoHelper()
 	_ = c.Patch(original, modified)
-	return c.WaitUntil(original.Name, cond, condDesc, 2*time.Second, timeout)
+	return c.WaitUntil(original.Name, cond, condDesc, poll, timeout)
 }
 
 // Delete deletes an Endpoints resource if it exists
@@ -110,15 +110,15 @@ func (c *EndpointsClient) Delete(name string) {
 func (c *EndpointsClient) DeleteSync(name string) {
 	ginkgo.GinkgoHelper()
 	c.Delete(name)
-	gomega.Expect(c.WaitToDisappear(name, 2*time.Second, timeout)).To(gomega.Succeed(), "wait for endpoints %q to disappear", name)
+	gomega.Expect(c.WaitToDisappear(name, poll, timeout)).To(gomega.Succeed(), "wait for endpoints %q to disappear", name)
 }
 
 // WaitUntil waits the given timeout duration for the specified condition to be met.
 func (c *EndpointsClient) WaitUntil(name string, cond func(s *corev1.Endpoints) (bool, error), condDesc string, _, timeout time.Duration) *corev1.Endpoints {
 	ginkgo.GinkgoHelper()
 
 	var endpoints *corev1.Endpoints
-	err := wait.PollUntilContextTimeout(context.Background(), 2*time.Second, timeout, true, func(_ context.Context) (bool, error) {
+	err := wait.PollUntilContextTimeout(context.Background(), poll, timeout, true, func(_ context.Context) (bool, error) {
 		Logf("Waiting for endpoints %s to meet condition %q", name, condDesc)
 		endpoints = c.Get(name).DeepCopy()
 		met, err := cond(endpoints)

test/e2e/framework/framework.go

@@ -38,7 +38,7 @@ const (
 
 const (
 	// poll is how often to Poll resources.
-	poll = 2 * time.Second
+	poll = time.Second
 
 	timeout = 2 * time.Minute
 )

test/e2e/framework/ip.go

@@ -81,7 +81,7 @@ func (c *IPClient) Patch(original, modified *apiv1.IP, timeout time.Duration) *a
 	ExpectNoError(err)
 
 	var patchedIP *apiv1.IP
-	err = wait.PollUntilContextTimeout(context.Background(), 2*time.Second, timeout, true, func(ctx context.Context) (bool, error) {
+	err = wait.PollUntilContextTimeout(context.Background(), poll, timeout, true, func(ctx context.Context) (bool, error) {
 		p, err := c.IPInterface.Patch(ctx, original.Name, types.MergePatchType, patch, metav1.PatchOptions{}, "")
 		if err != nil {
 			return handleWaitingAPIError(err, false, "patch IP %q", original.Name)
@@ -115,7 +115,7 @@ func (c *IPClient) Delete(name string) {
 func (c *IPClient) DeleteSync(name string) {
 	ginkgo.GinkgoHelper()
 	c.Delete(name)
-	gomega.Expect(c.WaitToDisappear(name, 2*time.Second, timeout)).To(gomega.Succeed(), "wait for ovn eip %q to disappear", name)
+	gomega.Expect(c.WaitToDisappear(name, poll, timeout)).To(gomega.Succeed(), "wait for ovn eip %q to disappear", name)
 }
 
 // WaitToDisappear waits the given timeout duration for the specified IP to disappear.

test/e2e/framework/ippool.go

@@ -64,7 +64,7 @@ func (c *IPPoolClient) Update(ippool *apiv1.IPPool, options metav1.UpdateOptions
 	ginkgo.GinkgoHelper()
 
 	var updatedIPPool *apiv1.IPPool
-	err := wait.PollUntilContextTimeout(context.Background(), 2*time.Second, timeout, true, func(ctx context.Context) (bool, error) {
+	err := wait.PollUntilContextTimeout(context.Background(), poll, timeout, true, func(ctx context.Context) (bool, error) {
 		s, err := c.IPPoolInterface.Update(ctx, ippool, options)
 		if err != nil {
 			// On conflict, refresh the resource and retry
@@ -113,7 +113,7 @@ func (c *IPPoolClient) Patch(original, modified *apiv1.IPPool, timeout time.Dura
 	ExpectNoError(err)
 
 	var patchedIPPool *apiv1.IPPool
-	err = wait.PollUntilContextTimeout(context.Background(), 2*time.Second, timeout, true, func(ctx context.Context) (bool, error) {
+	err = wait.PollUntilContextTimeout(context.Background(), poll, timeout, true, func(ctx context.Context) (bool, error) {
 		s, err := c.IPPoolInterface.Patch(ctx, original.Name, types.MergePatchType, patch, metav1.PatchOptions{}, "")
 		if err != nil {
 			return handleWaitingAPIError(err, false, "patch ippool %q", original.Name)
@@ -158,7 +158,7 @@ func (c *IPPoolClient) Delete(name string) {
 func (c *IPPoolClient) DeleteSync(name string) {
 	ginkgo.GinkgoHelper()
 	c.Delete(name)
-	gomega.Expect(c.WaitToDisappear(name, 2*time.Second, timeout)).To(gomega.Succeed(), "wait for ippool %q to disappear", name)
+	gomega.Expect(c.WaitToDisappear(name, poll, timeout)).To(gomega.Succeed(), "wait for ippool %q to disappear", name)
 }
 
 func isIPPoolConditionSetAsExpected(ippool *apiv1.IPPool, conditionType apiv1.ConditionType, wantTrue, silent bool) bool {